Reviews

Maladaptive alcohol use has emerged as a risk for a subset of individuals who have undergone weight loss surgery (WLS); studies report they are vulnerable to consuming alcohol in greater quantities or more frequently.^1,2 Estimates of the prevalence of “high-risk” or “hazardous” alcohol use after WLS range from 4% to 28%,^3,4 while the prevalence of alcohol use meeting DSM-IV-TR⁵ criteria for alcohol use disorders (AUDs) hovers around 10%.⁶

Heavy alcohol users or patients who have active AUD at the time of WLS are at greater risk for continuation of these problems after surgery.^2,6 For patients with a long-remitted history of AUD, the evidence regarding risk for post-WLS relapse is lacking, and some evidence suggests they may have better weight loss outcomes after WLS.⁷

However, approximately two-third of cases of post-WLS alcohol problems occur in patients who have had no history of such problems before surgery.^5,8,9 Reported prevalence rates of new-onset alcohol problems range from 3% to 18%,^6,9 with the modal finding being approximately 7% to 8%. New-onset alcohol problems appear to occur at a considerable latency after surgery. One study found little risk at 1 year post-surgery, but a significant increase in AUD symptoms at 2 years.⁶ Another study identified 3 years post-surgery as a high-risk time point,⁸ and yet another reported a linear increase in the risk for developing alcohol problems for at least 10 years after WLS.¹⁰

This article describes a group treatment protocol developed specifically for patients with post-WLS substance use disorder (SUD), and explores:

• risk factors and causal mechanisms of post-WLS AUDs
• weight stigma and emotional stressors
• the role of specialized treatment
• group treatment based on the Health at Every Size® (HAES)-oriented, trauma-informed and fat acceptance framework.

Post-WLS patients with alcohol problems may be a distinct phenotype among people with substance abuse issues. For this reason, they may have a need to address their experiences and issues specific to WLS as part of their alcohol treatment.

Etiology

Risk factors. Empirical findings have identified few predictors or risk factors for post-WLS SUD. These patients are more likely to be male and of a younger age.⁶ Notably, the vast majority of individuals reporting post-WLS alcohol problems have undergone Roux-en-Y gastric bypass (RYGB), rather than other WLS procedures, such as the laparoscopic adjustable gastric band,^6,11 suggesting some physiological mechanism specific to RYGB.

Other potential predictors of postoperative alcohol problems include a pre-operative history of depression, generalized anxiety disorder, smoking, and/or recreational drug use.^3,6 Likewise, patients with depression or anxiety disorder symptoms after surgery also may be at higher risk for postoperative alcohol problems.⁴ The evidence of an association between postoperative weight outcomes and post-WLS alcohol problems is mixed.^3,12 Interestingly, patients who had no personal history of substance abuse but who have a family history may have a higher risk of new-onset alcohol problems after surgery.^9,12

Causal mechanisms. The etiology of post-WLS alcohol problems is not well understood. If anything, epidemiological data suggest that larger-bodied individuals tend to consume lower levels of alcohol and have lower rates of AUD than individuals in the general population with thinner bodies.¹³ However, an association has been found between a family history of SUD, but not a personal history, and being large.¹⁴ This suggests a shared etiological pathway between addiction and being “overweight,” of which the onset of AUD after RYGB may be a manifestation.

Human and animal studies have shown that WLS may affect alcohol use differently in specific subgroups. Studies have shown that wild-type rats greatly increase their consumption of, or operant responding for, alcohol after RYGB,¹⁵ while genetically “alcohol-preferring” rats decrease consumption of, or responding for, alcohol after RYGB.¹⁶ A human study likewise found some patients decreased alcohol use or experienced improvement of or remission of AUD symptoms after WLS.⁴ Combined with the finding that a family history of substance abuse is related to risk for post-operative AUD, these data suggest a potential genetic vulnerability or protection in some individuals.

Turning to potential psychosocial explanations, the lay media has popularized the concepts of “addiction transfer,” or “transfer addiction,”¹² with the implication that some patients, who had a preoperative history of “food addiction,” transfer that “addiction” after surgery to substances of abuse.

However, the “addiction transfer” model has a number of flaws:

• it is stigmatizing, because it assumes the patient possesses an innate, chronic, and inalterable pathology
• it relies upon the validity of the controversial construct of “food addiction,” a construct of mixed scientific evidence.¹⁷

Further, our knowledge of post-WLS SUD argues against “addiction transfer.” As noted, postoperative alcohol problems are more likely to develop years after surgery, rather than in the first few months afterward when eating is most significantly curtailed. Additionally, post-WLS alcohol problems are significantly more likely to occur after RYGB than other procedures, whereas the “addiction transfer” model would hypothesize that all WLS patients would be at equal risk for postoperative “addiction transfer,” because their eating is similarly affected after surgery.

Links to RYGB. Some clues to physiological mechanisms underlying alcohol problems after RYGB have been identified. After surgery, many RYGB patients report a quicker effect from a smaller amount of alcohol than was the case pre-surgery.¹⁸ Studies have demonstrated a number of changes in the pharmacodynamics of alcohol after RYGB not seen in other WLS procedures¹⁹:

• a much faster time to peak blood (or breath) alcohol content (BAC)
• significantly higher peak BAC
• a precipitous initial decline in perceived intoxication.^18,20

Anatomical features of RYGB may explain such changes.⁸ However, an increased response to both IV alcohol and IV morphine after RYGB^21,22 in rodents suggests that gastrointestinal tract changes are not solely responsible for changes in alcohol use. Emerging research reports that WLS has been found to cause alterations in brain reward pathways,²³ which may be an additional contributor to changes in alcohol misuse after surgery.

However, even combined, pharmacokinetic and neurobiological factors cannot entirely explain new-onset alcohol problems after WLS; if they could, one would expect to see a much higher prevalence of this complication. Some psychosocial factors are likely involved as well.

Emotional stressors. One possibility involves a mismatch between post-WLS stressors and coping skills. After WLS, these patients face a multitude of challenges inherent in adjusting to changes in lifestyle, weight, body image, and social functioning, which most individuals would find daunting. These challenges become even more acute in the absence of appropriate psychoeducation, preparation, and intervention from qualified professionals. Individuals who lack effective and adaptive coping skills and supports may have a particularly heightened vulnerability to increased alcohol use in the setting of post-surgery changes in brain reward circuits and pharmacodynamics in alcohol metabolism. For example, one patient reported that her spouse’s pressure to “do something about her weight” was a significant factor in her decision to undergo surgery, but that her spouse was blaming and unsupportive when post-WLS complications developed. The patient believed that these experiences helped fuel development of her post-RYGB alcohol abuse.

Specialized treatment

The number of patients experiencing post-WLS alcohol problems likely will continue to grow, given that the risk of onset of has been shown increase over years. Already, post-WLS patients are proportionally overrepresented among substance abuse treatment populations.²⁴ Empirically, however, we do not know yet if these patients need a different type of addiction treatment than patients who have not had WLS.

Some evidence suggests that post-WLS patients with alcohol problems may be a distinct phenotype within the general population with alcohol problems, as their presentations differ in several ways, including their demographics, alcohol use patterns, and premorbid functioning. A number of studies have found that, despite their increased pharmacodynamic sensitivity to alcohol, people with post-WLS AUDs actually consume a larger amount of alcohol on both typical and maximum drinking days than other individuals with AUDs.²⁴ Additionally, although the median age of onset for AUD is around age 20,²⁵ patients presenting with new-onset, post-WLS alcohol problems are usually in their late 30s, or even 40s or 50s. Further, many of these patients were quite high functioning before their alcohol problems, and are unlikely to identify with the cultural stereotype of a person with AUD (eg, homeless, unemployed), which may hamper or delay their own willingness to accept that they have a problem. These phenotypic differences suggest that post-WLS patients may require substance abuse treatment approaches tailored to their unique presentation. There are additional factors specific to the experiences of being larger-bodied and WLS that also may need to be addressed in specialized treatment for post-WLS addiction patients.

Weight stigma. By definition, patients who have undergone WLS have spent a significant portion of their lives inhabiting larger bodies, an experience that, in our culture, can produce adverse psychosocial effects. Compared with the general population, patients seeking WLS exhibit psychological distress equivalent to psychiatric patients.²⁶ Weight stigma or weight bias—negative judgments directed toward people in larger bodies—is pervasive and continues to increase.²⁷ Further, evidence suggests that, unlike almost all other stigmatized groups, people in larger bodies tend to internalize this stigma, holding an unfavorable attitude toward their own social group.²⁸ Weight stigma impacts the well-being of people all along the weight spectrum, affecting many domains including educational achievements and classroom experiences, job opportunities, salaries, and medical care.²⁷ Weight stigma increases the likelihood of bullying, teasing, and harassment for both adults and children.²⁷ Weight bias has been associated with any number of adverse psychosocial effects, including symptoms of depression, anxiety, and eating pathology; poor body image; and a decrease in healthy self-care behaviors.^29-33

Weight stigma makes it more difficult for people to enjoy physical activities, nourish their bodies, and manage stress, which contributes to poorer health outcomes and lower quality of life.^33,34 For example, one study showed that, regardless of actual body mass index, people experiencing weight stigma have significantly increased risk of developing an illness or dying.³⁵

Factors specific to WLS. WLS may lead to significant changes in eating habits, and some patients experience a sense of loss, particularly if eating represented one of their primary coping strategies—this may represent a heightened emotional vulnerability for developing AUD.

The fairly rapid and substantial weight loss that WLS produces can lead to sweeping changes in lifestyle, body image, and functional factors for many individuals. Patients often report profound changes, both positive and negative, in their relationships and interactions not only with people in their support network, but also with strangers.³⁶

After the first year or 2 post-WLS, it is fairly common for patients to regain some weight, sometimes in significant amounts.³⁷ This can lead to a sense of “failure.” Life stressors, including difficulties in important relationships, can further add to patients’ vulnerability. For example, one patient noticed that when she was at her thinnest after WLS, drivers were more likely to stop for her when she crossed the street, which pleased but also angered her because they hadn’t extended the same courtesy before WLS. After she regained a significant amount of weight, she began to notice drivers stopping for her less and less frequently. This took her back to her previous feelings of being ignored but now with the certainty that she would be treated better if she were thinner.

Patients also may experience ambivalence about changes in their body size. One might expect that body image would improve after weight loss, but the evidence is mixed.³⁸ Although there is some evidence that body image improves in the short term after WLS,³⁸ other research indicates that body image does not improve with weight loss.³⁹ However, the evidence is clear that the appearance of excess skin after weight loss worsens some patients’ body image.⁴⁰

To date, there has been no research examining treatment modalities for this population. Because experiences common to individuals who have had WLS could play a role in the development of AUD after surgery, it is intuitive that it would be important to address these factors when designing a treatment plan for post-WLS substance abuse.

Group treatment approach

In 2013, in response to the increase in rates of post-WLS addictions presenting to West End Clinic, an outpatient dual-diagnosis (addiction and psychiatry) service at Massachusetts General Hospital, a specialized treatment group was developed. Nine patients have enrolled since October 2013.

The Post-WLS Addictions Group (PWAG) was designed to be HAES-oriented, trauma-informed, and run within a fat acceptance framework. The HAES model prioritizes a weight-neutral approach that sees health and well-being as multifaceted. This approach directs both patient and clinician to focus on improving health behaviors and reducing internalized weight bias, while building a supportive community that buffers against external cultural weight bias.⁴¹

Trauma-informed care⁴² emphasizes the principles of safety, trustworthiness, and transparency; peer support; collaboration and mutuality; empowerment; and awareness of cultural, historical, and gender issues. In the context of PWAG, weight stigma is conceptualized as a traumatic experience.⁴³ The fat acceptance approach promotes a culture that accepts people of every size with dignity and equality in all aspects of life.⁴⁴

Self-care emphasis. The HAES model encourages patients to allow their bodies to determine what weight to settle at, and to focus on sustainable health-enhancing behaviors rather than weight loss. Patients who asked about the PWAG were told that this group would not explicitly support, or even encourage, continued pursuit of weight loss per se, but instead would assist patients with relapse prevention, mindful eating, improving self-care, and ongoing stress management. Moving away from a focus on weight loss and toward improvement of self-care skills allowed patients to focus on behaviors and outcomes over which they had more direct control and were more likely to yield immediate benefits.

All of the PWAG group members were in early recovery from an SUD, with a minimum of 4 weeks of abstinence; all had at least 1 co-occurring mental health diagnosis. A licensed independent clinical social worker (LICSW) and a physician familiar with bariatric surgery ran the sessions. The group met weekly for 1 hour. The 8 weekly sessions included both psychoeducation and discussion, with each session covering different topics (Table). The first 20 minutes of each session were devoted to an educational presentation; the remaining 40 minutes for reflection and discussion. In sessions 2 through 8, participants were asked about any recent use or cravings, and problem-solving techniques were employed as needed.

Process and observations. Participants described high levels of excitement and hopefulness about being in a group with other WLS patients who had developed SUDs. They had a particular interest in reviewing medical/anatomical information about WLS and understanding more about the potential reasons for the elevated risk for developing SUD following WLS. Discussions regarding weight stigma proved to be quite emotional; most participants reported that this material readily related to their own experiences with weight stigma, but they had never discussed these ideas before.

Participants explored the role that grief, loss, guilt, and shame had in the decision to have WLS, the development of SUDs, weight regain or medical complications from the surgery or from substance abuse, career and relationship changes, and worsened body image. Another theme that emerged was the various reasons that prompted the members have WLS that they may not have been conscious of, or willing to discuss with others, such as pressure from a spouse, fears of remaining single due to their size, and a desire to finally “fit in.”

Repeatedly, group members expressed how satisfied and emotionally validated they felt being with people with similar experiences. Most of them had felt alone. They reported a belief that “everyone else” who had WLS was doing well, and that they were the exceptions. Such beliefs and emotions increased the risk of relapse and decreased participants’ ability to develop more positive coping strategies and self-care skills.

Participants reported that feeling less alone, understanding how stigma impacts health and well-being, and focusing on the general benefits of good self-care rather than the pursuit of weight loss were particularly helpful. The HAES and fat acceptance approaches have given group members new ways to think about their bodies and decreased shame. Several group members reported that if they had learned about the HAES approach prior to having a WLS, they might have made a different decision about having surgery, or at least might have been better prepared to handle the emotional and psychological challenges after WLS.

Although evidence for post-WLS addictions is fairly robust, causal mechanisms are not well understood, and research identifying specific risk factors is lacking. Because post-WLS patients with addictions seem to represent a specific phenotype, specialized treatment might be indicated. Future research will be needed to determine optimal treatment approaches for post-WLS addictions. However, a number of aspects are likely to be important. For example, it is likely that unaddressed experiences of weight stigma contribute to challenges, including substance abuse, after WLS; therefore, clinicians involved in the care of individuals presenting with post-WLS SUD should be knowledgeable about weight stigma and how to address it. Because of the specific nature of post-WLS addictions, patients often feel alone and isolated, and seem to benefit from the specialized group setting. We note that the PWAG group leader is herself a large person who models fat acceptance and follows the HAES approach, and therefore led the group using this experience and her specialized clinical training. As with other addiction recovery treatment modalities, clinicians who have lived the experience can add a valuable component to the program design and patient experience.

Maladaptive alcohol use has emerged as a risk for a subset of individuals who have undergone weight loss surgery (WLS); studies report they are vulnerable to consuming alcohol in greater quantities or more frequently.^1,2 Estimates of the prevalence of “high-risk” or “hazardous” alcohol use after WLS range from 4% to 28%,^3,4 while the prevalence of alcohol use meeting DSM-IV-TR⁵ criteria for alcohol use disorders (AUDs) hovers around 10%.⁶

Heavy alcohol users or patients who have active AUD at the time of WLS are at greater risk for continuation of these problems after surgery.^2,6 For patients with a long-remitted history of AUD, the evidence regarding risk for post-WLS relapse is lacking, and some evidence suggests they may have better weight loss outcomes after WLS.⁷

However, approximately two-third of cases of post-WLS alcohol problems occur in patients who have had no history of such problems before surgery.^5,8,9 Reported prevalence rates of new-onset alcohol problems range from 3% to 18%,^6,9 with the modal finding being approximately 7% to 8%. New-onset alcohol problems appear to occur at a considerable latency after surgery. One study found little risk at 1 year post-surgery, but a significant increase in AUD symptoms at 2 years.⁶ Another study identified 3 years post-surgery as a high-risk time point,⁸ and yet another reported a linear increase in the risk for developing alcohol problems for at least 10 years after WLS.¹⁰

This article describes a group treatment protocol developed specifically for patients with post-WLS substance use disorder (SUD), and explores:

• risk factors and causal mechanisms of post-WLS AUDs
• weight stigma and emotional stressors
• the role of specialized treatment
• group treatment based on the Health at Every Size® (HAES)-oriented, trauma-informed and fat acceptance framework.

Post-WLS patients with alcohol problems may be a distinct phenotype among people with substance abuse issues. For this reason, they may have a need to address their experiences and issues specific to WLS as part of their alcohol treatment.

Etiology

Risk factors. Empirical findings have identified few predictors or risk factors for post-WLS SUD. These patients are more likely to be male and of a younger age.⁶ Notably, the vast majority of individuals reporting post-WLS alcohol problems have undergone Roux-en-Y gastric bypass (RYGB), rather than other WLS procedures, such as the laparoscopic adjustable gastric band,^6,11 suggesting some physiological mechanism specific to RYGB.

Other potential predictors of postoperative alcohol problems include a pre-operative history of depression, generalized anxiety disorder, smoking, and/or recreational drug use.^3,6 Likewise, patients with depression or anxiety disorder symptoms after surgery also may be at higher risk for postoperative alcohol problems.⁴ The evidence of an association between postoperative weight outcomes and post-WLS alcohol problems is mixed.^3,12 Interestingly, patients who had no personal history of substance abuse but who have a family history may have a higher risk of new-onset alcohol problems after surgery.^9,12

Causal mechanisms. The etiology of post-WLS alcohol problems is not well understood. If anything, epidemiological data suggest that larger-bodied individuals tend to consume lower levels of alcohol and have lower rates of AUD than individuals in the general population with thinner bodies.¹³ However, an association has been found between a family history of SUD, but not a personal history, and being large.¹⁴ This suggests a shared etiological pathway between addiction and being “overweight,” of which the onset of AUD after RYGB may be a manifestation.

Human and animal studies have shown that WLS may affect alcohol use differently in specific subgroups. Studies have shown that wild-type rats greatly increase their consumption of, or operant responding for, alcohol after RYGB,¹⁵ while genetically “alcohol-preferring” rats decrease consumption of, or responding for, alcohol after RYGB.¹⁶ A human study likewise found some patients decreased alcohol use or experienced improvement of or remission of AUD symptoms after WLS.⁴ Combined with the finding that a family history of substance abuse is related to risk for post-operative AUD, these data suggest a potential genetic vulnerability or protection in some individuals.

Turning to potential psychosocial explanations, the lay media has popularized the concepts of “addiction transfer,” or “transfer addiction,”¹² with the implication that some patients, who had a preoperative history of “food addiction,” transfer that “addiction” after surgery to substances of abuse.

However, the “addiction transfer” model has a number of flaws:

• it is stigmatizing, because it assumes the patient possesses an innate, chronic, and inalterable pathology
• it relies upon the validity of the controversial construct of “food addiction,” a construct of mixed scientific evidence.¹⁷

Further, our knowledge of post-WLS SUD argues against “addiction transfer.” As noted, postoperative alcohol problems are more likely to develop years after surgery, rather than in the first few months afterward when eating is most significantly curtailed. Additionally, post-WLS alcohol problems are significantly more likely to occur after RYGB than other procedures, whereas the “addiction transfer” model would hypothesize that all WLS patients would be at equal risk for postoperative “addiction transfer,” because their eating is similarly affected after surgery.

Links to RYGB. Some clues to physiological mechanisms underlying alcohol problems after RYGB have been identified. After surgery, many RYGB patients report a quicker effect from a smaller amount of alcohol than was the case pre-surgery.¹⁸ Studies have demonstrated a number of changes in the pharmacodynamics of alcohol after RYGB not seen in other WLS procedures¹⁹:

• a much faster time to peak blood (or breath) alcohol content (BAC)
• significantly higher peak BAC
• a precipitous initial decline in perceived intoxication.^18,20

Anatomical features of RYGB may explain such changes.⁸ However, an increased response to both IV alcohol and IV morphine after RYGB^21,22 in rodents suggests that gastrointestinal tract changes are not solely responsible for changes in alcohol use. Emerging research reports that WLS has been found to cause alterations in brain reward pathways,²³ which may be an additional contributor to changes in alcohol misuse after surgery.

However, even combined, pharmacokinetic and neurobiological factors cannot entirely explain new-onset alcohol problems after WLS; if they could, one would expect to see a much higher prevalence of this complication. Some psychosocial factors are likely involved as well.

Emotional stressors. One possibility involves a mismatch between post-WLS stressors and coping skills. After WLS, these patients face a multitude of challenges inherent in adjusting to changes in lifestyle, weight, body image, and social functioning, which most individuals would find daunting. These challenges become even more acute in the absence of appropriate psychoeducation, preparation, and intervention from qualified professionals. Individuals who lack effective and adaptive coping skills and supports may have a particularly heightened vulnerability to increased alcohol use in the setting of post-surgery changes in brain reward circuits and pharmacodynamics in alcohol metabolism. For example, one patient reported that her spouse’s pressure to “do something about her weight” was a significant factor in her decision to undergo surgery, but that her spouse was blaming and unsupportive when post-WLS complications developed. The patient believed that these experiences helped fuel development of her post-RYGB alcohol abuse.

Specialized treatment

The number of patients experiencing post-WLS alcohol problems likely will continue to grow, given that the risk of onset of has been shown increase over years. Already, post-WLS patients are proportionally overrepresented among substance abuse treatment populations.²⁴ Empirically, however, we do not know yet if these patients need a different type of addiction treatment than patients who have not had WLS.

Some evidence suggests that post-WLS patients with alcohol problems may be a distinct phenotype within the general population with alcohol problems, as their presentations differ in several ways, including their demographics, alcohol use patterns, and premorbid functioning. A number of studies have found that, despite their increased pharmacodynamic sensitivity to alcohol, people with post-WLS AUDs actually consume a larger amount of alcohol on both typical and maximum drinking days than other individuals with AUDs.²⁴ Additionally, although the median age of onset for AUD is around age 20,²⁵ patients presenting with new-onset, post-WLS alcohol problems are usually in their late 30s, or even 40s or 50s. Further, many of these patients were quite high functioning before their alcohol problems, and are unlikely to identify with the cultural stereotype of a person with AUD (eg, homeless, unemployed), which may hamper or delay their own willingness to accept that they have a problem. These phenotypic differences suggest that post-WLS patients may require substance abuse treatment approaches tailored to their unique presentation. There are additional factors specific to the experiences of being larger-bodied and WLS that also may need to be addressed in specialized treatment for post-WLS addiction patients.

Weight stigma. By definition, patients who have undergone WLS have spent a significant portion of their lives inhabiting larger bodies, an experience that, in our culture, can produce adverse psychosocial effects. Compared with the general population, patients seeking WLS exhibit psychological distress equivalent to psychiatric patients.²⁶ Weight stigma or weight bias—negative judgments directed toward people in larger bodies—is pervasive and continues to increase.²⁷ Further, evidence suggests that, unlike almost all other stigmatized groups, people in larger bodies tend to internalize this stigma, holding an unfavorable attitude toward their own social group.²⁸ Weight stigma impacts the well-being of people all along the weight spectrum, affecting many domains including educational achievements and classroom experiences, job opportunities, salaries, and medical care.²⁷ Weight stigma increases the likelihood of bullying, teasing, and harassment for both adults and children.²⁷ Weight bias has been associated with any number of adverse psychosocial effects, including symptoms of depression, anxiety, and eating pathology; poor body image; and a decrease in healthy self-care behaviors.^29-33

Weight stigma makes it more difficult for people to enjoy physical activities, nourish their bodies, and manage stress, which contributes to poorer health outcomes and lower quality of life.^33,34 For example, one study showed that, regardless of actual body mass index, people experiencing weight stigma have significantly increased risk of developing an illness or dying.³⁵

Factors specific to WLS. WLS may lead to significant changes in eating habits, and some patients experience a sense of loss, particularly if eating represented one of their primary coping strategies—this may represent a heightened emotional vulnerability for developing AUD.

The fairly rapid and substantial weight loss that WLS produces can lead to sweeping changes in lifestyle, body image, and functional factors for many individuals. Patients often report profound changes, both positive and negative, in their relationships and interactions not only with people in their support network, but also with strangers.³⁶

After the first year or 2 post-WLS, it is fairly common for patients to regain some weight, sometimes in significant amounts.³⁷ This can lead to a sense of “failure.” Life stressors, including difficulties in important relationships, can further add to patients’ vulnerability. For example, one patient noticed that when she was at her thinnest after WLS, drivers were more likely to stop for her when she crossed the street, which pleased but also angered her because they hadn’t extended the same courtesy before WLS. After she regained a significant amount of weight, she began to notice drivers stopping for her less and less frequently. This took her back to her previous feelings of being ignored but now with the certainty that she would be treated better if she were thinner.

Patients also may experience ambivalence about changes in their body size. One might expect that body image would improve after weight loss, but the evidence is mixed.³⁸ Although there is some evidence that body image improves in the short term after WLS,³⁸ other research indicates that body image does not improve with weight loss.³⁹ However, the evidence is clear that the appearance of excess skin after weight loss worsens some patients’ body image.⁴⁰

To date, there has been no research examining treatment modalities for this population. Because experiences common to individuals who have had WLS could play a role in the development of AUD after surgery, it is intuitive that it would be important to address these factors when designing a treatment plan for post-WLS substance abuse.

Group treatment approach

In 2013, in response to the increase in rates of post-WLS addictions presenting to West End Clinic, an outpatient dual-diagnosis (addiction and psychiatry) service at Massachusetts General Hospital, a specialized treatment group was developed. Nine patients have enrolled since October 2013.

The Post-WLS Addictions Group (PWAG) was designed to be HAES-oriented, trauma-informed, and run within a fat acceptance framework. The HAES model prioritizes a weight-neutral approach that sees health and well-being as multifaceted. This approach directs both patient and clinician to focus on improving health behaviors and reducing internalized weight bias, while building a supportive community that buffers against external cultural weight bias.⁴¹

Trauma-informed care⁴² emphasizes the principles of safety, trustworthiness, and transparency; peer support; collaboration and mutuality; empowerment; and awareness of cultural, historical, and gender issues. In the context of PWAG, weight stigma is conceptualized as a traumatic experience.⁴³ The fat acceptance approach promotes a culture that accepts people of every size with dignity and equality in all aspects of life.⁴⁴

Self-care emphasis. The HAES model encourages patients to allow their bodies to determine what weight to settle at, and to focus on sustainable health-enhancing behaviors rather than weight loss. Patients who asked about the PWAG were told that this group would not explicitly support, or even encourage, continued pursuit of weight loss per se, but instead would assist patients with relapse prevention, mindful eating, improving self-care, and ongoing stress management. Moving away from a focus on weight loss and toward improvement of self-care skills allowed patients to focus on behaviors and outcomes over which they had more direct control and were more likely to yield immediate benefits.

All of the PWAG group members were in early recovery from an SUD, with a minimum of 4 weeks of abstinence; all had at least 1 co-occurring mental health diagnosis. A licensed independent clinical social worker (LICSW) and a physician familiar with bariatric surgery ran the sessions. The group met weekly for 1 hour. The 8 weekly sessions included both psychoeducation and discussion, with each session covering different topics (Table). The first 20 minutes of each session were devoted to an educational presentation; the remaining 40 minutes for reflection and discussion. In sessions 2 through 8, participants were asked about any recent use or cravings, and problem-solving techniques were employed as needed.

Process and observations. Participants described high levels of excitement and hopefulness about being in a group with other WLS patients who had developed SUDs. They had a particular interest in reviewing medical/anatomical information about WLS and understanding more about the potential reasons for the elevated risk for developing SUD following WLS. Discussions regarding weight stigma proved to be quite emotional; most participants reported that this material readily related to their own experiences with weight stigma, but they had never discussed these ideas before.

Participants explored the role that grief, loss, guilt, and shame had in the decision to have WLS, the development of SUDs, weight regain or medical complications from the surgery or from substance abuse, career and relationship changes, and worsened body image. Another theme that emerged was the various reasons that prompted the members have WLS that they may not have been conscious of, or willing to discuss with others, such as pressure from a spouse, fears of remaining single due to their size, and a desire to finally “fit in.”

Repeatedly, group members expressed how satisfied and emotionally validated they felt being with people with similar experiences. Most of them had felt alone. They reported a belief that “everyone else” who had WLS was doing well, and that they were the exceptions. Such beliefs and emotions increased the risk of relapse and decreased participants’ ability to develop more positive coping strategies and self-care skills.

Participants reported that feeling less alone, understanding how stigma impacts health and well-being, and focusing on the general benefits of good self-care rather than the pursuit of weight loss were particularly helpful. The HAES and fat acceptance approaches have given group members new ways to think about their bodies and decreased shame. Several group members reported that if they had learned about the HAES approach prior to having a WLS, they might have made a different decision about having surgery, or at least might have been better prepared to handle the emotional and psychological challenges after WLS.

Although evidence for post-WLS addictions is fairly robust, causal mechanisms are not well understood, and research identifying specific risk factors is lacking. Because post-WLS patients with addictions seem to represent a specific phenotype, specialized treatment might be indicated. Future research will be needed to determine optimal treatment approaches for post-WLS addictions. However, a number of aspects are likely to be important. For example, it is likely that unaddressed experiences of weight stigma contribute to challenges, including substance abuse, after WLS; therefore, clinicians involved in the care of individuals presenting with post-WLS SUD should be knowledgeable about weight stigma and how to address it. Because of the specific nature of post-WLS addictions, patients often feel alone and isolated, and seem to benefit from the specialized group setting. We note that the PWAG group leader is herself a large person who models fat acceptance and follows the HAES approach, and therefore led the group using this experience and her specialized clinical training. As with other addiction recovery treatment modalities, clinicians who have lived the experience can add a valuable component to the program design and patient experience.

Maladaptive alcohol use has emerged as a risk for a subset of individuals who have undergone weight loss surgery (WLS); studies report they are vulnerable to consuming alcohol in greater quantities or more frequently.^1,2 Estimates of the prevalence of “high-risk” or “hazardous” alcohol use after WLS range from 4% to 28%,^3,4 while the prevalence of alcohol use meeting DSM-IV-TR⁵ criteria for alcohol use disorders (AUDs) hovers around 10%.⁶

Heavy alcohol users or patients who have active AUD at the time of WLS are at greater risk for continuation of these problems after surgery.^2,6 For patients with a long-remitted history of AUD, the evidence regarding risk for post-WLS relapse is lacking, and some evidence suggests they may have better weight loss outcomes after WLS.⁷

However, approximately two-third of cases of post-WLS alcohol problems occur in patients who have had no history of such problems before surgery.^5,8,9 Reported prevalence rates of new-onset alcohol problems range from 3% to 18%,^6,9 with the modal finding being approximately 7% to 8%. New-onset alcohol problems appear to occur at a considerable latency after surgery. One study found little risk at 1 year post-surgery, but a significant increase in AUD symptoms at 2 years.⁶ Another study identified 3 years post-surgery as a high-risk time point,⁸ and yet another reported a linear increase in the risk for developing alcohol problems for at least 10 years after WLS.¹⁰

This article describes a group treatment protocol developed specifically for patients with post-WLS substance use disorder (SUD), and explores:

• risk factors and causal mechanisms of post-WLS AUDs
• weight stigma and emotional stressors
• the role of specialized treatment
• group treatment based on the Health at Every Size® (HAES)-oriented, trauma-informed and fat acceptance framework.

Post-WLS patients with alcohol problems may be a distinct phenotype among people with substance abuse issues. For this reason, they may have a need to address their experiences and issues specific to WLS as part of their alcohol treatment.

Etiology

Risk factors. Empirical findings have identified few predictors or risk factors for post-WLS SUD. These patients are more likely to be male and of a younger age.⁶ Notably, the vast majority of individuals reporting post-WLS alcohol problems have undergone Roux-en-Y gastric bypass (RYGB), rather than other WLS procedures, such as the laparoscopic adjustable gastric band,^6,11 suggesting some physiological mechanism specific to RYGB.

Other potential predictors of postoperative alcohol problems include a pre-operative history of depression, generalized anxiety disorder, smoking, and/or recreational drug use.^3,6 Likewise, patients with depression or anxiety disorder symptoms after surgery also may be at higher risk for postoperative alcohol problems.⁴ The evidence of an association between postoperative weight outcomes and post-WLS alcohol problems is mixed.^3,12 Interestingly, patients who had no personal history of substance abuse but who have a family history may have a higher risk of new-onset alcohol problems after surgery.^9,12

Causal mechanisms. The etiology of post-WLS alcohol problems is not well understood. If anything, epidemiological data suggest that larger-bodied individuals tend to consume lower levels of alcohol and have lower rates of AUD than individuals in the general population with thinner bodies.¹³ However, an association has been found between a family history of SUD, but not a personal history, and being large.¹⁴ This suggests a shared etiological pathway between addiction and being “overweight,” of which the onset of AUD after RYGB may be a manifestation.

Human and animal studies have shown that WLS may affect alcohol use differently in specific subgroups. Studies have shown that wild-type rats greatly increase their consumption of, or operant responding for, alcohol after RYGB,¹⁵ while genetically “alcohol-preferring” rats decrease consumption of, or responding for, alcohol after RYGB.¹⁶ A human study likewise found some patients decreased alcohol use or experienced improvement of or remission of AUD symptoms after WLS.⁴ Combined with the finding that a family history of substance abuse is related to risk for post-operative AUD, these data suggest a potential genetic vulnerability or protection in some individuals.

Turning to potential psychosocial explanations, the lay media has popularized the concepts of “addiction transfer,” or “transfer addiction,”¹² with the implication that some patients, who had a preoperative history of “food addiction,” transfer that “addiction” after surgery to substances of abuse.

However, the “addiction transfer” model has a number of flaws:

• it is stigmatizing, because it assumes the patient possesses an innate, chronic, and inalterable pathology
• it relies upon the validity of the controversial construct of “food addiction,” a construct of mixed scientific evidence.¹⁷

Further, our knowledge of post-WLS SUD argues against “addiction transfer.” As noted, postoperative alcohol problems are more likely to develop years after surgery, rather than in the first few months afterward when eating is most significantly curtailed. Additionally, post-WLS alcohol problems are significantly more likely to occur after RYGB than other procedures, whereas the “addiction transfer” model would hypothesize that all WLS patients would be at equal risk for postoperative “addiction transfer,” because their eating is similarly affected after surgery.

Links to RYGB. Some clues to physiological mechanisms underlying alcohol problems after RYGB have been identified. After surgery, many RYGB patients report a quicker effect from a smaller amount of alcohol than was the case pre-surgery.¹⁸ Studies have demonstrated a number of changes in the pharmacodynamics of alcohol after RYGB not seen in other WLS procedures¹⁹:

• a much faster time to peak blood (or breath) alcohol content (BAC)
• significantly higher peak BAC
• a precipitous initial decline in perceived intoxication.^18,20

Anatomical features of RYGB may explain such changes.⁸ However, an increased response to both IV alcohol and IV morphine after RYGB^21,22 in rodents suggests that gastrointestinal tract changes are not solely responsible for changes in alcohol use. Emerging research reports that WLS has been found to cause alterations in brain reward pathways,²³ which may be an additional contributor to changes in alcohol misuse after surgery.

However, even combined, pharmacokinetic and neurobiological factors cannot entirely explain new-onset alcohol problems after WLS; if they could, one would expect to see a much higher prevalence of this complication. Some psychosocial factors are likely involved as well.

Emotional stressors. One possibility involves a mismatch between post-WLS stressors and coping skills. After WLS, these patients face a multitude of challenges inherent in adjusting to changes in lifestyle, weight, body image, and social functioning, which most individuals would find daunting. These challenges become even more acute in the absence of appropriate psychoeducation, preparation, and intervention from qualified professionals. Individuals who lack effective and adaptive coping skills and supports may have a particularly heightened vulnerability to increased alcohol use in the setting of post-surgery changes in brain reward circuits and pharmacodynamics in alcohol metabolism. For example, one patient reported that her spouse’s pressure to “do something about her weight” was a significant factor in her decision to undergo surgery, but that her spouse was blaming and unsupportive when post-WLS complications developed. The patient believed that these experiences helped fuel development of her post-RYGB alcohol abuse.

Specialized treatment

The number of patients experiencing post-WLS alcohol problems likely will continue to grow, given that the risk of onset of has been shown increase over years. Already, post-WLS patients are proportionally overrepresented among substance abuse treatment populations.²⁴ Empirically, however, we do not know yet if these patients need a different type of addiction treatment than patients who have not had WLS.

Some evidence suggests that post-WLS patients with alcohol problems may be a distinct phenotype within the general population with alcohol problems, as their presentations differ in several ways, including their demographics, alcohol use patterns, and premorbid functioning. A number of studies have found that, despite their increased pharmacodynamic sensitivity to alcohol, people with post-WLS AUDs actually consume a larger amount of alcohol on both typical and maximum drinking days than other individuals with AUDs.²⁴ Additionally, although the median age of onset for AUD is around age 20,²⁵ patients presenting with new-onset, post-WLS alcohol problems are usually in their late 30s, or even 40s or 50s. Further, many of these patients were quite high functioning before their alcohol problems, and are unlikely to identify with the cultural stereotype of a person with AUD (eg, homeless, unemployed), which may hamper or delay their own willingness to accept that they have a problem. These phenotypic differences suggest that post-WLS patients may require substance abuse treatment approaches tailored to their unique presentation. There are additional factors specific to the experiences of being larger-bodied and WLS that also may need to be addressed in specialized treatment for post-WLS addiction patients.

Weight stigma. By definition, patients who have undergone WLS have spent a significant portion of their lives inhabiting larger bodies, an experience that, in our culture, can produce adverse psychosocial effects. Compared with the general population, patients seeking WLS exhibit psychological distress equivalent to psychiatric patients.²⁶ Weight stigma or weight bias—negative judgments directed toward people in larger bodies—is pervasive and continues to increase.²⁷ Further, evidence suggests that, unlike almost all other stigmatized groups, people in larger bodies tend to internalize this stigma, holding an unfavorable attitude toward their own social group.²⁸ Weight stigma impacts the well-being of people all along the weight spectrum, affecting many domains including educational achievements and classroom experiences, job opportunities, salaries, and medical care.²⁷ Weight stigma increases the likelihood of bullying, teasing, and harassment for both adults and children.²⁷ Weight bias has been associated with any number of adverse psychosocial effects, including symptoms of depression, anxiety, and eating pathology; poor body image; and a decrease in healthy self-care behaviors.^29-33

Weight stigma makes it more difficult for people to enjoy physical activities, nourish their bodies, and manage stress, which contributes to poorer health outcomes and lower quality of life.^33,34 For example, one study showed that, regardless of actual body mass index, people experiencing weight stigma have significantly increased risk of developing an illness or dying.³⁵

Factors specific to WLS. WLS may lead to significant changes in eating habits, and some patients experience a sense of loss, particularly if eating represented one of their primary coping strategies—this may represent a heightened emotional vulnerability for developing AUD.

The fairly rapid and substantial weight loss that WLS produces can lead to sweeping changes in lifestyle, body image, and functional factors for many individuals. Patients often report profound changes, both positive and negative, in their relationships and interactions not only with people in their support network, but also with strangers.³⁶

After the first year or 2 post-WLS, it is fairly common for patients to regain some weight, sometimes in significant amounts.³⁷ This can lead to a sense of “failure.” Life stressors, including difficulties in important relationships, can further add to patients’ vulnerability. For example, one patient noticed that when she was at her thinnest after WLS, drivers were more likely to stop for her when she crossed the street, which pleased but also angered her because they hadn’t extended the same courtesy before WLS. After she regained a significant amount of weight, she began to notice drivers stopping for her less and less frequently. This took her back to her previous feelings of being ignored but now with the certainty that she would be treated better if she were thinner.

Patients also may experience ambivalence about changes in their body size. One might expect that body image would improve after weight loss, but the evidence is mixed.³⁸ Although there is some evidence that body image improves in the short term after WLS,³⁸ other research indicates that body image does not improve with weight loss.³⁹ However, the evidence is clear that the appearance of excess skin after weight loss worsens some patients’ body image.⁴⁰

To date, there has been no research examining treatment modalities for this population. Because experiences common to individuals who have had WLS could play a role in the development of AUD after surgery, it is intuitive that it would be important to address these factors when designing a treatment plan for post-WLS substance abuse.

Group treatment approach

In 2013, in response to the increase in rates of post-WLS addictions presenting to West End Clinic, an outpatient dual-diagnosis (addiction and psychiatry) service at Massachusetts General Hospital, a specialized treatment group was developed. Nine patients have enrolled since October 2013.

The Post-WLS Addictions Group (PWAG) was designed to be HAES-oriented, trauma-informed, and run within a fat acceptance framework. The HAES model prioritizes a weight-neutral approach that sees health and well-being as multifaceted. This approach directs both patient and clinician to focus on improving health behaviors and reducing internalized weight bias, while building a supportive community that buffers against external cultural weight bias.⁴¹

Trauma-informed care⁴² emphasizes the principles of safety, trustworthiness, and transparency; peer support; collaboration and mutuality; empowerment; and awareness of cultural, historical, and gender issues. In the context of PWAG, weight stigma is conceptualized as a traumatic experience.⁴³ The fat acceptance approach promotes a culture that accepts people of every size with dignity and equality in all aspects of life.⁴⁴

Self-care emphasis. The HAES model encourages patients to allow their bodies to determine what weight to settle at, and to focus on sustainable health-enhancing behaviors rather than weight loss. Patients who asked about the PWAG were told that this group would not explicitly support, or even encourage, continued pursuit of weight loss per se, but instead would assist patients with relapse prevention, mindful eating, improving self-care, and ongoing stress management. Moving away from a focus on weight loss and toward improvement of self-care skills allowed patients to focus on behaviors and outcomes over which they had more direct control and were more likely to yield immediate benefits.

All of the PWAG group members were in early recovery from an SUD, with a minimum of 4 weeks of abstinence; all had at least 1 co-occurring mental health diagnosis. A licensed independent clinical social worker (LICSW) and a physician familiar with bariatric surgery ran the sessions. The group met weekly for 1 hour. The 8 weekly sessions included both psychoeducation and discussion, with each session covering different topics (Table). The first 20 minutes of each session were devoted to an educational presentation; the remaining 40 minutes for reflection and discussion. In sessions 2 through 8, participants were asked about any recent use or cravings, and problem-solving techniques were employed as needed.

Process and observations. Participants described high levels of excitement and hopefulness about being in a group with other WLS patients who had developed SUDs. They had a particular interest in reviewing medical/anatomical information about WLS and understanding more about the potential reasons for the elevated risk for developing SUD following WLS. Discussions regarding weight stigma proved to be quite emotional; most participants reported that this material readily related to their own experiences with weight stigma, but they had never discussed these ideas before.

Participants explored the role that grief, loss, guilt, and shame had in the decision to have WLS, the development of SUDs, weight regain or medical complications from the surgery or from substance abuse, career and relationship changes, and worsened body image. Another theme that emerged was the various reasons that prompted the members have WLS that they may not have been conscious of, or willing to discuss with others, such as pressure from a spouse, fears of remaining single due to their size, and a desire to finally “fit in.”

Repeatedly, group members expressed how satisfied and emotionally validated they felt being with people with similar experiences. Most of them had felt alone. They reported a belief that “everyone else” who had WLS was doing well, and that they were the exceptions. Such beliefs and emotions increased the risk of relapse and decreased participants’ ability to develop more positive coping strategies and self-care skills.

Participants reported that feeling less alone, understanding how stigma impacts health and well-being, and focusing on the general benefits of good self-care rather than the pursuit of weight loss were particularly helpful. The HAES and fat acceptance approaches have given group members new ways to think about their bodies and decreased shame. Several group members reported that if they had learned about the HAES approach prior to having a WLS, they might have made a different decision about having surgery, or at least might have been better prepared to handle the emotional and psychological challenges after WLS.

Although evidence for post-WLS addictions is fairly robust, causal mechanisms are not well understood, and research identifying specific risk factors is lacking. Because post-WLS patients with addictions seem to represent a specific phenotype, specialized treatment might be indicated. Future research will be needed to determine optimal treatment approaches for post-WLS addictions. However, a number of aspects are likely to be important. For example, it is likely that unaddressed experiences of weight stigma contribute to challenges, including substance abuse, after WLS; therefore, clinicians involved in the care of individuals presenting with post-WLS SUD should be knowledgeable about weight stigma and how to address it. Because of the specific nature of post-WLS addictions, patients often feel alone and isolated, and seem to benefit from the specialized group setting. We note that the PWAG group leader is herself a large person who models fat acceptance and follows the HAES approach, and therefore led the group using this experience and her specialized clinical training. As with other addiction recovery treatment modalities, clinicians who have lived the experience can add a valuable component to the program design and patient experience.

Sexually transmitted infections (STIs) continue to be a significant public health problem with potentially serious complications.¹ The incidence of new STIs, including viral STIs, in the United States is estimated at 19 million cases per year.²Chlamydia trachomatis remains the most common bacterial STI with an estimated annual incidence of 2.8 million cases in the United States and 50 million worldwide. Second in prevalence is gonococcal infection. Herpes simplex virus is one of the most common viral STIs, but the incidence of human papillomavirus virus (HPV), which is associated with cervical cancer, has steadily increased worldwide.³ Young persons age 15 to 24 are at the highest risk of acquiring new STIs with almost 50% of new cases reported among this age group.⁴

STIs can have serious complications and sequelae. For example, 20% to 40% of women who have chlamydia infections and 10% to 20% of women who have gonococcal infections develop pelvic inflammatory disease (PID),² which increases the risk for ectopic pregnancy, infertility, and chronic pelvic pain.

Patients with mental illness are at high risk of acquiring STIs. In the United States, the prevalence of HIV among patients with psychiatric illness is 10 to 20 times higher than in the general population.^4,5 Factors contributing to increased vulnerability to STIs among psychiatric patients include:

• impaired autonomy
• increased impulsivity
• increased susceptibility to coerced sex.⁶
  

Furthermore, a higher incidence of poverty, placement in risky environments, and overall poor health and medical care also contribute to the high prevalence of STIs and their complications in this population (Table 1). Because of risk factors specific to psychiatric illness, standard STI prevention interventions are not always successful and novel and innovative behavioral approaches are necessary.⁷

Case Abdominal pain and fever

Ms. K, age 25, has a history of bipolar disorder treated with lithium and presents to the community psychiatrist with lower abdominal pain. She recently recovered from a manic episode and has started to reintegrate with the community mental health team. She refuses to see her primary care physician and is adamant that she wishes to see her psychiatrist, who is the only doctor she has rapport with.

Ms. K reports lower abdominal pain for 3 or 4 days and fever for 1 day. The pain is dull in character. She denies diarrhea, vomiting, or urinary symptoms, but on further questioning describes new-onset, foul-smelling vaginal discharge without vaginal bleeding. Her menstrual cycle usually is regular, but her last menstrual period occurred 2 months ago. Her medical history includes an appendectomy at age 10 and she is a current cigarette smoker. Chart notes taken during her manic episode describe high-risk behavior, including having unprotected sexual intercourse with several partners. On examination, she is febrile and tachycardic with a tender lower abdomen.

Diagnosing STIs

To diagnose an STI, first a clinician must consider its likelihood. Taking a thorough sexual history allows assessment of the need for further investigation and provides an opportunity to discuss risk reduction. In accordance with recent guidelines,⁸ all health care providers are encouraged to consider the sexual history a routine aspect of the clinical encounter. The Centers for Disease Control and Prevention’s (CDC’s) “Five Ps” approach (Table 2) is an excellent tool for guiding investigation and counseling.⁹

The Figure provides health care providers with an algorithm to guide testing for STIs among psychiatric patients. Note that chlamydia, gonorrhea, syphilis, chancroid, viral hepatitis, and HIV must be reported to state public health agencies and the CDC.

Modern laboratory techniques make diagnosing STIs easier. Analysis of urine or serum reduces the need for invasive sampling. If swabs are required for diagnosis, patient self-collection of urethral, vulvovaginal, rectal, or pharyngeal specimens is as accurate as clinician collected samples and is better tolerated.⁸ Because of variation in diagnostic assays, we recommend contacting the laboratory before sending non-standard samples to ensure accurate collection and analysis.

Guidelines for preventing and screening for STIs

There are no prevention guidelines for STIs specific to the psychiatric population, although there is a clear need for focused intervention in this vulnerable patient group.¹⁰ Rates of STI screening generally are low in the psychiatric setting,¹¹ which results in a considerable burden of disease. All psychiatric patients should be encouraged to engage with STI screening programs that are in line with national guidelines. In the inpatient psychiatric or medical environment, clinicians have a responsibility to ensure that STI screening is considered for each patient.

Patients with mental illness should be assumed to be sexually active, even if they do not volunteer this information to clinicians. Employ a low threshold for recommending safer sex practices including condom use. Encourage women to develop a relationship with a family practitioner, internist, or gynecologist. Advise men who have sex with men (MSM) to visit a doctor regularly for screening of HIV and rectal, anal, and oral STIs as behavior and symptoms dictate.

HPV. Routine HPV screening is not recommended; however, 2 vaccines are available to prevent oncogenic HPV (types 16 and 18). All females age 13 to 26 should receive 3 doses of HPV vaccine over a 6-month period. The quadrivalent vaccine (Gardasil) also protects against HPV types 6 and 11, which cause 90% of genital warts and is preferred when available. Males age 9 to 26 also can receive the vaccine, although ideally it should be administered before sexual activity begins.¹⁵ Women still should attend routine cervical cancer screening even if they have the vaccine because 30% of cervical cancers are not caused by HPV 16/18. However, this means that 70% of cervical cancers are associated with HPV 16/18, making screening and the vaccine an important public health initiative. There also is a link between HPV and oral cancers.

Treating STIs among mentally ill individuals

Treatment of STIs among mentally ill individuals is important to prevent medical complications and to reduce transmission. Here are a few additional questions to keep in mind when treating a patient with psychiatric illness:

Does the patient have a primary psychiatric disorder, or is the patient’s current psychiatric presentation a result of the infection?
Some STIs can manifest with psychiatric symptoms—for example, neurosyphilis and HIV-associated neurocognitive disorders—and pose a diagnostic challenge. Obtaining a longitudinal history of the patient’s mental health, age of onset, and family history can help clarify the cause.

Are there any psychiatric adverse effects of STI treatment?

Most drugs used for treating common STIs are not known to cause psychiatric adverse effects (See the American Psychiatric Association¹⁶ and Sockalingham et al¹⁷ for a thorough discussion of HIV and hepatitis C treatment). The exception is fluoroquinolones, which could be prescribed for PID if cephalosporin therapy is not feasible. CNS effects of fluoroquinolones include insomnia, restlessness, confusion, and, in rare cases, mania and psychosis.

What are possible medication interactions to keep in mind when treating a psychiatric patient?
Nonsteroidal anti-inflammatory drugs (NSAIDs), other than sulindac, could increase serum lithium levels. Although NSAIDs are not contraindicated in patients taking lithium, other pain relievers, such as acetaminophen, may be preferred as a first-line choice.

Carbamazepine could lower serum levels of doxycycline.¹⁸
Azithromycin and other macrolides, as well as fluoroquinolones, could have QTc prolonging effects and has been associated with torsades de pointes.¹⁹ Several psychiatric medications, in particular, atypical antipsychotics, also could prolong the QTc interval. This could be a consideration in patients with underlying long QT intervals at baseline or a family history of sudden cardiac death.

Psychiatric patients might refuse or not adhere to their medication. Refusals could be the result of grandiose delusions (“I don’t need treatment”) or paranoia (“The doctor is trying to poison me”). Consider 1-time doses of antibiotics that can be given in the clinic for uncomplicated infections when adherence is an issue. Because psychiatric patients are at higher risk for acquiring STIs, education and counseling—especially substance abuse counseling—are vital as both primary and secondary prevention strategies. Treatment of STIs should be accompanied by referrals to the social work team or a therapist when appropriate.

Finally, as with any proposed treatment, it is important to consider whether the patient has capacity to consent to or refuse treatment. To assess for capacity, a patient must be able to:

• communicate a choice
• understand the relevant information
• appreciate the medical consequences of the decision
• demonstrate the ability to reason about treatment choices.²⁰

Case continued

In the emergency department, Ms. K’s vital signs are: temperature 39.5°C; pulse 110 beats per minute; blood pressure 96/67 mm Hg; and breathing 20 respirations per minute. She complains of nausea and has 2 episodes of emesis. She allows clinicians to perform a complete physical examination, including pelvic exam. Her cervix is inflamed, and she is noted to have adnexal and cervical motion tenderness.

Labs and imaging confirm a diagnosis of PID due to gonorrhea and she is admitted to the hospital for IV antibiotics. She continues to experience nausea and vomiting, but also complains of dizziness and diarrhea. Her speech is slurred and a coarse tremor is noticed in her hands. Renal function tests show slight impairment, probably due to dehydration. A pregnancy test is negative.

Lithium is held. Her nausea, vomiting, and diarrhea resolve quickly, and Ms. K asks to leave. When she is told that she is not ready for discharge, Ms. K becomes upset and rips out her IV yelling, “I don’t need treatment from you guys!” A psychiatry consult is called to assess for her capacity to refuse treatment. The team determines that she has capacity, but she becomes agreeable to remaining in the hospital after a phone conversation with her community mental health team.

Ms. K improves with antibiotic treatment. HIV and syphilis serology tests are negative. Before discharge, both the community psychiatrist and her primary care physicians are informed her lithium was held during hospitalization and restarted before discharge. Ms. K also is educated about the signs and symptoms of lithium toxicity, as well as common STIs.

Clinical considerations

• Physicians should have a low threshold of suspicion for PID in a sexually active young woman who presents with abdominal pain and shuffling gait, which is a natural attempt to reduce cervical irritation and is associated with PID.
• Ask about sexual history and symptoms of STIs.
• Rule out STIs in men presenting with urinary tract infections.
• If chlamydia is diagnosed, treatment for gonorrhea also is essential, and vice versa.
• Always think about HIV and hepatatis B and C in a patient with a STI.
• Treatment with single-dose medications can be effective.
• Risk of STIs is higher during episodes of mania or psychosis.
• Consider hospitalization if medically indicated or if you suspect non-adherence to therapy. It is important to remember that all kinds of systemic infections—including PID—can result in dehydration and alter renal metabolism leading to lithium accumulation.
• Mentally ill patients might require placement under involuntary commitment if they are found to be a danger to themselves or others. It is important to liaise with both the community psychiatry team and primary care physician both during hospitalization and before discharge to ensure a smooth transition.

Sexually transmitted infections (STIs) continue to be a significant public health problem with potentially serious complications.¹ The incidence of new STIs, including viral STIs, in the United States is estimated at 19 million cases per year.²Chlamydia trachomatis remains the most common bacterial STI with an estimated annual incidence of 2.8 million cases in the United States and 50 million worldwide. Second in prevalence is gonococcal infection. Herpes simplex virus is one of the most common viral STIs, but the incidence of human papillomavirus virus (HPV), which is associated with cervical cancer, has steadily increased worldwide.³ Young persons age 15 to 24 are at the highest risk of acquiring new STIs with almost 50% of new cases reported among this age group.⁴

STIs can have serious complications and sequelae. For example, 20% to 40% of women who have chlamydia infections and 10% to 20% of women who have gonococcal infections develop pelvic inflammatory disease (PID),² which increases the risk for ectopic pregnancy, infertility, and chronic pelvic pain.

Patients with mental illness are at high risk of acquiring STIs. In the United States, the prevalence of HIV among patients with psychiatric illness is 10 to 20 times higher than in the general population.^4,5 Factors contributing to increased vulnerability to STIs among psychiatric patients include:

• impaired autonomy
• increased impulsivity
• increased susceptibility to coerced sex.⁶
  

Furthermore, a higher incidence of poverty, placement in risky environments, and overall poor health and medical care also contribute to the high prevalence of STIs and their complications in this population (Table 1). Because of risk factors specific to psychiatric illness, standard STI prevention interventions are not always successful and novel and innovative behavioral approaches are necessary.⁷

Case Abdominal pain and fever

Ms. K, age 25, has a history of bipolar disorder treated with lithium and presents to the community psychiatrist with lower abdominal pain. She recently recovered from a manic episode and has started to reintegrate with the community mental health team. She refuses to see her primary care physician and is adamant that she wishes to see her psychiatrist, who is the only doctor she has rapport with.

Ms. K reports lower abdominal pain for 3 or 4 days and fever for 1 day. The pain is dull in character. She denies diarrhea, vomiting, or urinary symptoms, but on further questioning describes new-onset, foul-smelling vaginal discharge without vaginal bleeding. Her menstrual cycle usually is regular, but her last menstrual period occurred 2 months ago. Her medical history includes an appendectomy at age 10 and she is a current cigarette smoker. Chart notes taken during her manic episode describe high-risk behavior, including having unprotected sexual intercourse with several partners. On examination, she is febrile and tachycardic with a tender lower abdomen.

Diagnosing STIs

To diagnose an STI, first a clinician must consider its likelihood. Taking a thorough sexual history allows assessment of the need for further investigation and provides an opportunity to discuss risk reduction. In accordance with recent guidelines,⁸ all health care providers are encouraged to consider the sexual history a routine aspect of the clinical encounter. The Centers for Disease Control and Prevention’s (CDC’s) “Five Ps” approach (Table 2) is an excellent tool for guiding investigation and counseling.⁹

The Figure provides health care providers with an algorithm to guide testing for STIs among psychiatric patients. Note that chlamydia, gonorrhea, syphilis, chancroid, viral hepatitis, and HIV must be reported to state public health agencies and the CDC.

Modern laboratory techniques make diagnosing STIs easier. Analysis of urine or serum reduces the need for invasive sampling. If swabs are required for diagnosis, patient self-collection of urethral, vulvovaginal, rectal, or pharyngeal specimens is as accurate as clinician collected samples and is better tolerated.⁸ Because of variation in diagnostic assays, we recommend contacting the laboratory before sending non-standard samples to ensure accurate collection and analysis.

Guidelines for preventing and screening for STIs

There are no prevention guidelines for STIs specific to the psychiatric population, although there is a clear need for focused intervention in this vulnerable patient group.¹⁰ Rates of STI screening generally are low in the psychiatric setting,¹¹ which results in a considerable burden of disease. All psychiatric patients should be encouraged to engage with STI screening programs that are in line with national guidelines. In the inpatient psychiatric or medical environment, clinicians have a responsibility to ensure that STI screening is considered for each patient.

Patients with mental illness should be assumed to be sexually active, even if they do not volunteer this information to clinicians. Employ a low threshold for recommending safer sex practices including condom use. Encourage women to develop a relationship with a family practitioner, internist, or gynecologist. Advise men who have sex with men (MSM) to visit a doctor regularly for screening of HIV and rectal, anal, and oral STIs as behavior and symptoms dictate.

HPV. Routine HPV screening is not recommended; however, 2 vaccines are available to prevent oncogenic HPV (types 16 and 18). All females age 13 to 26 should receive 3 doses of HPV vaccine over a 6-month period. The quadrivalent vaccine (Gardasil) also protects against HPV types 6 and 11, which cause 90% of genital warts and is preferred when available. Males age 9 to 26 also can receive the vaccine, although ideally it should be administered before sexual activity begins.¹⁵ Women still should attend routine cervical cancer screening even if they have the vaccine because 30% of cervical cancers are not caused by HPV 16/18. However, this means that 70% of cervical cancers are associated with HPV 16/18, making screening and the vaccine an important public health initiative. There also is a link between HPV and oral cancers.

Treating STIs among mentally ill individuals

Treatment of STIs among mentally ill individuals is important to prevent medical complications and to reduce transmission. Here are a few additional questions to keep in mind when treating a patient with psychiatric illness:

Does the patient have a primary psychiatric disorder, or is the patient’s current psychiatric presentation a result of the infection?
Some STIs can manifest with psychiatric symptoms—for example, neurosyphilis and HIV-associated neurocognitive disorders—and pose a diagnostic challenge. Obtaining a longitudinal history of the patient’s mental health, age of onset, and family history can help clarify the cause.

Are there any psychiatric adverse effects of STI treatment?

Most drugs used for treating common STIs are not known to cause psychiatric adverse effects (See the American Psychiatric Association¹⁶ and Sockalingham et al¹⁷ for a thorough discussion of HIV and hepatitis C treatment). The exception is fluoroquinolones, which could be prescribed for PID if cephalosporin therapy is not feasible. CNS effects of fluoroquinolones include insomnia, restlessness, confusion, and, in rare cases, mania and psychosis.

What are possible medication interactions to keep in mind when treating a psychiatric patient?
Nonsteroidal anti-inflammatory drugs (NSAIDs), other than sulindac, could increase serum lithium levels. Although NSAIDs are not contraindicated in patients taking lithium, other pain relievers, such as acetaminophen, may be preferred as a first-line choice.

Carbamazepine could lower serum levels of doxycycline.¹⁸
Azithromycin and other macrolides, as well as fluoroquinolones, could have QTc prolonging effects and has been associated with torsades de pointes.¹⁹ Several psychiatric medications, in particular, atypical antipsychotics, also could prolong the QTc interval. This could be a consideration in patients with underlying long QT intervals at baseline or a family history of sudden cardiac death.

Psychiatric patients might refuse or not adhere to their medication. Refusals could be the result of grandiose delusions (“I don’t need treatment”) or paranoia (“The doctor is trying to poison me”). Consider 1-time doses of antibiotics that can be given in the clinic for uncomplicated infections when adherence is an issue. Because psychiatric patients are at higher risk for acquiring STIs, education and counseling—especially substance abuse counseling—are vital as both primary and secondary prevention strategies. Treatment of STIs should be accompanied by referrals to the social work team or a therapist when appropriate.

Finally, as with any proposed treatment, it is important to consider whether the patient has capacity to consent to or refuse treatment. To assess for capacity, a patient must be able to:

• communicate a choice
• understand the relevant information
• appreciate the medical consequences of the decision
• demonstrate the ability to reason about treatment choices.²⁰

Case continued

In the emergency department, Ms. K’s vital signs are: temperature 39.5°C; pulse 110 beats per minute; blood pressure 96/67 mm Hg; and breathing 20 respirations per minute. She complains of nausea and has 2 episodes of emesis. She allows clinicians to perform a complete physical examination, including pelvic exam. Her cervix is inflamed, and she is noted to have adnexal and cervical motion tenderness.

Labs and imaging confirm a diagnosis of PID due to gonorrhea and she is admitted to the hospital for IV antibiotics. She continues to experience nausea and vomiting, but also complains of dizziness and diarrhea. Her speech is slurred and a coarse tremor is noticed in her hands. Renal function tests show slight impairment, probably due to dehydration. A pregnancy test is negative.

Lithium is held. Her nausea, vomiting, and diarrhea resolve quickly, and Ms. K asks to leave. When she is told that she is not ready for discharge, Ms. K becomes upset and rips out her IV yelling, “I don’t need treatment from you guys!” A psychiatry consult is called to assess for her capacity to refuse treatment. The team determines that she has capacity, but she becomes agreeable to remaining in the hospital after a phone conversation with her community mental health team.

Ms. K improves with antibiotic treatment. HIV and syphilis serology tests are negative. Before discharge, both the community psychiatrist and her primary care physicians are informed her lithium was held during hospitalization and restarted before discharge. Ms. K also is educated about the signs and symptoms of lithium toxicity, as well as common STIs.

Clinical considerations

• Physicians should have a low threshold of suspicion for PID in a sexually active young woman who presents with abdominal pain and shuffling gait, which is a natural attempt to reduce cervical irritation and is associated with PID.
• Ask about sexual history and symptoms of STIs.
• Rule out STIs in men presenting with urinary tract infections.
• If chlamydia is diagnosed, treatment for gonorrhea also is essential, and vice versa.
• Always think about HIV and hepatatis B and C in a patient with a STI.
• Treatment with single-dose medications can be effective.
• Risk of STIs is higher during episodes of mania or psychosis.
• Consider hospitalization if medically indicated or if you suspect non-adherence to therapy. It is important to remember that all kinds of systemic infections—including PID—can result in dehydration and alter renal metabolism leading to lithium accumulation.
• Mentally ill patients might require placement under involuntary commitment if they are found to be a danger to themselves or others. It is important to liaise with both the community psychiatry team and primary care physician both during hospitalization and before discharge to ensure a smooth transition.

Sexually transmitted infections (STIs) continue to be a significant public health problem with potentially serious complications.¹ The incidence of new STIs, including viral STIs, in the United States is estimated at 19 million cases per year.²Chlamydia trachomatis remains the most common bacterial STI with an estimated annual incidence of 2.8 million cases in the United States and 50 million worldwide. Second in prevalence is gonococcal infection. Herpes simplex virus is one of the most common viral STIs, but the incidence of human papillomavirus virus (HPV), which is associated with cervical cancer, has steadily increased worldwide.³ Young persons age 15 to 24 are at the highest risk of acquiring new STIs with almost 50% of new cases reported among this age group.⁴

STIs can have serious complications and sequelae. For example, 20% to 40% of women who have chlamydia infections and 10% to 20% of women who have gonococcal infections develop pelvic inflammatory disease (PID),² which increases the risk for ectopic pregnancy, infertility, and chronic pelvic pain.

Patients with mental illness are at high risk of acquiring STIs. In the United States, the prevalence of HIV among patients with psychiatric illness is 10 to 20 times higher than in the general population.^4,5 Factors contributing to increased vulnerability to STIs among psychiatric patients include:

• impaired autonomy
• increased impulsivity
• increased susceptibility to coerced sex.⁶
  

Furthermore, a higher incidence of poverty, placement in risky environments, and overall poor health and medical care also contribute to the high prevalence of STIs and their complications in this population (Table 1). Because of risk factors specific to psychiatric illness, standard STI prevention interventions are not always successful and novel and innovative behavioral approaches are necessary.⁷

Case Abdominal pain and fever

Ms. K, age 25, has a history of bipolar disorder treated with lithium and presents to the community psychiatrist with lower abdominal pain. She recently recovered from a manic episode and has started to reintegrate with the community mental health team. She refuses to see her primary care physician and is adamant that she wishes to see her psychiatrist, who is the only doctor she has rapport with.

Ms. K reports lower abdominal pain for 3 or 4 days and fever for 1 day. The pain is dull in character. She denies diarrhea, vomiting, or urinary symptoms, but on further questioning describes new-onset, foul-smelling vaginal discharge without vaginal bleeding. Her menstrual cycle usually is regular, but her last menstrual period occurred 2 months ago. Her medical history includes an appendectomy at age 10 and she is a current cigarette smoker. Chart notes taken during her manic episode describe high-risk behavior, including having unprotected sexual intercourse with several partners. On examination, she is febrile and tachycardic with a tender lower abdomen.

Diagnosing STIs

To diagnose an STI, first a clinician must consider its likelihood. Taking a thorough sexual history allows assessment of the need for further investigation and provides an opportunity to discuss risk reduction. In accordance with recent guidelines,⁸ all health care providers are encouraged to consider the sexual history a routine aspect of the clinical encounter. The Centers for Disease Control and Prevention’s (CDC’s) “Five Ps” approach (Table 2) is an excellent tool for guiding investigation and counseling.⁹

The Figure provides health care providers with an algorithm to guide testing for STIs among psychiatric patients. Note that chlamydia, gonorrhea, syphilis, chancroid, viral hepatitis, and HIV must be reported to state public health agencies and the CDC.

Modern laboratory techniques make diagnosing STIs easier. Analysis of urine or serum reduces the need for invasive sampling. If swabs are required for diagnosis, patient self-collection of urethral, vulvovaginal, rectal, or pharyngeal specimens is as accurate as clinician collected samples and is better tolerated.⁸ Because of variation in diagnostic assays, we recommend contacting the laboratory before sending non-standard samples to ensure accurate collection and analysis.

Guidelines for preventing and screening for STIs

There are no prevention guidelines for STIs specific to the psychiatric population, although there is a clear need for focused intervention in this vulnerable patient group.¹⁰ Rates of STI screening generally are low in the psychiatric setting,¹¹ which results in a considerable burden of disease. All psychiatric patients should be encouraged to engage with STI screening programs that are in line with national guidelines. In the inpatient psychiatric or medical environment, clinicians have a responsibility to ensure that STI screening is considered for each patient.

Patients with mental illness should be assumed to be sexually active, even if they do not volunteer this information to clinicians. Employ a low threshold for recommending safer sex practices including condom use. Encourage women to develop a relationship with a family practitioner, internist, or gynecologist. Advise men who have sex with men (MSM) to visit a doctor regularly for screening of HIV and rectal, anal, and oral STIs as behavior and symptoms dictate.

HPV. Routine HPV screening is not recommended; however, 2 vaccines are available to prevent oncogenic HPV (types 16 and 18). All females age 13 to 26 should receive 3 doses of HPV vaccine over a 6-month period. The quadrivalent vaccine (Gardasil) also protects against HPV types 6 and 11, which cause 90% of genital warts and is preferred when available. Males age 9 to 26 also can receive the vaccine, although ideally it should be administered before sexual activity begins.¹⁵ Women still should attend routine cervical cancer screening even if they have the vaccine because 30% of cervical cancers are not caused by HPV 16/18. However, this means that 70% of cervical cancers are associated with HPV 16/18, making screening and the vaccine an important public health initiative. There also is a link between HPV and oral cancers.

Treating STIs among mentally ill individuals

Treatment of STIs among mentally ill individuals is important to prevent medical complications and to reduce transmission. Here are a few additional questions to keep in mind when treating a patient with psychiatric illness:

Does the patient have a primary psychiatric disorder, or is the patient’s current psychiatric presentation a result of the infection?
Some STIs can manifest with psychiatric symptoms—for example, neurosyphilis and HIV-associated neurocognitive disorders—and pose a diagnostic challenge. Obtaining a longitudinal history of the patient’s mental health, age of onset, and family history can help clarify the cause.

Are there any psychiatric adverse effects of STI treatment?

Most drugs used for treating common STIs are not known to cause psychiatric adverse effects (See the American Psychiatric Association¹⁶ and Sockalingham et al¹⁷ for a thorough discussion of HIV and hepatitis C treatment). The exception is fluoroquinolones, which could be prescribed for PID if cephalosporin therapy is not feasible. CNS effects of fluoroquinolones include insomnia, restlessness, confusion, and, in rare cases, mania and psychosis.

What are possible medication interactions to keep in mind when treating a psychiatric patient?
Nonsteroidal anti-inflammatory drugs (NSAIDs), other than sulindac, could increase serum lithium levels. Although NSAIDs are not contraindicated in patients taking lithium, other pain relievers, such as acetaminophen, may be preferred as a first-line choice.

Carbamazepine could lower serum levels of doxycycline.¹⁸
Azithromycin and other macrolides, as well as fluoroquinolones, could have QTc prolonging effects and has been associated with torsades de pointes.¹⁹ Several psychiatric medications, in particular, atypical antipsychotics, also could prolong the QTc interval. This could be a consideration in patients with underlying long QT intervals at baseline or a family history of sudden cardiac death.

Psychiatric patients might refuse or not adhere to their medication. Refusals could be the result of grandiose delusions (“I don’t need treatment”) or paranoia (“The doctor is trying to poison me”). Consider 1-time doses of antibiotics that can be given in the clinic for uncomplicated infections when adherence is an issue. Because psychiatric patients are at higher risk for acquiring STIs, education and counseling—especially substance abuse counseling—are vital as both primary and secondary prevention strategies. Treatment of STIs should be accompanied by referrals to the social work team or a therapist when appropriate.

Finally, as with any proposed treatment, it is important to consider whether the patient has capacity to consent to or refuse treatment. To assess for capacity, a patient must be able to:

• communicate a choice
• understand the relevant information
• appreciate the medical consequences of the decision
• demonstrate the ability to reason about treatment choices.²⁰

Case continued

In the emergency department, Ms. K’s vital signs are: temperature 39.5°C; pulse 110 beats per minute; blood pressure 96/67 mm Hg; and breathing 20 respirations per minute. She complains of nausea and has 2 episodes of emesis. She allows clinicians to perform a complete physical examination, including pelvic exam. Her cervix is inflamed, and she is noted to have adnexal and cervical motion tenderness.

Labs and imaging confirm a diagnosis of PID due to gonorrhea and she is admitted to the hospital for IV antibiotics. She continues to experience nausea and vomiting, but also complains of dizziness and diarrhea. Her speech is slurred and a coarse tremor is noticed in her hands. Renal function tests show slight impairment, probably due to dehydration. A pregnancy test is negative.

Lithium is held. Her nausea, vomiting, and diarrhea resolve quickly, and Ms. K asks to leave. When she is told that she is not ready for discharge, Ms. K becomes upset and rips out her IV yelling, “I don’t need treatment from you guys!” A psychiatry consult is called to assess for her capacity to refuse treatment. The team determines that she has capacity, but she becomes agreeable to remaining in the hospital after a phone conversation with her community mental health team.

Ms. K improves with antibiotic treatment. HIV and syphilis serology tests are negative. Before discharge, both the community psychiatrist and her primary care physicians are informed her lithium was held during hospitalization and restarted before discharge. Ms. K also is educated about the signs and symptoms of lithium toxicity, as well as common STIs.

Clinical considerations

• Physicians should have a low threshold of suspicion for PID in a sexually active young woman who presents with abdominal pain and shuffling gait, which is a natural attempt to reduce cervical irritation and is associated with PID.
• Ask about sexual history and symptoms of STIs.
• Rule out STIs in men presenting with urinary tract infections.
• If chlamydia is diagnosed, treatment for gonorrhea also is essential, and vice versa.
• Always think about HIV and hepatatis B and C in a patient with a STI.
• Treatment with single-dose medications can be effective.
• Risk of STIs is higher during episodes of mania or psychosis.
• Consider hospitalization if medically indicated or if you suspect non-adherence to therapy. It is important to remember that all kinds of systemic infections—including PID—can result in dehydration and alter renal metabolism leading to lithium accumulation.
• Mentally ill patients might require placement under involuntary commitment if they are found to be a danger to themselves or others. It is important to liaise with both the community psychiatry team and primary care physician both during hospitalization and before discharge to ensure a smooth transition.

INTRODUCTION

Renal cell carcinoma (RCC) is the most common malignancy arising in the kidney, comprising 90% of all renal tumors.¹ Approximately 55,000 new RCC cases are diagnosed each year.² Patients with RCC are often asymptomatic, and most cases are discovered as incidental findings on abdominal imaging performed during evaluation of nonrenal complaints. Limited-stage RCC that is found early can be cured surgically, with estimated 5-year survival rates approaching 90%; however, long-term survival for metastatic disease is poor, with rates ranging from 0% to 20%.² Advanced RCC is resistant to conventional chemotherapy and radiotherapy, and outcomes for patients with metastatic or unresectable RCC remain poor. However, the recent development of new therapeutic modalities that target tumor molecular pathways has expanded the treatment options for these patients and changed the management of RCC.

EPIDEMIOLOGY AND CLASSIFICATION

Median age at diagnosis in the United States is 64 years. Men have a higher incidence of RCC than women, with the highest incidence seen in American Indian and Alaska Native men (30.1 per 100,000 population). Genetic syndromes account for 2% to 4% of all RCCs.² Risk factors for RCC include smoking, hypertension, obesity, and acquired cystic kidney disease that is associated with end-stage renal failure.³ Longer duration of tobacco use is associated with a more aggressive course.

The 2004 World Health Organization (WHO) classification of renal tumors summarizes the previous classification systems (including the Heidelberg and Mainz classification systems) to describe different categories of RCC based on histologic and molecular genetics characteristics.² Using the WHO classification criteria, RCC comprises 90% of all renal tumors, with clear cell being the most common type (80%).² Other types of renal tumors include papillary, chromophobe, oncocytoma, and collecting-duct or Bellini duct tumors. Approximately 3% to 5% of tumors are unclassified. Oncocytomas are generally considered benign, and chromophobe tumors typically have an indolent course and rarely metastasize. Sarcomatoid differentiation can be seen in any histologic type and is associated with a worse prognosis. While different types of tumors may be seen in the kidney (such as transitional cell or lymphomas), the focus of this review is the primary malignancies of the renal parenchyma.

FAMILIAL SYNDROMES

Several genetic syndromes have been identified by studying families with inherited RCC. Among these, von Hippel-Lindau (VHL) gene mutation is the most commonly found inherited genetic defect. Table 1 summarizes the incidence of gene mutations and the corresponding histologic appearance of the most common sporadic and hereditary RCCs.⁴

VHL disease is an autosomal dominant familial syndrome. Patients with this mutation are at higher risk for developing RCC (clear cell histology), retinal angiomas, pheochromocytomas, as well as hemangioblastomas of the central nervous system (CNS).⁴ Of all the genetic mutations seen in RCC, the somatic mutation in the VHL tumor-suppressor gene is by far the most common.⁵ VHL targets hypoxia–inducible factor-1 alpha (HIF-α) for ubiquitination and subsequent degradation, which has been shown to suppress the growth of clear-cell RCC in mouse models.^6–8 HIF expression under hypoxic conditions leads to activation of a number of genes important in blood vessel development, cell proliferation, and glucose metabolism, including vascular endothelial growth factor (VEGF), erythropoietin, platelet-derived growth factor beta (PDGF-β), transforming growth factor alpha (TGF-α), and glucose transporter-1 (GLUT-1). Mutation in the VHL gene prevents degradation of the HIF-α protein, thereby leading to increased expression of these downstream proteins, including MET and Axl. The upregulation of these angiogenic factors is thought to be the underlying process for increased vascularity of CNS hemangioblastomas and clear-cell renal tumors in VHL disease.^4–8

Other less common genetic syndromes seen in hereditary RCC include hereditary papillary RCC, hereditary leiomyomatosis, and Birt-Hogg-Dubé (BHD) syndrome.⁹ In hereditary papillary RCC, the MET gene is mutated. BHD syndrome is a rare, autosomal dominant syndrome characterized by hair follicle hamartomas of the face and neck. About 15% of patients have multiple renal tumors, the majority of which are of the chromophobe or mixed chromophobe-oncocytoma histology. The BHD gene encodes the protein folliculin, which is thought to be a tumor-suppressor gene.

DIAGNOSIS AND STAGING

CASE PRESENTATION

A 74-year-old man who works as an airplane mechanic repairman presents to the emergency department with sudden worsening of chronic right upper arm and shoulder pain after lifting a jug of orange juice. He does not have a significant past medical history and initially thought that his pain was due to a work-related injury. Upon initial evaluation in the emergency department he is found to have a fracture of his right humerus. Given that the fracture appears to be pathologic, further work-up is recommended.

Most patients are asymptomatic until the disease becomes advanced. The classic triad of flank pain, hematuria, and palpable abdominal mass is seen in approximately 10% of patients with RCC, partly because of earlier detection of renal masses by imaging performed for other purposes.¹⁰ Less frequently, patients present with signs or symptoms of metastatic disease such as bone pain or fracture (as seen in the case patient), painful adenopathy, and pulmonary symptoms related to mediastinal masses. Fever, weight loss, anemia, and/or varicocele often occur in young patients (≤ 46 years) and may indicate the presence of a hereditary form of the disease. Patients may present with paraneoplastic syndromes seen as abnormalities on routine blood work. These can include polycythemia or elevated liver function tests (LFTs) without the presence of liver metastases (known as Stauffer syndrome), which can be seen in localized renal tumors. Nearly half (45%) of patients present with localized disease, 25% present with locally advanced disease, and 30% present with metastatic disease.¹¹ Bone is the second most common site of distant metastatic spread (following lung) in patients with advanced RCC.

• What is the approach to initial evaluation for a patient with suspected RCC?

Initial evaluation consists of a physical exam, laboratory tests including complete blood count (CBC) and comprehensive metabolic panel (calcium, serum creatinine, LFTs, lactate dehydrogenase [LDH], and urinalysis), and imaging. Imaging studies include computed tomography (CT) scan with contrast of the abdomen and pelvis or magnetic resonance imaging (MRI) of the abdomen and chest imaging. A chest radiograph may be obtained, although a chest CT is more sensitive for the presence of pulmonary metastases. MRI can be used in patients with renal dysfunction to evaluate the renal vein and inferior vena cava (IVC) for thrombus or to determine the presence of local invasion.¹² Although bone and brain are common sites for metastases, routine imaging is not indicated unless the patient is symptomatic. The value of positron emission tomography in RCC remains undetermined at this time.

Staging is done according to the American Joint Committee on Cancer (AJCC) staging classification for RCC; the Figure summarizes the staging and 5-year survival data based on this classification scheme.^4,13

LIMITED-STAGE DISEASE

• What are the therapeutic options for limited-stage disease?

For patients with nondistant metastases, or limited-stage disease, surgical intervention with curative intent is considered. Convention suggests considering definitive surgery for patients with stage I and II disease, select patients with stage III disease with pathologically enlarged retroperitoneal lymph nodes, patients with IVC and/or cardiac atrium involvement of tumor thrombus, and patients with direct extension of the renal tumor into the ipsilateral adrenal gland if there is no evidence of distant disease. While there may be a role for aggressive surgical intervention in patients with distant metastatic disease, this topic will not be covered in this review.

SURGICAL INTERVENTION

Once patients are determined to be appropriate candidates for surgical removal of a renal tumor, the urologist will perform either a radical nephrectomy or a nephron-sparing nephrectomy, also called a partial nephrectomy. The urologist will evaluate the patient based on his or her body habitus, the location of the tumor, whether multiple tumors in one kidney or bilateral tumors are present, whether the patient has a solitary kidney or otherwise impaired kidney function, and whether the patient has a history of a hereditary syndrome involving kidney cancer as this affects the risk of future kidney tumors.

A radical nephrectomy is surgically preferred in the presence of the following factors: tumor larger than 7 cm in diameter, a more centrally located tumor, suspicion of lymph node involvement, tumor involvement with renal vein or IVC, and/or direct extension of the tumor into the ipsilateral adrenal gland. Nephrectomy involves ligation of the vascular supply (renal artery and vein) followed by removal of the kidney and surrounding Gerota’s fascia. The ipsilateral adrenal gland is removed if there is a high-risk for or presence of invasion of the adrenal gland. Removal of the adrenal gland is not standard since the literature demonstrates there is less than a 10% chance of solitary, ipsilateral adrenal gland involvement of tumor at the time of nephrectomy in the absence of high-risk features, and a recent systematic review suggests that the chance may be as low as 1.8%.¹⁴ Preoperative factors that correlated with adrenal involvement included upper pole kidney location, renal vein thrombosis, higher T stage (T3a and greater), multifocal tumors, and evidence for distant metastases or lymph node involvement. Lymphadenectomy previously had been included in radical nephrectomy but now is performed selectively. Radical nephrectomy may be performed as

Conversely, a nephron-sparing approach is preferred for tumors less than 7 cm in diameter, for patients with a solitary kidney or impaired renal function, for patients with multiple small ipsilateral tumors or with bilateral tumors, or for radical nephrectomy candidates with comorbidities for whom a limited intervention is deemed to be a lower-risk procedure. A nephron-sparing procedure may also be performed open or laparoscopically. In nephron-sparing procedures, the tumor is removed along with a small margin of normal parenchyma.¹⁵

In summary, the goal of surgical intervention is curative intent with removal of the tumor while maintaining as much residual renal function as possible to limit long-term morbidity of chronic kidney disease and associated cardiovascular events.¹⁸ Oncologic outcomes for radical nephrectomy and partial nephrectomy are similar. In one study, overall survival was slightly lower in the partial nephrectomy cohort, but only a small number of the deaths were due to RCC.¹⁹

ADJUVANT THERAPY

Adjuvant systemic therapy currently has no role following nephrectomy for RCC because no systemic therapy has been able to reduce the likelihood of relapse. Randomized trials of cytokine therapy (eg, interferon, interleukin 2) or tyrosine kinase inhibitors (TKIs; eg, sorafenib, sunitinib) with observation alone in patients with locally advanced completely resected RCC have shown no delay in time to relapse or improvement of survival with adjuvant therapy.²⁰ Similarly, adjuvant radiation therapy has not shown benefit even in patients with nodal involvement or incomplete resection.²¹ Therefore, observation remains the standard of care after nephrectomy.

RENAL TUMOR ABLATION

For patients who are deemed not to be surgical candidates due to age, comorbidities, or patient preference and who have tumors less than 4 cm in size (stage I tumors), ablative techniques may be considered. The 2 most well-studied and effective techniques at present are cryoablation and radiofrequency ablation (RFA). Microwave ablation may be an option in some facilities, but the data in RCC are limited. An emerging ablative technique under investigation is irreversible electroporation. At present, the long-term efficacy of all ablative techniques is unknown.

Patient selection is undertaken by urologists and interventional radiologists who evaluate the patient with ultrasound, CT, and/or MRI to determine the location and size of the tumor and the presence or absence of metastatic disease. A pretreatment biopsy is recommended to document the histology of the lesion to confirm a malignancy and to guide future treatment for recurrent or metastatic disease. Contraindications to the procedure include the presence of metastatic disease, a life expectancy of less than 1 year, general medical instability, or uncorrectable coagulopathy due to increased risk of bleeding complications. Tumors in close proximity to the renal hilum or collecting system are a contraindication to the procedure because of the risk for hemorrhage or damage to the collecting system. The location of the tumor in relation to the vasculature is also important to maximize efficacy because the vasculature acts as a “heat sink,” causing dissipation of the thermal energy. Occasionally, stenting of the proximal ureter due to upper tumor location is necessary to prevent thermal injury that could lead to urine leaks.

Selection of the modality to be used primarily depends on operator comfort, which translates to good patient outcomes, such as better cancer control and fewer complications. Cryoablation and RFA have both demonstrated good clinical efficacy and cancer control of 89% and 90%, respectively, with comparable complication rates.²² There have been no studies performed directly comparing the modalities.

Cryoablation

Cryoablation is performed through the insertion of a probe into the tumor, which may be done through a surgical or percutaneous approach. Once the probe is in place, a high- pressure gas (argon, nitrogen) is passed through the probe and upon entering a low pressure region the gas cools. The gas is able to cool to temperatures as low as –185°C. The tissue is then rewarmed through the use of helium, which conversely warms when entering a low pressure area. The process of freezing followed by rewarming subsequently causes cell death/tissue destruction through direct cell injury from cellular dehydration and vascular injury. Clinically, 2 freeze-thaw cycles are used to treat a tumor.^23,24

Radiofrequency ablation, or RFA, targets tumors via an electrode placed within the mass that produces intense frictional heat from medium-frequency alternating current (approximately 500 kHz) produced by a connected generator that is grounded on the patient. The thermal energy created causes coagulative necrosis. Due to the reliance on heat for tumor destruction, central lesions are less amenable to this approach because of the “heat sink” effect from the hilum.²⁴

Microwave Ablation

Microwave ablation, like RFA, relies on the generation of frictional heat to cause cell death by coagulative necrosis. In this case, the friction is created through the activation of water molecules; because of the different thermal kinetics involved with microwave ablation, the “heat sink” effect is minimized when treatment is employed near large vessels, in comparison to RFA.²⁴ The data on this mechanism of ablation are still maturing, with varied outcomes thus far. One study demonstrated outcomes comparable to RFA and cryoablation, with cancer-specific survival of 97.8% at 3 years.²⁵ However, a study by Castle and colleagues²⁶ demonstrated higher recurrence rates. The overarching impediment to widespread adoption of microwave ablation is inconclusive data gleaned from studies with small numbers of patients with limited follow up. The role of this modality will need to be revisited.

Irreversible Electroporation

Irreversible electroporation (IRE) is under investigation. IRE is a non-thermal ablative technique that employs rapid electrical pulses to create pores in cell membranes, leading to cell death. The postulated benefits of IRE include the lack of an effect from “heat sinks” and less collateral damage to the surrounding tissues, when compared with the thermal modalities. In a human phase 1 study of patients undergoing IRE prior to immediate surgical resection, the procedure appeared feasible and safe.²⁷ Significant concerns for this method of ablation possibly inducing cardiac arrhythmias, and the resultant need for sedation with neuromuscular blockade and associated electrocardiography monitoring, may impede its implementation in nonresearch settings.²⁴

ACTIVE SURVEILLANCE

Due to the more frequent use of imaging for various indications, there has been an increase in the discovery of small renal masses (SRM); 85% of RCC that present in an asymptomatic or incidental manner are tumors under 4 cm in diameter.^28,29 The role of active surveillance is evolving, but is primarily suggested for patients who are not candidates for more aggressive intervention based on comorbidities. A recent prospective, nonrandomized analysis of data from the Delayed Intervention and Surveillance for Small Renal Masses (DISSRM) registry evaluated outcomes for patients with SRM looking at primary intervention compared with active surveillance.³⁰ The primary intervention selected was at the discretion of the provider; treatments included partial nephrectomy, RFA, and cryoablation, and active surveillance patients were followed with imaging every 6 months. Progression of SRM, with recommendation for delayed intervention, was defined as a growth rate of mass greater than 0.5 cm/year, size greater than 4 cm, or hematuria. Thirty-six of 158 patients on active surveillance met criteria for progression; 21 underwent delayed intervention. Of note, even the patients who progressed but did not undergo delayed intervention did not develop metastatic disease during the follow-up interval. With a median follow-up of 2 years, cancer-specific survival was noted to be 99% and 100% at 5 years for primary intervention and active surveillance, respectively. Overall survival at 2 years for primary intervention was 98% and 96% for active surveillance; at 5 years, the survival rates were 92% and 75% (P = 0.06). Of note, 2 patients in the primary intervention arm died of RCC, while none in the active surveillance arm died. As would be expected, active surveillance patients were older, had a worse performance status, and had more comorbidities. Interestingly, 40% of patients enrolled selected active surveillance as their preferred management for SRM. The DISSRM results were consistent with data from the Renal Cell Consortium of Canada and other retrospective reviews.^31–33

• What is the approach to follow-up after treatment of localized RCC?

After a patient undergoes treatment for a localized RCC, the goal is to optimize oncologic outcomes, monitor for treatment sequelae, such as renal failure, and focus on survivorship. At this time, there is no consensus in the literature or across published national and international guidelines with regards to the appropriate schedule for surveillance to achieve these goals. In principle, the greatest risk for recurrence occurs within the first 3 years, so many guidelines focus on this timeframe. Likewise, the route of spread tends to be hematogenous, so patients present with pulmonary, bone, and brain metastases, in addition to local recurrence within the renal bed. Symptomatic recurrences often are seen

METASTATIC DISEASE

CASE CONTINUED

CT scan with contrast of the chest, abdomen, and pelvis as well as bone scan are done. CT of the abdomen and pelvis demonstrates a 7.8-cm left renal mass arising from the lower pole of the left kidney. Paraesophageal lymphadenopathy and mesenteric nodules are also noted. CT of the chest demonstrates bilateral pulmonary emboli. Bone scan is significant for increased activity related to the pathological fracture involving the right humerus. The patient undergoes surgery to stabilize the pathologic fracture of his humerus. He is diagnosed with metastatic RCC (clear cell histology) and undergoes palliative debulking nephrectomy.

• How is prognosis defined for metastatic RCC?

PROGNOSTIC MODELS

Limited-stage RCC that is found early can be cured surgically, with estimated 5-year survival rates for stage T1 and T2 disease approaching 90%; however, long-term survival for metastatic disease is poor, with rates ranging from 0% to 20%.¹³ Approximately 30% of patients have metastatic disease at diagnosis, and about one-third of patients who have undergone treatment for localized disease experience relapse.^36,37 Common sites of metastases include lung, lymph nodes, bone, liver, adrenal gland, and brain.

Prognostic scoring systems have been developed to define risk groups and assist with determining appropriate therapy in the metastatic setting. The most widely used validated prognostic factor model is that from the Memorial Sloan-Kettering Cancer Center (MSKCC), which was developed using a multivariate analysis derived from data of patients enrolled in clinical trials and treated with interferon alfa.³⁸ The factors included in the MSKCC model are Karnofsky performance status less than 80, time from diagnosis to treatment with interferon alfa less than 12 months, hemoglobin level less than lower limit of laboratory’s reference range, LDH level greater than 1.5 times the upper limit of laboratory’s reference range, and corrected serum calcium level greater than 10 mg/dL. Risk groups are categorized as favorable (0 risk factors), intermediate (1 to 2 risk factors), and poor (3 or more risk factors).³⁹ Median survival for favorable-, intermediate-, and poor-risk patients was 20, 10, and 4 months, respectively.⁴⁰

Another prognostic model, the International Metastatic RCC Database Consortium, or Heng, model was developed to evaluate prognosis in patients treated with VEGF-targeted therapy.⁴¹ This model was developed from a retrospective study of patients treated with sunitinib, sorafenib, and bevacizumab plus interferon alfa or prior immunotherapy. Prognostic factors in this model include 4 of the 5 MSKCC risk factors (hemoglobin level, corrected serum calcium level, Karnofsky performance status, and time to initial diagnosis). Additionally, this model includes both absolute neutrophil and platelet counts greater than the upper limit of normal. Risk groups are identified as favorable (0 risk factors), intermediate (1 to 2 risk factors), and poor (3 or more risk factors). Median survival for favorable-, intermediate-, and poor-risk patients was not reached, 27 months, and 8.8 months, respectively. The University of California, Los Angeles scoring algorithm to predict survival after nephrectomy and immunotherapy (SANI) in patients with metastatic RCC is another prognostic model that can be used. This simplified scoring system incorporates lymph node status, constitutional symptoms, metastases location, histology, and thyroid stimulating hormone (TSH) level.⁴²

The role of debulking or cytoreductive nephrectomy in treatment of metastatic RCC is well established. Large randomized studies have demonstrated a statistically significant median survival benefit for patients undergoing nephrectomy plus interferon alfa therapy compared with patients treated with interferon alfa alone (13.6 months versus 7.8 months, respectively).⁴³ The role of cytoreductive nephrectomy in combination with antiangiogenic agents is less clear. While a retrospective study investigating outcomes of patients with metastatic RCC receiving anti-VEGF agents showed a prolonged survival with nephrectomy, results of large randomized trials are not yet available.^44,45 Patients with lung-only metastases, good prognostic features, and a good performance status are historically the most likely to benefit from cytoreductive surgery.

CASE CONTINUED

Based on the MSKCC prognostic factor model, the patient is considered to be in the intermediate-risk group (Karnofsky performance status of 80, calcium 9.5 mg/dL, LDH 204 U/L, hemoglobin 13.6 g/dL). He is started on treatment for his bilateral pulmonary emboli and recovers well from orthopedic surgery as well as palliative debulking nephrectomy.

Several approaches to systemic therapy for advanced RCC have been taken based on the histologic type of the tumor. Clear-cell is by far the predominant histologic type in RCC. Several options are available as first-line treatment for patients with metastatic clear-cell RCC (Table 2).^46–54 These include biologic agents such as high-dose interleukin-2 (IL-2) immune therapy, as well as targeted therapies including TKIs and anti-VEGF antibodies. The mammalian target of rapamycin (mTOR) inhibitor temsirolimus is recommended as first-line therapy in patients with poor prognosis only. Second-line therapies for clear-cell RCC following antiangiogenic therapy include TKIs, mTOR inhibitors, nivolumab (PD-1 inhibitor), and the combination of the TKI lenvatinib and mTOR inhibitor everolimus.⁵⁵ In addition, after initial cytokine therapy, TKIs, temsirolimus, and the anti-VEGF antibody bevacizumab are other treatment options available to patients. Best supportive care should always be provided along with initial and subsequent therapies. Clinical trials are also an appropriate choice as first-line or subsequent therapies. All of these therapies require periodic monitoring to prevent and quickly treat adverse effects. Table 3 lists recommended monitoring parameters for each of these agents.⁵⁶

Based on several studies, TKIs seem to be less effective in patients with non–clear-cell type histology.^57,58 In these patients, risk factors can guide therapy. In the ASPEN trial, where 108 patients were randomly assigned to everolimus or sunitinib, patients in the good- and intermediate-risk groups had longer overall and median progression-free survival (PFS) on sunitinib (8.3 months versus 5.3 months, respectively). However, those in the poor-risk group had a longer median overall survival with everolimus.⁵⁹ Given that the role of targeted therapies in non–clear-cell RCCs is less well established, enrollment in clinical trials should be considered as a first-line treatment option.²¹

Sarcomatoid features can be observed in any of the histologic types of RCC, and RCC with these features has an aggressive course and a poor prognosis. Currently, there is no standard therapy for treatment of patients with metastatic or unresectable RCC with sarcomatoid features.⁶⁰ Chemotherapeutic regimens used for soft tissue sarcomas, including a trial of ifosfamide and doxorubicin, did not show any objective response.⁶¹ A small trial of 10 patients treated with doxorubicin and gemcitabine resulted in complete response in 2 patients and partial response in 1 patient.⁶²

Enrollment in a clinical trial remains a first-line treatment option for these patients. More recently, a phase 2 trial of sunitinib and gemcitabine in patients with sarcomatoid (39 patients) and/or poor-risk (33 patients) metastatic RCC showed overall response rates (ORR) of 26% and 24%, respectively. A higher clinical benefit rate (defined as ORR plus stable disease) was seen in patients with tumors containing more than 10% sarcomatoid histology, as compared with patients whose tumors contained less than 10% sarcomatoid histology. Neutropenia (n = 20), anemia (n = 10), and fatigue (n = 7) were the most common grade 3 toxicities seen in all the patients. Although this was a small study, the results showed a trend towards better efficacy of the combination therapy as compared with the single-agent regimen. Currently, another study is underway to further investigate this in a larger group of patients.⁶³

BIOLOGICS

Cytokine therapy, including high-dose IL-2 and interferon alfa, had long been the only first-line treatment option for patients with metastatic or unresectable RCC. Studies of high-dose IL-2 have shown an ORR of 25% and durable response in up to 11% of patients with clear-cell histology.⁶⁴ Toxicities were similar to those previously observed with high-dose IL-2 treatment; the most commonly observed grade 3 toxicities were hypotension and capillary leak syndrome. IL-2 requires strict monitoring (Table 3). It is important to note that retrospective studies evaluating the safety and efficacy of using IL-2 as second-line treatment in patients previously treated with TKIs demonstrated significant toxicity without achieving partial or complete response in any of the patients.⁶⁵

Prior to the advent of TKIs in the treatment of RCC, interferon alfa was a first-line treatment option for those who could not receive high-dose IL-2. It has been shown to produce response rates of approximately 20%, with maximum response seen with a higher dose range of 5 to 20 million units daily in 1 study.^66,67 However, with the introduction of TKIs, which produce a higher and more durable response, interferon alfa alone is no longer recommended as a treatment option.

Bevacizumab is a recombinant humanized monoclonal antibody that binds and neutralizes VEGF-A. Given overexpression of VEGF in RCC, the role of bevacizumab both as a single agent and in combination with interferon alfa has been investigated. In a randomized phase 2 study involving patients with cytokine-refractory disease, bevacizumab produced a 10% response rate and PFS of 4.8 months compared to patients treated with placebo.⁶⁸ In the AVOREN trial, the addition of bevacizumab (10 mg/kg intravenously [IV] every 2 weeks) to interferon alfa (9 million units subcutaneously [SC] 3 times weekly) was shown to significantly increase PFS compared with interferon alfa alone (10.2 months versus 5.4 months; P = 0.0001).^47,48 Adverse effects of this combination therapy include fatigue and asthenia. Additionally, hypertension, proteinuria, and bleeding occurred.

TYROSINE KINASE INHIBITORS

TKIs have largely replaced IL-2 as first-line therapy for metastatic RCC. Axitinib, pazopanib, sorafenib, and sunitinib and can be used as first-line therapy. All of the TKIs can be used as subsequent therapy.

Sunitinib

Sunitinib is an orally administered TKI that inhibits VEGF receptor (VEGFR) types 1 and 2, PDGF receptors (PDGFR) α and β, stem cell factor receptor (c-Kit), and FLT-3 and RET kinases. Motzer and colleagues^52,53 compared sunitinib 50 mg daily orally for 4 weeks with 2 weeks off to the then standard of care, interferon alfa 9 million units SC 3 times weekly. Sunitinib significantly increased the overall objective response rate (47% versus 12%; P < 0.001), PFS (11 versus 5 months; P < 0.001), and overall survival (26.4 versus 21.8 months; hazard ratio [HR], 0.821). The most common side effects are diarrhea, fatigue, nausea/vomiting, anorexia, hypertension, stomatitis, and hand-foot syndrome, occurring in more than 30% of patients. Often patients will require dose reductions or temporary discontinuations to tolerate therapy. Alternative dosing strategies (eg, 50 mg dose orally daily for 2 weeks alternating with 1-week free interval) have been attempted but not prospectively evaluated for efficacy.^69–71

Pazopanib

Pazopanib is an oral multi-kinase inhibitor of VEGFR types 1 and 2, PDGFR, and c-KIT. Results of a phase 3 trial comparing pazopanib (800 mg orally daily) to placebo favored the TKI, with a PFS of 9.2 months versus 4.2 months. A subset of treatment-naïve patients had a longer PFS of 11.1 versus 2.8 months and a response rate of 32% versus 4%.⁷² This led to a noninferiority phase 3 trial comparing pazopanib with sunitinib as first-line therapy.⁵⁰ In this study, PFS was similar (8.4 versus 9.5 months; HR 1.05), and overall safety and quality-of-life endpoints favored pazopanib. Much less fatigue, stomatitis, hand-foot syndrome, and thrombocytopenia occurred with pazopanib, whereas hair color changes, weight loss, alopecia, and elevations of LFT enzymes occurred more frequently with pazopanib. Hypertension is common with the administration of pazopanib as well.

Sorafenib

Sorafenib is an orally administered inhibitor of Raf, serine/threonine kinase, VEGFR, PDGFR, FLT-3, c-Kit, and RET. The pivotal phase 3 Treatment Approaches in Renal Cancer Global Evaluation Trial (TARGET) compared sorafenib (400 mg orally twice daily) with placebo in patients who had progressed on prior cytokine-based therapy.⁷³ A final analysis, which excluded patients who were allowed to cross over therapies, found improved overall survival times (14.3 versus 1.8 months, P = 0.029).⁵¹ Sorafenib is associated with lower rates of diarrhea, rash, fatigue, hand-foot syndrome, alopecia, hypertension, and nausea than sunitinib, although these agents have not been compared to one another.

Axitinib

Axitinib is an oral inhibitor of VEGFRs 1, 2, and 3. Results of the phase 3 AXIS trial comparing axitinib (5 mg orally twice daily) with sorafenib (400 mg orally twice daily) in patients receiving 1 prior systemic therapy showed axitinib was more active than sorafenib in improving ORR (19% versus 9%; P = 0.001) and PFS (6.7 versus 4.7 months; P < 0.001), although no difference in overall survival times was noted.⁷⁴ In a subsequent phase 3 trial comparing these drugs in the first-line setting, axitinib showed a nonsignificantly higher response rate and PFS. Despite this, the National Comprehensive Cancer Network guidelines consider axitinib an acceptable first-line therapy because activity with acceptable toxicity was demonstrated (Table 2).⁴⁶ The most common adverse effects of axitinib are diarrhea, hypertension, fatigue, decreased appetite, dysphonia, hypothyroidism, and upper abdominal pain.

CABOZANTINIB

Given that resistance eventually develops in most patients treated with standard treatments, including bevacizumab and TKIs, the need to evaluate the safety and efficacy of novel agents targeting VEGFR and overcoming this resistance is of vital importance. Cabozantinib is an oral small-molecule inhibitor of VEGFR, Met, and Axl, all tyrosine kinases implicated in metastatic RCC. Overexpression of Met and Axl, which occurs as a result of inactivation of the VHL gene, is associated with a poor prognosis in patients with RCC. In a

MTOR INHIBITORS

The mTOR inhibitors, temsirolimus and everolimus, are also approved for the treatment of metastatic or advanced RCC. These drugs block mTOR’s phosphorylation and subsequent translation of mRNA to inhibit cell proliferation, cell growth, and angiogenesis.⁷⁷ Temsirolimus can be used as first-line therapy for patients with a poor prognosis, and everolimus is appropriate as a subsequent therapy.

Temsirolimus is an intravenous prodrug of rapamycin. It was the first of the class to be approved for metastatic RCC for treatment-naïve patients with a poor prognosis (ie, at least 3 of 6 predictors of poor survival based on MSKCC model).⁵⁴ The pivotal ARCC trial compared temsirolimus (25 mg IV weekly) alone, interferon alfa (3 million units SC 3 times weekly) alone, or the combination (temsirolimus 15 mg IV weekly plus interferon alfa 6 million units SC 3 times weekly). In this trial, temsirolimus monotherapy produced a significantly longer overall survival time than interferon alfa alone (10.9 versus 7.3 months; P = 0.008) and improved PFS time when administered alone or in combination with interferon alfa (3.8 and 3.7 months, respectively, versus 1.9 months). Because no real efficacy advantage of the combination was demonstrated, temsirolimus is administered alone. The most common adverse effects of temsirolimus are asthenia, rash, anemia, nausea, anorexia, pain, and dyspnea. Additionally, hyperglycemia, hyper-cholesterolemia, and hyperlipidemia occur with these agents. Noninfectious pneumonitis is a rare but often fatal complication.

Everolimus is also an orally administered derivative of rapamycin that is approved for use after failure of VEGF-targeted therapies. The results of the landmark trial RECORD-1 demonstrated that everolimus (10 mg orally daily) is effective at prolonging PFS (4 versus 1.9 months; P < 0.001) when compared with best supportive care, a viable treatment option at the time of approval.⁷⁸ The most common adverse effects of everolimus are stomatitis, rash, fatigue, asthenia, and diarrhea. As with temsirolimus, elevations in glucose, lipids, and triglycerides and noninfectious pneumonitis can occur.

TKI + MTOR INHIBITOR

Lenvatinib is also a small molecule targeting multiple tyrosine kinases, primarily VEGF2. Combined with the mTOR inhibitor everolimus, it has been shown to be an effective regimen in patients with metastatic RCC who have failed other therapies. In a randomized phase 2 study involving patients with advanced or metastatic clear-cell RCC, patients were randomly assigned to receive either lenvatinib (24 mg/day), everolimus (10 mg/day), or lenvatinib plus everolimus (18 mg/day and 5 mg/day, respectively). Patients received the treatment continuously on a 28-day cycle until progression or inability to tolerate toxicity. Patients in the lenvatinib plus everolimus arm had median PFS of 14.6 months (95% CI 5.9 to 20.1) versus 5.5 months (95% CI 3.5 to 7.1) with everlolimus alone (HR 0.40 [95% CI 0.24 to 0.68]; P = 0.0005). PFS with levantinib alone was 7.4 months (95% CI 5.6 to 10.20; HR 0.66 [95% CI 0.30 to 1.10]; P = 0.12). In addition, PFS with levantinib alone was significantly prolonged in comparison with everolimus alone (HR 0.61 [95% CI 0.38 to 0.98]; P = 0.048). Grade 3 or 4 toxicity were less frequent in the everolimus only arm and the most common grade 3 or 4 toxicity in the lenvatinib plus everolimus arm was diarrhea. The results of this study show that the combination of lenvatinib plus everolimus is an acceptable second-line option for treatment of patients with advanced or metastatic RCC.⁵⁵

The patient is initially started on pazopanib and tolerates the medication well, with partial response to the treatment. However, on restaging scans he is noted to have small bowel perforation. Pazopanib is discontinued until the patient has a full recovery. He is then started on everolimus. Restaging scans done 3 months after starting everolimus demonstrate disease progression.

• What is the appropriate next step in treatment?

PD1 BLOCKADE

Programmed death 1 (PD-1) protein is a T-cell inhibitory receptor with 2 ligands, PD-L1 and PD-L2. PD-L1 is expressed on many tumors. Blocking the interaction between PD-1 and PD-L1 by anti-PD-1 humanized antibodies potentiates a robust immune response and has been a breakthrough in the field of cancer immunotherapy.⁷⁹ Previous studies have demonstrated that overexpression of PD-L1 leads to worse outcomes and poor prognosis in patients with RCC.⁸⁰ Nivolumab, a fully human IgG4 PD-1 immune checkpoint inhibitor, blocks the interaction between PD-1 and its ligands, PD-L1 and PD-L2. In a randomized, open-label, phase 3 study comparing nivolumab with everolimus in patients with RCC who had previously undergone treatment with other standard therapies, Motzer and colleagues⁸¹ demonstrated a longer overall survival time and fewer adverse effects with nivolumab. In this study, 821 patients with clear-cell RCC were randomly assigned to receive nivolumab (3 mg/kg of body weight IV every 2 weeks) or everolimus (10 mg orally once daily). The median overall survival time with nivolumab was 25 months versus 19.6 months with everolimus (P < 0.0148). Nineteen percent of patients receiving nivolumab experienced grade 3 or 4 toxicities, with fatigue being the most common adverse effect. Grade 3 or 4 toxicities were observed in 37% of patients treated with everolimus, with anemia being the most common. Based on the results of this trial, on November 23, 2015, the U.S. Food and Drug Administration approved nivolumab to treat patients with metastatic RCC who have received a prior antiangiogenic therapy.

CASE CONCLUSION

Both TKI and mTOR inhibitor therapy fail, and the patient is eligible for third-line therapy. Because of his previous GI perforation, other TKIs are not an option. The patient opts for enrollment in hospice due to declining performance status. For other patients in this situation with a good performance status, nivolumab would be a reasonable option.

FUTURE DIRECTIONS

With the approval of nivolumab, multiple treatment options are now available for patients with metastatic or unresectable RCC. Development of other PD-1 inhibitors and immunotherapies as well as multi-targeted TKIs will only serve to expand treatment options for these patients. Given the aggressive course and poor prognosis of non-clear cell renal cell tumors and those with sarcomatoid features, evaluation of systemic and targeted therapies for these subtypes should remain active areas of research and investigation.

INTRODUCTION

Renal cell carcinoma (RCC) is the most common malignancy arising in the kidney, comprising 90% of all renal tumors.¹ Approximately 55,000 new RCC cases are diagnosed each year.² Patients with RCC are often asymptomatic, and most cases are discovered as incidental findings on abdominal imaging performed during evaluation of nonrenal complaints. Limited-stage RCC that is found early can be cured surgically, with estimated 5-year survival rates approaching 90%; however, long-term survival for metastatic disease is poor, with rates ranging from 0% to 20%.² Advanced RCC is resistant to conventional chemotherapy and radiotherapy, and outcomes for patients with metastatic or unresectable RCC remain poor. However, the recent development of new therapeutic modalities that target tumor molecular pathways has expanded the treatment options for these patients and changed the management of RCC.

EPIDEMIOLOGY AND CLASSIFICATION

Median age at diagnosis in the United States is 64 years. Men have a higher incidence of RCC than women, with the highest incidence seen in American Indian and Alaska Native men (30.1 per 100,000 population). Genetic syndromes account for 2% to 4% of all RCCs.² Risk factors for RCC include smoking, hypertension, obesity, and acquired cystic kidney disease that is associated with end-stage renal failure.³ Longer duration of tobacco use is associated with a more aggressive course.

The 2004 World Health Organization (WHO) classification of renal tumors summarizes the previous classification systems (including the Heidelberg and Mainz classification systems) to describe different categories of RCC based on histologic and molecular genetics characteristics.² Using the WHO classification criteria, RCC comprises 90% of all renal tumors, with clear cell being the most common type (80%).² Other types of renal tumors include papillary, chromophobe, oncocytoma, and collecting-duct or Bellini duct tumors. Approximately 3% to 5% of tumors are unclassified. Oncocytomas are generally considered benign, and chromophobe tumors typically have an indolent course and rarely metastasize. Sarcomatoid differentiation can be seen in any histologic type and is associated with a worse prognosis. While different types of tumors may be seen in the kidney (such as transitional cell or lymphomas), the focus of this review is the primary malignancies of the renal parenchyma.

FAMILIAL SYNDROMES

Several genetic syndromes have been identified by studying families with inherited RCC. Among these, von Hippel-Lindau (VHL) gene mutation is the most commonly found inherited genetic defect. Table 1 summarizes the incidence of gene mutations and the corresponding histologic appearance of the most common sporadic and hereditary RCCs.⁴

VHL disease is an autosomal dominant familial syndrome. Patients with this mutation are at higher risk for developing RCC (clear cell histology), retinal angiomas, pheochromocytomas, as well as hemangioblastomas of the central nervous system (CNS).⁴ Of all the genetic mutations seen in RCC, the somatic mutation in the VHL tumor-suppressor gene is by far the most common.⁵ VHL targets hypoxia–inducible factor-1 alpha (HIF-α) for ubiquitination and subsequent degradation, which has been shown to suppress the growth of clear-cell RCC in mouse models.^6–8 HIF expression under hypoxic conditions leads to activation of a number of genes important in blood vessel development, cell proliferation, and glucose metabolism, including vascular endothelial growth factor (VEGF), erythropoietin, platelet-derived growth factor beta (PDGF-β), transforming growth factor alpha (TGF-α), and glucose transporter-1 (GLUT-1). Mutation in the VHL gene prevents degradation of the HIF-α protein, thereby leading to increased expression of these downstream proteins, including MET and Axl. The upregulation of these angiogenic factors is thought to be the underlying process for increased vascularity of CNS hemangioblastomas and clear-cell renal tumors in VHL disease.^4–8

Other less common genetic syndromes seen in hereditary RCC include hereditary papillary RCC, hereditary leiomyomatosis, and Birt-Hogg-Dubé (BHD) syndrome.⁹ In hereditary papillary RCC, the MET gene is mutated. BHD syndrome is a rare, autosomal dominant syndrome characterized by hair follicle hamartomas of the face and neck. About 15% of patients have multiple renal tumors, the majority of which are of the chromophobe or mixed chromophobe-oncocytoma histology. The BHD gene encodes the protein folliculin, which is thought to be a tumor-suppressor gene.

DIAGNOSIS AND STAGING

CASE PRESENTATION

A 74-year-old man who works as an airplane mechanic repairman presents to the emergency department with sudden worsening of chronic right upper arm and shoulder pain after lifting a jug of orange juice. He does not have a significant past medical history and initially thought that his pain was due to a work-related injury. Upon initial evaluation in the emergency department he is found to have a fracture of his right humerus. Given that the fracture appears to be pathologic, further work-up is recommended.

Most patients are asymptomatic until the disease becomes advanced. The classic triad of flank pain, hematuria, and palpable abdominal mass is seen in approximately 10% of patients with RCC, partly because of earlier detection of renal masses by imaging performed for other purposes.¹⁰ Less frequently, patients present with signs or symptoms of metastatic disease such as bone pain or fracture (as seen in the case patient), painful adenopathy, and pulmonary symptoms related to mediastinal masses. Fever, weight loss, anemia, and/or varicocele often occur in young patients (≤ 46 years) and may indicate the presence of a hereditary form of the disease. Patients may present with paraneoplastic syndromes seen as abnormalities on routine blood work. These can include polycythemia or elevated liver function tests (LFTs) without the presence of liver metastases (known as Stauffer syndrome), which can be seen in localized renal tumors. Nearly half (45%) of patients present with localized disease, 25% present with locally advanced disease, and 30% present with metastatic disease.¹¹ Bone is the second most common site of distant metastatic spread (following lung) in patients with advanced RCC.

• What is the approach to initial evaluation for a patient with suspected RCC?

Initial evaluation consists of a physical exam, laboratory tests including complete blood count (CBC) and comprehensive metabolic panel (calcium, serum creatinine, LFTs, lactate dehydrogenase [LDH], and urinalysis), and imaging. Imaging studies include computed tomography (CT) scan with contrast of the abdomen and pelvis or magnetic resonance imaging (MRI) of the abdomen and chest imaging. A chest radiograph may be obtained, although a chest CT is more sensitive for the presence of pulmonary metastases. MRI can be used in patients with renal dysfunction to evaluate the renal vein and inferior vena cava (IVC) for thrombus or to determine the presence of local invasion.¹² Although bone and brain are common sites for metastases, routine imaging is not indicated unless the patient is symptomatic. The value of positron emission tomography in RCC remains undetermined at this time.

Staging is done according to the American Joint Committee on Cancer (AJCC) staging classification for RCC; the Figure summarizes the staging and 5-year survival data based on this classification scheme.^4,13

LIMITED-STAGE DISEASE

• What are the therapeutic options for limited-stage disease?

For patients with nondistant metastases, or limited-stage disease, surgical intervention with curative intent is considered. Convention suggests considering definitive surgery for patients with stage I and II disease, select patients with stage III disease with pathologically enlarged retroperitoneal lymph nodes, patients with IVC and/or cardiac atrium involvement of tumor thrombus, and patients with direct extension of the renal tumor into the ipsilateral adrenal gland if there is no evidence of distant disease. While there may be a role for aggressive surgical intervention in patients with distant metastatic disease, this topic will not be covered in this review.

SURGICAL INTERVENTION

Once patients are determined to be appropriate candidates for surgical removal of a renal tumor, the urologist will perform either a radical nephrectomy or a nephron-sparing nephrectomy, also called a partial nephrectomy. The urologist will evaluate the patient based on his or her body habitus, the location of the tumor, whether multiple tumors in one kidney or bilateral tumors are present, whether the patient has a solitary kidney or otherwise impaired kidney function, and whether the patient has a history of a hereditary syndrome involving kidney cancer as this affects the risk of future kidney tumors.

A radical nephrectomy is surgically preferred in the presence of the following factors: tumor larger than 7 cm in diameter, a more centrally located tumor, suspicion of lymph node involvement, tumor involvement with renal vein or IVC, and/or direct extension of the tumor into the ipsilateral adrenal gland. Nephrectomy involves ligation of the vascular supply (renal artery and vein) followed by removal of the kidney and surrounding Gerota’s fascia. The ipsilateral adrenal gland is removed if there is a high-risk for or presence of invasion of the adrenal gland. Removal of the adrenal gland is not standard since the literature demonstrates there is less than a 10% chance of solitary, ipsilateral adrenal gland involvement of tumor at the time of nephrectomy in the absence of high-risk features, and a recent systematic review suggests that the chance may be as low as 1.8%.¹⁴ Preoperative factors that correlated with adrenal involvement included upper pole kidney location, renal vein thrombosis, higher T stage (T3a and greater), multifocal tumors, and evidence for distant metastases or lymph node involvement. Lymphadenectomy previously had been included in radical nephrectomy but now is performed selectively. Radical nephrectomy may be performed as

Conversely, a nephron-sparing approach is preferred for tumors less than 7 cm in diameter, for patients with a solitary kidney or impaired renal function, for patients with multiple small ipsilateral tumors or with bilateral tumors, or for radical nephrectomy candidates with comorbidities for whom a limited intervention is deemed to be a lower-risk procedure. A nephron-sparing procedure may also be performed open or laparoscopically. In nephron-sparing procedures, the tumor is removed along with a small margin of normal parenchyma.¹⁵

In summary, the goal of surgical intervention is curative intent with removal of the tumor while maintaining as much residual renal function as possible to limit long-term morbidity of chronic kidney disease and associated cardiovascular events.¹⁸ Oncologic outcomes for radical nephrectomy and partial nephrectomy are similar. In one study, overall survival was slightly lower in the partial nephrectomy cohort, but only a small number of the deaths were due to RCC.¹⁹

ADJUVANT THERAPY

Adjuvant systemic therapy currently has no role following nephrectomy for RCC because no systemic therapy has been able to reduce the likelihood of relapse. Randomized trials of cytokine therapy (eg, interferon, interleukin 2) or tyrosine kinase inhibitors (TKIs; eg, sorafenib, sunitinib) with observation alone in patients with locally advanced completely resected RCC have shown no delay in time to relapse or improvement of survival with adjuvant therapy.²⁰ Similarly, adjuvant radiation therapy has not shown benefit even in patients with nodal involvement or incomplete resection.²¹ Therefore, observation remains the standard of care after nephrectomy.

RENAL TUMOR ABLATION

For patients who are deemed not to be surgical candidates due to age, comorbidities, or patient preference and who have tumors less than 4 cm in size (stage I tumors), ablative techniques may be considered. The 2 most well-studied and effective techniques at present are cryoablation and radiofrequency ablation (RFA). Microwave ablation may be an option in some facilities, but the data in RCC are limited. An emerging ablative technique under investigation is irreversible electroporation. At present, the long-term efficacy of all ablative techniques is unknown.

Patient selection is undertaken by urologists and interventional radiologists who evaluate the patient with ultrasound, CT, and/or MRI to determine the location and size of the tumor and the presence or absence of metastatic disease. A pretreatment biopsy is recommended to document the histology of the lesion to confirm a malignancy and to guide future treatment for recurrent or metastatic disease. Contraindications to the procedure include the presence of metastatic disease, a life expectancy of less than 1 year, general medical instability, or uncorrectable coagulopathy due to increased risk of bleeding complications. Tumors in close proximity to the renal hilum or collecting system are a contraindication to the procedure because of the risk for hemorrhage or damage to the collecting system. The location of the tumor in relation to the vasculature is also important to maximize efficacy because the vasculature acts as a “heat sink,” causing dissipation of the thermal energy. Occasionally, stenting of the proximal ureter due to upper tumor location is necessary to prevent thermal injury that could lead to urine leaks.

Selection of the modality to be used primarily depends on operator comfort, which translates to good patient outcomes, such as better cancer control and fewer complications. Cryoablation and RFA have both demonstrated good clinical efficacy and cancer control of 89% and 90%, respectively, with comparable complication rates.²² There have been no studies performed directly comparing the modalities.

Cryoablation

Cryoablation is performed through the insertion of a probe into the tumor, which may be done through a surgical or percutaneous approach. Once the probe is in place, a high- pressure gas (argon, nitrogen) is passed through the probe and upon entering a low pressure region the gas cools. The gas is able to cool to temperatures as low as –185°C. The tissue is then rewarmed through the use of helium, which conversely warms when entering a low pressure area. The process of freezing followed by rewarming subsequently causes cell death/tissue destruction through direct cell injury from cellular dehydration and vascular injury. Clinically, 2 freeze-thaw cycles are used to treat a tumor.^23,24

Radiofrequency ablation, or RFA, targets tumors via an electrode placed within the mass that produces intense frictional heat from medium-frequency alternating current (approximately 500 kHz) produced by a connected generator that is grounded on the patient. The thermal energy created causes coagulative necrosis. Due to the reliance on heat for tumor destruction, central lesions are less amenable to this approach because of the “heat sink” effect from the hilum.²⁴

Microwave Ablation

Microwave ablation, like RFA, relies on the generation of frictional heat to cause cell death by coagulative necrosis. In this case, the friction is created through the activation of water molecules; because of the different thermal kinetics involved with microwave ablation, the “heat sink” effect is minimized when treatment is employed near large vessels, in comparison to RFA.²⁴ The data on this mechanism of ablation are still maturing, with varied outcomes thus far. One study demonstrated outcomes comparable to RFA and cryoablation, with cancer-specific survival of 97.8% at 3 years.²⁵ However, a study by Castle and colleagues²⁶ demonstrated higher recurrence rates. The overarching impediment to widespread adoption of microwave ablation is inconclusive data gleaned from studies with small numbers of patients with limited follow up. The role of this modality will need to be revisited.

Irreversible Electroporation

Irreversible electroporation (IRE) is under investigation. IRE is a non-thermal ablative technique that employs rapid electrical pulses to create pores in cell membranes, leading to cell death. The postulated benefits of IRE include the lack of an effect from “heat sinks” and less collateral damage to the surrounding tissues, when compared with the thermal modalities. In a human phase 1 study of patients undergoing IRE prior to immediate surgical resection, the procedure appeared feasible and safe.²⁷ Significant concerns for this method of ablation possibly inducing cardiac arrhythmias, and the resultant need for sedation with neuromuscular blockade and associated electrocardiography monitoring, may impede its implementation in nonresearch settings.²⁴

ACTIVE SURVEILLANCE

Due to the more frequent use of imaging for various indications, there has been an increase in the discovery of small renal masses (SRM); 85% of RCC that present in an asymptomatic or incidental manner are tumors under 4 cm in diameter.^28,29 The role of active surveillance is evolving, but is primarily suggested for patients who are not candidates for more aggressive intervention based on comorbidities. A recent prospective, nonrandomized analysis of data from the Delayed Intervention and Surveillance for Small Renal Masses (DISSRM) registry evaluated outcomes for patients with SRM looking at primary intervention compared with active surveillance.³⁰ The primary intervention selected was at the discretion of the provider; treatments included partial nephrectomy, RFA, and cryoablation, and active surveillance patients were followed with imaging every 6 months. Progression of SRM, with recommendation for delayed intervention, was defined as a growth rate of mass greater than 0.5 cm/year, size greater than 4 cm, or hematuria. Thirty-six of 158 patients on active surveillance met criteria for progression; 21 underwent delayed intervention. Of note, even the patients who progressed but did not undergo delayed intervention did not develop metastatic disease during the follow-up interval. With a median follow-up of 2 years, cancer-specific survival was noted to be 99% and 100% at 5 years for primary intervention and active surveillance, respectively. Overall survival at 2 years for primary intervention was 98% and 96% for active surveillance; at 5 years, the survival rates were 92% and 75% (P = 0.06). Of note, 2 patients in the primary intervention arm died of RCC, while none in the active surveillance arm died. As would be expected, active surveillance patients were older, had a worse performance status, and had more comorbidities. Interestingly, 40% of patients enrolled selected active surveillance as their preferred management for SRM. The DISSRM results were consistent with data from the Renal Cell Consortium of Canada and other retrospective reviews.^31–33

• What is the approach to follow-up after treatment of localized RCC?

After a patient undergoes treatment for a localized RCC, the goal is to optimize oncologic outcomes, monitor for treatment sequelae, such as renal failure, and focus on survivorship. At this time, there is no consensus in the literature or across published national and international guidelines with regards to the appropriate schedule for surveillance to achieve these goals. In principle, the greatest risk for recurrence occurs within the first 3 years, so many guidelines focus on this timeframe. Likewise, the route of spread tends to be hematogenous, so patients present with pulmonary, bone, and brain metastases, in addition to local recurrence within the renal bed. Symptomatic recurrences often are seen

METASTATIC DISEASE

CASE CONTINUED

CT scan with contrast of the chest, abdomen, and pelvis as well as bone scan are done. CT of the abdomen and pelvis demonstrates a 7.8-cm left renal mass arising from the lower pole of the left kidney. Paraesophageal lymphadenopathy and mesenteric nodules are also noted. CT of the chest demonstrates bilateral pulmonary emboli. Bone scan is significant for increased activity related to the pathological fracture involving the right humerus. The patient undergoes surgery to stabilize the pathologic fracture of his humerus. He is diagnosed with metastatic RCC (clear cell histology) and undergoes palliative debulking nephrectomy.

• How is prognosis defined for metastatic RCC?

PROGNOSTIC MODELS

Limited-stage RCC that is found early can be cured surgically, with estimated 5-year survival rates for stage T1 and T2 disease approaching 90%; however, long-term survival for metastatic disease is poor, with rates ranging from 0% to 20%.¹³ Approximately 30% of patients have metastatic disease at diagnosis, and about one-third of patients who have undergone treatment for localized disease experience relapse.^36,37 Common sites of metastases include lung, lymph nodes, bone, liver, adrenal gland, and brain.

Prognostic scoring systems have been developed to define risk groups and assist with determining appropriate therapy in the metastatic setting. The most widely used validated prognostic factor model is that from the Memorial Sloan-Kettering Cancer Center (MSKCC), which was developed using a multivariate analysis derived from data of patients enrolled in clinical trials and treated with interferon alfa.³⁸ The factors included in the MSKCC model are Karnofsky performance status less than 80, time from diagnosis to treatment with interferon alfa less than 12 months, hemoglobin level less than lower limit of laboratory’s reference range, LDH level greater than 1.5 times the upper limit of laboratory’s reference range, and corrected serum calcium level greater than 10 mg/dL. Risk groups are categorized as favorable (0 risk factors), intermediate (1 to 2 risk factors), and poor (3 or more risk factors).³⁹ Median survival for favorable-, intermediate-, and poor-risk patients was 20, 10, and 4 months, respectively.⁴⁰

Another prognostic model, the International Metastatic RCC Database Consortium, or Heng, model was developed to evaluate prognosis in patients treated with VEGF-targeted therapy.⁴¹ This model was developed from a retrospective study of patients treated with sunitinib, sorafenib, and bevacizumab plus interferon alfa or prior immunotherapy. Prognostic factors in this model include 4 of the 5 MSKCC risk factors (hemoglobin level, corrected serum calcium level, Karnofsky performance status, and time to initial diagnosis). Additionally, this model includes both absolute neutrophil and platelet counts greater than the upper limit of normal. Risk groups are identified as favorable (0 risk factors), intermediate (1 to 2 risk factors), and poor (3 or more risk factors). Median survival for favorable-, intermediate-, and poor-risk patients was not reached, 27 months, and 8.8 months, respectively. The University of California, Los Angeles scoring algorithm to predict survival after nephrectomy and immunotherapy (SANI) in patients with metastatic RCC is another prognostic model that can be used. This simplified scoring system incorporates lymph node status, constitutional symptoms, metastases location, histology, and thyroid stimulating hormone (TSH) level.⁴²

The role of debulking or cytoreductive nephrectomy in treatment of metastatic RCC is well established. Large randomized studies have demonstrated a statistically significant median survival benefit for patients undergoing nephrectomy plus interferon alfa therapy compared with patients treated with interferon alfa alone (13.6 months versus 7.8 months, respectively).⁴³ The role of cytoreductive nephrectomy in combination with antiangiogenic agents is less clear. While a retrospective study investigating outcomes of patients with metastatic RCC receiving anti-VEGF agents showed a prolonged survival with nephrectomy, results of large randomized trials are not yet available.^44,45 Patients with lung-only metastases, good prognostic features, and a good performance status are historically the most likely to benefit from cytoreductive surgery.

CASE CONTINUED

Based on the MSKCC prognostic factor model, the patient is considered to be in the intermediate-risk group (Karnofsky performance status of 80, calcium 9.5 mg/dL, LDH 204 U/L, hemoglobin 13.6 g/dL). He is started on treatment for his bilateral pulmonary emboli and recovers well from orthopedic surgery as well as palliative debulking nephrectomy.

Several approaches to systemic therapy for advanced RCC have been taken based on the histologic type of the tumor. Clear-cell is by far the predominant histologic type in RCC. Several options are available as first-line treatment for patients with metastatic clear-cell RCC (Table 2).^46–54 These include biologic agents such as high-dose interleukin-2 (IL-2) immune therapy, as well as targeted therapies including TKIs and anti-VEGF antibodies. The mammalian target of rapamycin (mTOR) inhibitor temsirolimus is recommended as first-line therapy in patients with poor prognosis only. Second-line therapies for clear-cell RCC following antiangiogenic therapy include TKIs, mTOR inhibitors, nivolumab (PD-1 inhibitor), and the combination of the TKI lenvatinib and mTOR inhibitor everolimus.⁵⁵ In addition, after initial cytokine therapy, TKIs, temsirolimus, and the anti-VEGF antibody bevacizumab are other treatment options available to patients. Best supportive care should always be provided along with initial and subsequent therapies. Clinical trials are also an appropriate choice as first-line or subsequent therapies. All of these therapies require periodic monitoring to prevent and quickly treat adverse effects. Table 3 lists recommended monitoring parameters for each of these agents.⁵⁶

Based on several studies, TKIs seem to be less effective in patients with non–clear-cell type histology.^57,58 In these patients, risk factors can guide therapy. In the ASPEN trial, where 108 patients were randomly assigned to everolimus or sunitinib, patients in the good- and intermediate-risk groups had longer overall and median progression-free survival (PFS) on sunitinib (8.3 months versus 5.3 months, respectively). However, those in the poor-risk group had a longer median overall survival with everolimus.⁵⁹ Given that the role of targeted therapies in non–clear-cell RCCs is less well established, enrollment in clinical trials should be considered as a first-line treatment option.²¹

Sarcomatoid features can be observed in any of the histologic types of RCC, and RCC with these features has an aggressive course and a poor prognosis. Currently, there is no standard therapy for treatment of patients with metastatic or unresectable RCC with sarcomatoid features.⁶⁰ Chemotherapeutic regimens used for soft tissue sarcomas, including a trial of ifosfamide and doxorubicin, did not show any objective response.⁶¹ A small trial of 10 patients treated with doxorubicin and gemcitabine resulted in complete response in 2 patients and partial response in 1 patient.⁶²

Enrollment in a clinical trial remains a first-line treatment option for these patients. More recently, a phase 2 trial of sunitinib and gemcitabine in patients with sarcomatoid (39 patients) and/or poor-risk (33 patients) metastatic RCC showed overall response rates (ORR) of 26% and 24%, respectively. A higher clinical benefit rate (defined as ORR plus stable disease) was seen in patients with tumors containing more than 10% sarcomatoid histology, as compared with patients whose tumors contained less than 10% sarcomatoid histology. Neutropenia (n = 20), anemia (n = 10), and fatigue (n = 7) were the most common grade 3 toxicities seen in all the patients. Although this was a small study, the results showed a trend towards better efficacy of the combination therapy as compared with the single-agent regimen. Currently, another study is underway to further investigate this in a larger group of patients.⁶³

BIOLOGICS

Cytokine therapy, including high-dose IL-2 and interferon alfa, had long been the only first-line treatment option for patients with metastatic or unresectable RCC. Studies of high-dose IL-2 have shown an ORR of 25% and durable response in up to 11% of patients with clear-cell histology.⁶⁴ Toxicities were similar to those previously observed with high-dose IL-2 treatment; the most commonly observed grade 3 toxicities were hypotension and capillary leak syndrome. IL-2 requires strict monitoring (Table 3). It is important to note that retrospective studies evaluating the safety and efficacy of using IL-2 as second-line treatment in patients previously treated with TKIs demonstrated significant toxicity without achieving partial or complete response in any of the patients.⁶⁵

Prior to the advent of TKIs in the treatment of RCC, interferon alfa was a first-line treatment option for those who could not receive high-dose IL-2. It has been shown to produce response rates of approximately 20%, with maximum response seen with a higher dose range of 5 to 20 million units daily in 1 study.^66,67 However, with the introduction of TKIs, which produce a higher and more durable response, interferon alfa alone is no longer recommended as a treatment option.

Bevacizumab is a recombinant humanized monoclonal antibody that binds and neutralizes VEGF-A. Given overexpression of VEGF in RCC, the role of bevacizumab both as a single agent and in combination with interferon alfa has been investigated. In a randomized phase 2 study involving patients with cytokine-refractory disease, bevacizumab produced a 10% response rate and PFS of 4.8 months compared to patients treated with placebo.⁶⁸ In the AVOREN trial, the addition of bevacizumab (10 mg/kg intravenously [IV] every 2 weeks) to interferon alfa (9 million units subcutaneously [SC] 3 times weekly) was shown to significantly increase PFS compared with interferon alfa alone (10.2 months versus 5.4 months; P = 0.0001).^47,48 Adverse effects of this combination therapy include fatigue and asthenia. Additionally, hypertension, proteinuria, and bleeding occurred.

TYROSINE KINASE INHIBITORS

TKIs have largely replaced IL-2 as first-line therapy for metastatic RCC. Axitinib, pazopanib, sorafenib, and sunitinib and can be used as first-line therapy. All of the TKIs can be used as subsequent therapy.

Sunitinib

Sunitinib is an orally administered TKI that inhibits VEGF receptor (VEGFR) types 1 and 2, PDGF receptors (PDGFR) α and β, stem cell factor receptor (c-Kit), and FLT-3 and RET kinases. Motzer and colleagues^52,53 compared sunitinib 50 mg daily orally for 4 weeks with 2 weeks off to the then standard of care, interferon alfa 9 million units SC 3 times weekly. Sunitinib significantly increased the overall objective response rate (47% versus 12%; P < 0.001), PFS (11 versus 5 months; P < 0.001), and overall survival (26.4 versus 21.8 months; hazard ratio [HR], 0.821). The most common side effects are diarrhea, fatigue, nausea/vomiting, anorexia, hypertension, stomatitis, and hand-foot syndrome, occurring in more than 30% of patients. Often patients will require dose reductions or temporary discontinuations to tolerate therapy. Alternative dosing strategies (eg, 50 mg dose orally daily for 2 weeks alternating with 1-week free interval) have been attempted but not prospectively evaluated for efficacy.^69–71

Pazopanib

Pazopanib is an oral multi-kinase inhibitor of VEGFR types 1 and 2, PDGFR, and c-KIT. Results of a phase 3 trial comparing pazopanib (800 mg orally daily) to placebo favored the TKI, with a PFS of 9.2 months versus 4.2 months. A subset of treatment-naïve patients had a longer PFS of 11.1 versus 2.8 months and a response rate of 32% versus 4%.⁷² This led to a noninferiority phase 3 trial comparing pazopanib with sunitinib as first-line therapy.⁵⁰ In this study, PFS was similar (8.4 versus 9.5 months; HR 1.05), and overall safety and quality-of-life endpoints favored pazopanib. Much less fatigue, stomatitis, hand-foot syndrome, and thrombocytopenia occurred with pazopanib, whereas hair color changes, weight loss, alopecia, and elevations of LFT enzymes occurred more frequently with pazopanib. Hypertension is common with the administration of pazopanib as well.

Sorafenib

Sorafenib is an orally administered inhibitor of Raf, serine/threonine kinase, VEGFR, PDGFR, FLT-3, c-Kit, and RET. The pivotal phase 3 Treatment Approaches in Renal Cancer Global Evaluation Trial (TARGET) compared sorafenib (400 mg orally twice daily) with placebo in patients who had progressed on prior cytokine-based therapy.⁷³ A final analysis, which excluded patients who were allowed to cross over therapies, found improved overall survival times (14.3 versus 1.8 months, P = 0.029).⁵¹ Sorafenib is associated with lower rates of diarrhea, rash, fatigue, hand-foot syndrome, alopecia, hypertension, and nausea than sunitinib, although these agents have not been compared to one another.

Axitinib

Axitinib is an oral inhibitor of VEGFRs 1, 2, and 3. Results of the phase 3 AXIS trial comparing axitinib (5 mg orally twice daily) with sorafenib (400 mg orally twice daily) in patients receiving 1 prior systemic therapy showed axitinib was more active than sorafenib in improving ORR (19% versus 9%; P = 0.001) and PFS (6.7 versus 4.7 months; P < 0.001), although no difference in overall survival times was noted.⁷⁴ In a subsequent phase 3 trial comparing these drugs in the first-line setting, axitinib showed a nonsignificantly higher response rate and PFS. Despite this, the National Comprehensive Cancer Network guidelines consider axitinib an acceptable first-line therapy because activity with acceptable toxicity was demonstrated (Table 2).⁴⁶ The most common adverse effects of axitinib are diarrhea, hypertension, fatigue, decreased appetite, dysphonia, hypothyroidism, and upper abdominal pain.

CABOZANTINIB

Given that resistance eventually develops in most patients treated with standard treatments, including bevacizumab and TKIs, the need to evaluate the safety and efficacy of novel agents targeting VEGFR and overcoming this resistance is of vital importance. Cabozantinib is an oral small-molecule inhibitor of VEGFR, Met, and Axl, all tyrosine kinases implicated in metastatic RCC. Overexpression of Met and Axl, which occurs as a result of inactivation of the VHL gene, is associated with a poor prognosis in patients with RCC. In a

MTOR INHIBITORS

The mTOR inhibitors, temsirolimus and everolimus, are also approved for the treatment of metastatic or advanced RCC. These drugs block mTOR’s phosphorylation and subsequent translation of mRNA to inhibit cell proliferation, cell growth, and angiogenesis.⁷⁷ Temsirolimus can be used as first-line therapy for patients with a poor prognosis, and everolimus is appropriate as a subsequent therapy.

Temsirolimus is an intravenous prodrug of rapamycin. It was the first of the class to be approved for metastatic RCC for treatment-naïve patients with a poor prognosis (ie, at least 3 of 6 predictors of poor survival based on MSKCC model).⁵⁴ The pivotal ARCC trial compared temsirolimus (25 mg IV weekly) alone, interferon alfa (3 million units SC 3 times weekly) alone, or the combination (temsirolimus 15 mg IV weekly plus interferon alfa 6 million units SC 3 times weekly). In this trial, temsirolimus monotherapy produced a significantly longer overall survival time than interferon alfa alone (10.9 versus 7.3 months; P = 0.008) and improved PFS time when administered alone or in combination with interferon alfa (3.8 and 3.7 months, respectively, versus 1.9 months). Because no real efficacy advantage of the combination was demonstrated, temsirolimus is administered alone. The most common adverse effects of temsirolimus are asthenia, rash, anemia, nausea, anorexia, pain, and dyspnea. Additionally, hyperglycemia, hyper-cholesterolemia, and hyperlipidemia occur with these agents. Noninfectious pneumonitis is a rare but often fatal complication.

Everolimus is also an orally administered derivative of rapamycin that is approved for use after failure of VEGF-targeted therapies. The results of the landmark trial RECORD-1 demonstrated that everolimus (10 mg orally daily) is effective at prolonging PFS (4 versus 1.9 months; P < 0.001) when compared with best supportive care, a viable treatment option at the time of approval.⁷⁸ The most common adverse effects of everolimus are stomatitis, rash, fatigue, asthenia, and diarrhea. As with temsirolimus, elevations in glucose, lipids, and triglycerides and noninfectious pneumonitis can occur.

TKI + MTOR INHIBITOR

Lenvatinib is also a small molecule targeting multiple tyrosine kinases, primarily VEGF2. Combined with the mTOR inhibitor everolimus, it has been shown to be an effective regimen in patients with metastatic RCC who have failed other therapies. In a randomized phase 2 study involving patients with advanced or metastatic clear-cell RCC, patients were randomly assigned to receive either lenvatinib (24 mg/day), everolimus (10 mg/day), or lenvatinib plus everolimus (18 mg/day and 5 mg/day, respectively). Patients received the treatment continuously on a 28-day cycle until progression or inability to tolerate toxicity. Patients in the lenvatinib plus everolimus arm had median PFS of 14.6 months (95% CI 5.9 to 20.1) versus 5.5 months (95% CI 3.5 to 7.1) with everlolimus alone (HR 0.40 [95% CI 0.24 to 0.68]; P = 0.0005). PFS with levantinib alone was 7.4 months (95% CI 5.6 to 10.20; HR 0.66 [95% CI 0.30 to 1.10]; P = 0.12). In addition, PFS with levantinib alone was significantly prolonged in comparison with everolimus alone (HR 0.61 [95% CI 0.38 to 0.98]; P = 0.048). Grade 3 or 4 toxicity were less frequent in the everolimus only arm and the most common grade 3 or 4 toxicity in the lenvatinib plus everolimus arm was diarrhea. The results of this study show that the combination of lenvatinib plus everolimus is an acceptable second-line option for treatment of patients with advanced or metastatic RCC.⁵⁵

The patient is initially started on pazopanib and tolerates the medication well, with partial response to the treatment. However, on restaging scans he is noted to have small bowel perforation. Pazopanib is discontinued until the patient has a full recovery. He is then started on everolimus. Restaging scans done 3 months after starting everolimus demonstrate disease progression.

• What is the appropriate next step in treatment?

PD1 BLOCKADE

Programmed death 1 (PD-1) protein is a T-cell inhibitory receptor with 2 ligands, PD-L1 and PD-L2. PD-L1 is expressed on many tumors. Blocking the interaction between PD-1 and PD-L1 by anti-PD-1 humanized antibodies potentiates a robust immune response and has been a breakthrough in the field of cancer immunotherapy.⁷⁹ Previous studies have demonstrated that overexpression of PD-L1 leads to worse outcomes and poor prognosis in patients with RCC.⁸⁰ Nivolumab, a fully human IgG4 PD-1 immune checkpoint inhibitor, blocks the interaction between PD-1 and its ligands, PD-L1 and PD-L2. In a randomized, open-label, phase 3 study comparing nivolumab with everolimus in patients with RCC who had previously undergone treatment with other standard therapies, Motzer and colleagues⁸¹ demonstrated a longer overall survival time and fewer adverse effects with nivolumab. In this study, 821 patients with clear-cell RCC were randomly assigned to receive nivolumab (3 mg/kg of body weight IV every 2 weeks) or everolimus (10 mg orally once daily). The median overall survival time with nivolumab was 25 months versus 19.6 months with everolimus (P < 0.0148). Nineteen percent of patients receiving nivolumab experienced grade 3 or 4 toxicities, with fatigue being the most common adverse effect. Grade 3 or 4 toxicities were observed in 37% of patients treated with everolimus, with anemia being the most common. Based on the results of this trial, on November 23, 2015, the U.S. Food and Drug Administration approved nivolumab to treat patients with metastatic RCC who have received a prior antiangiogenic therapy.

CASE CONCLUSION

Both TKI and mTOR inhibitor therapy fail, and the patient is eligible for third-line therapy. Because of his previous GI perforation, other TKIs are not an option. The patient opts for enrollment in hospice due to declining performance status. For other patients in this situation with a good performance status, nivolumab would be a reasonable option.

FUTURE DIRECTIONS

With the approval of nivolumab, multiple treatment options are now available for patients with metastatic or unresectable RCC. Development of other PD-1 inhibitors and immunotherapies as well as multi-targeted TKIs will only serve to expand treatment options for these patients. Given the aggressive course and poor prognosis of non-clear cell renal cell tumors and those with sarcomatoid features, evaluation of systemic and targeted therapies for these subtypes should remain active areas of research and investigation.

INTRODUCTION

Renal cell carcinoma (RCC) is the most common malignancy arising in the kidney, comprising 90% of all renal tumors.¹ Approximately 55,000 new RCC cases are diagnosed each year.² Patients with RCC are often asymptomatic, and most cases are discovered as incidental findings on abdominal imaging performed during evaluation of nonrenal complaints. Limited-stage RCC that is found early can be cured surgically, with estimated 5-year survival rates approaching 90%; however, long-term survival for metastatic disease is poor, with rates ranging from 0% to 20%.² Advanced RCC is resistant to conventional chemotherapy and radiotherapy, and outcomes for patients with metastatic or unresectable RCC remain poor. However, the recent development of new therapeutic modalities that target tumor molecular pathways has expanded the treatment options for these patients and changed the management of RCC.

EPIDEMIOLOGY AND CLASSIFICATION

Median age at diagnosis in the United States is 64 years. Men have a higher incidence of RCC than women, with the highest incidence seen in American Indian and Alaska Native men (30.1 per 100,000 population). Genetic syndromes account for 2% to 4% of all RCCs.² Risk factors for RCC include smoking, hypertension, obesity, and acquired cystic kidney disease that is associated with end-stage renal failure.³ Longer duration of tobacco use is associated with a more aggressive course.

The 2004 World Health Organization (WHO) classification of renal tumors summarizes the previous classification systems (including the Heidelberg and Mainz classification systems) to describe different categories of RCC based on histologic and molecular genetics characteristics.² Using the WHO classification criteria, RCC comprises 90% of all renal tumors, with clear cell being the most common type (80%).² Other types of renal tumors include papillary, chromophobe, oncocytoma, and collecting-duct or Bellini duct tumors. Approximately 3% to 5% of tumors are unclassified. Oncocytomas are generally considered benign, and chromophobe tumors typically have an indolent course and rarely metastasize. Sarcomatoid differentiation can be seen in any histologic type and is associated with a worse prognosis. While different types of tumors may be seen in the kidney (such as transitional cell or lymphomas), the focus of this review is the primary malignancies of the renal parenchyma.

FAMILIAL SYNDROMES

Several genetic syndromes have been identified by studying families with inherited RCC. Among these, von Hippel-Lindau (VHL) gene mutation is the most commonly found inherited genetic defect. Table 1 summarizes the incidence of gene mutations and the corresponding histologic appearance of the most common sporadic and hereditary RCCs.⁴

VHL disease is an autosomal dominant familial syndrome. Patients with this mutation are at higher risk for developing RCC (clear cell histology), retinal angiomas, pheochromocytomas, as well as hemangioblastomas of the central nervous system (CNS).⁴ Of all the genetic mutations seen in RCC, the somatic mutation in the VHL tumor-suppressor gene is by far the most common.⁵ VHL targets hypoxia–inducible factor-1 alpha (HIF-α) for ubiquitination and subsequent degradation, which has been shown to suppress the growth of clear-cell RCC in mouse models.^6–8 HIF expression under hypoxic conditions leads to activation of a number of genes important in blood vessel development, cell proliferation, and glucose metabolism, including vascular endothelial growth factor (VEGF), erythropoietin, platelet-derived growth factor beta (PDGF-β), transforming growth factor alpha (TGF-α), and glucose transporter-1 (GLUT-1). Mutation in the VHL gene prevents degradation of the HIF-α protein, thereby leading to increased expression of these downstream proteins, including MET and Axl. The upregulation of these angiogenic factors is thought to be the underlying process for increased vascularity of CNS hemangioblastomas and clear-cell renal tumors in VHL disease.^4–8

Other less common genetic syndromes seen in hereditary RCC include hereditary papillary RCC, hereditary leiomyomatosis, and Birt-Hogg-Dubé (BHD) syndrome.⁹ In hereditary papillary RCC, the MET gene is mutated. BHD syndrome is a rare, autosomal dominant syndrome characterized by hair follicle hamartomas of the face and neck. About 15% of patients have multiple renal tumors, the majority of which are of the chromophobe or mixed chromophobe-oncocytoma histology. The BHD gene encodes the protein folliculin, which is thought to be a tumor-suppressor gene.

DIAGNOSIS AND STAGING

CASE PRESENTATION

A 74-year-old man who works as an airplane mechanic repairman presents to the emergency department with sudden worsening of chronic right upper arm and shoulder pain after lifting a jug of orange juice. He does not have a significant past medical history and initially thought that his pain was due to a work-related injury. Upon initial evaluation in the emergency department he is found to have a fracture of his right humerus. Given that the fracture appears to be pathologic, further work-up is recommended.

Most patients are asymptomatic until the disease becomes advanced. The classic triad of flank pain, hematuria, and palpable abdominal mass is seen in approximately 10% of patients with RCC, partly because of earlier detection of renal masses by imaging performed for other purposes.¹⁰ Less frequently, patients present with signs or symptoms of metastatic disease such as bone pain or fracture (as seen in the case patient), painful adenopathy, and pulmonary symptoms related to mediastinal masses. Fever, weight loss, anemia, and/or varicocele often occur in young patients (≤ 46 years) and may indicate the presence of a hereditary form of the disease. Patients may present with paraneoplastic syndromes seen as abnormalities on routine blood work. These can include polycythemia or elevated liver function tests (LFTs) without the presence of liver metastases (known as Stauffer syndrome), which can be seen in localized renal tumors. Nearly half (45%) of patients present with localized disease, 25% present with locally advanced disease, and 30% present with metastatic disease.¹¹ Bone is the second most common site of distant metastatic spread (following lung) in patients with advanced RCC.

• What is the approach to initial evaluation for a patient with suspected RCC?

Initial evaluation consists of a physical exam, laboratory tests including complete blood count (CBC) and comprehensive metabolic panel (calcium, serum creatinine, LFTs, lactate dehydrogenase [LDH], and urinalysis), and imaging. Imaging studies include computed tomography (CT) scan with contrast of the abdomen and pelvis or magnetic resonance imaging (MRI) of the abdomen and chest imaging. A chest radiograph may be obtained, although a chest CT is more sensitive for the presence of pulmonary metastases. MRI can be used in patients with renal dysfunction to evaluate the renal vein and inferior vena cava (IVC) for thrombus or to determine the presence of local invasion.¹² Although bone and brain are common sites for metastases, routine imaging is not indicated unless the patient is symptomatic. The value of positron emission tomography in RCC remains undetermined at this time.

Staging is done according to the American Joint Committee on Cancer (AJCC) staging classification for RCC; the Figure summarizes the staging and 5-year survival data based on this classification scheme.^4,13

LIMITED-STAGE DISEASE

• What are the therapeutic options for limited-stage disease?

For patients with nondistant metastases, or limited-stage disease, surgical intervention with curative intent is considered. Convention suggests considering definitive surgery for patients with stage I and II disease, select patients with stage III disease with pathologically enlarged retroperitoneal lymph nodes, patients with IVC and/or cardiac atrium involvement of tumor thrombus, and patients with direct extension of the renal tumor into the ipsilateral adrenal gland if there is no evidence of distant disease. While there may be a role for aggressive surgical intervention in patients with distant metastatic disease, this topic will not be covered in this review.

SURGICAL INTERVENTION

Once patients are determined to be appropriate candidates for surgical removal of a renal tumor, the urologist will perform either a radical nephrectomy or a nephron-sparing nephrectomy, also called a partial nephrectomy. The urologist will evaluate the patient based on his or her body habitus, the location of the tumor, whether multiple tumors in one kidney or bilateral tumors are present, whether the patient has a solitary kidney or otherwise impaired kidney function, and whether the patient has a history of a hereditary syndrome involving kidney cancer as this affects the risk of future kidney tumors.

A radical nephrectomy is surgically preferred in the presence of the following factors: tumor larger than 7 cm in diameter, a more centrally located tumor, suspicion of lymph node involvement, tumor involvement with renal vein or IVC, and/or direct extension of the tumor into the ipsilateral adrenal gland. Nephrectomy involves ligation of the vascular supply (renal artery and vein) followed by removal of the kidney and surrounding Gerota’s fascia. The ipsilateral adrenal gland is removed if there is a high-risk for or presence of invasion of the adrenal gland. Removal of the adrenal gland is not standard since the literature demonstrates there is less than a 10% chance of solitary, ipsilateral adrenal gland involvement of tumor at the time of nephrectomy in the absence of high-risk features, and a recent systematic review suggests that the chance may be as low as 1.8%.¹⁴ Preoperative factors that correlated with adrenal involvement included upper pole kidney location, renal vein thrombosis, higher T stage (T3a and greater), multifocal tumors, and evidence for distant metastases or lymph node involvement. Lymphadenectomy previously had been included in radical nephrectomy but now is performed selectively. Radical nephrectomy may be performed as

Conversely, a nephron-sparing approach is preferred for tumors less than 7 cm in diameter, for patients with a solitary kidney or impaired renal function, for patients with multiple small ipsilateral tumors or with bilateral tumors, or for radical nephrectomy candidates with comorbidities for whom a limited intervention is deemed to be a lower-risk procedure. A nephron-sparing procedure may also be performed open or laparoscopically. In nephron-sparing procedures, the tumor is removed along with a small margin of normal parenchyma.¹⁵

In summary, the goal of surgical intervention is curative intent with removal of the tumor while maintaining as much residual renal function as possible to limit long-term morbidity of chronic kidney disease and associated cardiovascular events.¹⁸ Oncologic outcomes for radical nephrectomy and partial nephrectomy are similar. In one study, overall survival was slightly lower in the partial nephrectomy cohort, but only a small number of the deaths were due to RCC.¹⁹

ADJUVANT THERAPY

Adjuvant systemic therapy currently has no role following nephrectomy for RCC because no systemic therapy has been able to reduce the likelihood of relapse. Randomized trials of cytokine therapy (eg, interferon, interleukin 2) or tyrosine kinase inhibitors (TKIs; eg, sorafenib, sunitinib) with observation alone in patients with locally advanced completely resected RCC have shown no delay in time to relapse or improvement of survival with adjuvant therapy.²⁰ Similarly, adjuvant radiation therapy has not shown benefit even in patients with nodal involvement or incomplete resection.²¹ Therefore, observation remains the standard of care after nephrectomy.

RENAL TUMOR ABLATION

For patients who are deemed not to be surgical candidates due to age, comorbidities, or patient preference and who have tumors less than 4 cm in size (stage I tumors), ablative techniques may be considered. The 2 most well-studied and effective techniques at present are cryoablation and radiofrequency ablation (RFA). Microwave ablation may be an option in some facilities, but the data in RCC are limited. An emerging ablative technique under investigation is irreversible electroporation. At present, the long-term efficacy of all ablative techniques is unknown.

Patient selection is undertaken by urologists and interventional radiologists who evaluate the patient with ultrasound, CT, and/or MRI to determine the location and size of the tumor and the presence or absence of metastatic disease. A pretreatment biopsy is recommended to document the histology of the lesion to confirm a malignancy and to guide future treatment for recurrent or metastatic disease. Contraindications to the procedure include the presence of metastatic disease, a life expectancy of less than 1 year, general medical instability, or uncorrectable coagulopathy due to increased risk of bleeding complications. Tumors in close proximity to the renal hilum or collecting system are a contraindication to the procedure because of the risk for hemorrhage or damage to the collecting system. The location of the tumor in relation to the vasculature is also important to maximize efficacy because the vasculature acts as a “heat sink,” causing dissipation of the thermal energy. Occasionally, stenting of the proximal ureter due to upper tumor location is necessary to prevent thermal injury that could lead to urine leaks.

Selection of the modality to be used primarily depends on operator comfort, which translates to good patient outcomes, such as better cancer control and fewer complications. Cryoablation and RFA have both demonstrated good clinical efficacy and cancer control of 89% and 90%, respectively, with comparable complication rates.²² There have been no studies performed directly comparing the modalities.

Cryoablation

Cryoablation is performed through the insertion of a probe into the tumor, which may be done through a surgical or percutaneous approach. Once the probe is in place, a high- pressure gas (argon, nitrogen) is passed through the probe and upon entering a low pressure region the gas cools. The gas is able to cool to temperatures as low as –185°C. The tissue is then rewarmed through the use of helium, which conversely warms when entering a low pressure area. The process of freezing followed by rewarming subsequently causes cell death/tissue destruction through direct cell injury from cellular dehydration and vascular injury. Clinically, 2 freeze-thaw cycles are used to treat a tumor.^23,24

Radiofrequency ablation, or RFA, targets tumors via an electrode placed within the mass that produces intense frictional heat from medium-frequency alternating current (approximately 500 kHz) produced by a connected generator that is grounded on the patient. The thermal energy created causes coagulative necrosis. Due to the reliance on heat for tumor destruction, central lesions are less amenable to this approach because of the “heat sink” effect from the hilum.²⁴

Microwave Ablation

Microwave ablation, like RFA, relies on the generation of frictional heat to cause cell death by coagulative necrosis. In this case, the friction is created through the activation of water molecules; because of the different thermal kinetics involved with microwave ablation, the “heat sink” effect is minimized when treatment is employed near large vessels, in comparison to RFA.²⁴ The data on this mechanism of ablation are still maturing, with varied outcomes thus far. One study demonstrated outcomes comparable to RFA and cryoablation, with cancer-specific survival of 97.8% at 3 years.²⁵ However, a study by Castle and colleagues²⁶ demonstrated higher recurrence rates. The overarching impediment to widespread adoption of microwave ablation is inconclusive data gleaned from studies with small numbers of patients with limited follow up. The role of this modality will need to be revisited.

Irreversible Electroporation

Irreversible electroporation (IRE) is under investigation. IRE is a non-thermal ablative technique that employs rapid electrical pulses to create pores in cell membranes, leading to cell death. The postulated benefits of IRE include the lack of an effect from “heat sinks” and less collateral damage to the surrounding tissues, when compared with the thermal modalities. In a human phase 1 study of patients undergoing IRE prior to immediate surgical resection, the procedure appeared feasible and safe.²⁷ Significant concerns for this method of ablation possibly inducing cardiac arrhythmias, and the resultant need for sedation with neuromuscular blockade and associated electrocardiography monitoring, may impede its implementation in nonresearch settings.²⁴

ACTIVE SURVEILLANCE

Due to the more frequent use of imaging for various indications, there has been an increase in the discovery of small renal masses (SRM); 85% of RCC that present in an asymptomatic or incidental manner are tumors under 4 cm in diameter.^28,29 The role of active surveillance is evolving, but is primarily suggested for patients who are not candidates for more aggressive intervention based on comorbidities. A recent prospective, nonrandomized analysis of data from the Delayed Intervention and Surveillance for Small Renal Masses (DISSRM) registry evaluated outcomes for patients with SRM looking at primary intervention compared with active surveillance.³⁰ The primary intervention selected was at the discretion of the provider; treatments included partial nephrectomy, RFA, and cryoablation, and active surveillance patients were followed with imaging every 6 months. Progression of SRM, with recommendation for delayed intervention, was defined as a growth rate of mass greater than 0.5 cm/year, size greater than 4 cm, or hematuria. Thirty-six of 158 patients on active surveillance met criteria for progression; 21 underwent delayed intervention. Of note, even the patients who progressed but did not undergo delayed intervention did not develop metastatic disease during the follow-up interval. With a median follow-up of 2 years, cancer-specific survival was noted to be 99% and 100% at 5 years for primary intervention and active surveillance, respectively. Overall survival at 2 years for primary intervention was 98% and 96% for active surveillance; at 5 years, the survival rates were 92% and 75% (P = 0.06). Of note, 2 patients in the primary intervention arm died of RCC, while none in the active surveillance arm died. As would be expected, active surveillance patients were older, had a worse performance status, and had more comorbidities. Interestingly, 40% of patients enrolled selected active surveillance as their preferred management for SRM. The DISSRM results were consistent with data from the Renal Cell Consortium of Canada and other retrospective reviews.^31–33

• What is the approach to follow-up after treatment of localized RCC?

After a patient undergoes treatment for a localized RCC, the goal is to optimize oncologic outcomes, monitor for treatment sequelae, such as renal failure, and focus on survivorship. At this time, there is no consensus in the literature or across published national and international guidelines with regards to the appropriate schedule for surveillance to achieve these goals. In principle, the greatest risk for recurrence occurs within the first 3 years, so many guidelines focus on this timeframe. Likewise, the route of spread tends to be hematogenous, so patients present with pulmonary, bone, and brain metastases, in addition to local recurrence within the renal bed. Symptomatic recurrences often are seen

METASTATIC DISEASE

CASE CONTINUED

CT scan with contrast of the chest, abdomen, and pelvis as well as bone scan are done. CT of the abdomen and pelvis demonstrates a 7.8-cm left renal mass arising from the lower pole of the left kidney. Paraesophageal lymphadenopathy and mesenteric nodules are also noted. CT of the chest demonstrates bilateral pulmonary emboli. Bone scan is significant for increased activity related to the pathological fracture involving the right humerus. The patient undergoes surgery to stabilize the pathologic fracture of his humerus. He is diagnosed with metastatic RCC (clear cell histology) and undergoes palliative debulking nephrectomy.

• How is prognosis defined for metastatic RCC?

PROGNOSTIC MODELS

Limited-stage RCC that is found early can be cured surgically, with estimated 5-year survival rates for stage T1 and T2 disease approaching 90%; however, long-term survival for metastatic disease is poor, with rates ranging from 0% to 20%.¹³ Approximately 30% of patients have metastatic disease at diagnosis, and about one-third of patients who have undergone treatment for localized disease experience relapse.^36,37 Common sites of metastases include lung, lymph nodes, bone, liver, adrenal gland, and brain.

Prognostic scoring systems have been developed to define risk groups and assist with determining appropriate therapy in the metastatic setting. The most widely used validated prognostic factor model is that from the Memorial Sloan-Kettering Cancer Center (MSKCC), which was developed using a multivariate analysis derived from data of patients enrolled in clinical trials and treated with interferon alfa.³⁸ The factors included in the MSKCC model are Karnofsky performance status less than 80, time from diagnosis to treatment with interferon alfa less than 12 months, hemoglobin level less than lower limit of laboratory’s reference range, LDH level greater than 1.5 times the upper limit of laboratory’s reference range, and corrected serum calcium level greater than 10 mg/dL. Risk groups are categorized as favorable (0 risk factors), intermediate (1 to 2 risk factors), and poor (3 or more risk factors).³⁹ Median survival for favorable-, intermediate-, and poor-risk patients was 20, 10, and 4 months, respectively.⁴⁰

Another prognostic model, the International Metastatic RCC Database Consortium, or Heng, model was developed to evaluate prognosis in patients treated with VEGF-targeted therapy.⁴¹ This model was developed from a retrospective study of patients treated with sunitinib, sorafenib, and bevacizumab plus interferon alfa or prior immunotherapy. Prognostic factors in this model include 4 of the 5 MSKCC risk factors (hemoglobin level, corrected serum calcium level, Karnofsky performance status, and time to initial diagnosis). Additionally, this model includes both absolute neutrophil and platelet counts greater than the upper limit of normal. Risk groups are identified as favorable (0 risk factors), intermediate (1 to 2 risk factors), and poor (3 or more risk factors). Median survival for favorable-, intermediate-, and poor-risk patients was not reached, 27 months, and 8.8 months, respectively. The University of California, Los Angeles scoring algorithm to predict survival after nephrectomy and immunotherapy (SANI) in patients with metastatic RCC is another prognostic model that can be used. This simplified scoring system incorporates lymph node status, constitutional symptoms, metastases location, histology, and thyroid stimulating hormone (TSH) level.⁴²

The role of debulking or cytoreductive nephrectomy in treatment of metastatic RCC is well established. Large randomized studies have demonstrated a statistically significant median survival benefit for patients undergoing nephrectomy plus interferon alfa therapy compared with patients treated with interferon alfa alone (13.6 months versus 7.8 months, respectively).⁴³ The role of cytoreductive nephrectomy in combination with antiangiogenic agents is less clear. While a retrospective study investigating outcomes of patients with metastatic RCC receiving anti-VEGF agents showed a prolonged survival with nephrectomy, results of large randomized trials are not yet available.^44,45 Patients with lung-only metastases, good prognostic features, and a good performance status are historically the most likely to benefit from cytoreductive surgery.

CASE CONTINUED

Based on the MSKCC prognostic factor model, the patient is considered to be in the intermediate-risk group (Karnofsky performance status of 80, calcium 9.5 mg/dL, LDH 204 U/L, hemoglobin 13.6 g/dL). He is started on treatment for his bilateral pulmonary emboli and recovers well from orthopedic surgery as well as palliative debulking nephrectomy.

Several approaches to systemic therapy for advanced RCC have been taken based on the histologic type of the tumor. Clear-cell is by far the predominant histologic type in RCC. Several options are available as first-line treatment for patients with metastatic clear-cell RCC (Table 2).^46–54 These include biologic agents such as high-dose interleukin-2 (IL-2) immune therapy, as well as targeted therapies including TKIs and anti-VEGF antibodies. The mammalian target of rapamycin (mTOR) inhibitor temsirolimus is recommended as first-line therapy in patients with poor prognosis only. Second-line therapies for clear-cell RCC following antiangiogenic therapy include TKIs, mTOR inhibitors, nivolumab (PD-1 inhibitor), and the combination of the TKI lenvatinib and mTOR inhibitor everolimus.⁵⁵ In addition, after initial cytokine therapy, TKIs, temsirolimus, and the anti-VEGF antibody bevacizumab are other treatment options available to patients. Best supportive care should always be provided along with initial and subsequent therapies. Clinical trials are also an appropriate choice as first-line or subsequent therapies. All of these therapies require periodic monitoring to prevent and quickly treat adverse effects. Table 3 lists recommended monitoring parameters for each of these agents.⁵⁶

Based on several studies, TKIs seem to be less effective in patients with non–clear-cell type histology.^57,58 In these patients, risk factors can guide therapy. In the ASPEN trial, where 108 patients were randomly assigned to everolimus or sunitinib, patients in the good- and intermediate-risk groups had longer overall and median progression-free survival (PFS) on sunitinib (8.3 months versus 5.3 months, respectively). However, those in the poor-risk group had a longer median overall survival with everolimus.⁵⁹ Given that the role of targeted therapies in non–clear-cell RCCs is less well established, enrollment in clinical trials should be considered as a first-line treatment option.²¹

Sarcomatoid features can be observed in any of the histologic types of RCC, and RCC with these features has an aggressive course and a poor prognosis. Currently, there is no standard therapy for treatment of patients with metastatic or unresectable RCC with sarcomatoid features.⁶⁰ Chemotherapeutic regimens used for soft tissue sarcomas, including a trial of ifosfamide and doxorubicin, did not show any objective response.⁶¹ A small trial of 10 patients treated with doxorubicin and gemcitabine resulted in complete response in 2 patients and partial response in 1 patient.⁶²

Enrollment in a clinical trial remains a first-line treatment option for these patients. More recently, a phase 2 trial of sunitinib and gemcitabine in patients with sarcomatoid (39 patients) and/or poor-risk (33 patients) metastatic RCC showed overall response rates (ORR) of 26% and 24%, respectively. A higher clinical benefit rate (defined as ORR plus stable disease) was seen in patients with tumors containing more than 10% sarcomatoid histology, as compared with patients whose tumors contained less than 10% sarcomatoid histology. Neutropenia (n = 20), anemia (n = 10), and fatigue (n = 7) were the most common grade 3 toxicities seen in all the patients. Although this was a small study, the results showed a trend towards better efficacy of the combination therapy as compared with the single-agent regimen. Currently, another study is underway to further investigate this in a larger group of patients.⁶³

BIOLOGICS

Cytokine therapy, including high-dose IL-2 and interferon alfa, had long been the only first-line treatment option for patients with metastatic or unresectable RCC. Studies of high-dose IL-2 have shown an ORR of 25% and durable response in up to 11% of patients with clear-cell histology.⁶⁴ Toxicities were similar to those previously observed with high-dose IL-2 treatment; the most commonly observed grade 3 toxicities were hypotension and capillary leak syndrome. IL-2 requires strict monitoring (Table 3). It is important to note that retrospective studies evaluating the safety and efficacy of using IL-2 as second-line treatment in patients previously treated with TKIs demonstrated significant toxicity without achieving partial or complete response in any of the patients.⁶⁵

Prior to the advent of TKIs in the treatment of RCC, interferon alfa was a first-line treatment option for those who could not receive high-dose IL-2. It has been shown to produce response rates of approximately 20%, with maximum response seen with a higher dose range of 5 to 20 million units daily in 1 study.^66,67 However, with the introduction of TKIs, which produce a higher and more durable response, interferon alfa alone is no longer recommended as a treatment option.

Bevacizumab is a recombinant humanized monoclonal antibody that binds and neutralizes VEGF-A. Given overexpression of VEGF in RCC, the role of bevacizumab both as a single agent and in combination with interferon alfa has been investigated. In a randomized phase 2 study involving patients with cytokine-refractory disease, bevacizumab produced a 10% response rate and PFS of 4.8 months compared to patients treated with placebo.⁶⁸ In the AVOREN trial, the addition of bevacizumab (10 mg/kg intravenously [IV] every 2 weeks) to interferon alfa (9 million units subcutaneously [SC] 3 times weekly) was shown to significantly increase PFS compared with interferon alfa alone (10.2 months versus 5.4 months; P = 0.0001).^47,48 Adverse effects of this combination therapy include fatigue and asthenia. Additionally, hypertension, proteinuria, and bleeding occurred.

TYROSINE KINASE INHIBITORS

TKIs have largely replaced IL-2 as first-line therapy for metastatic RCC. Axitinib, pazopanib, sorafenib, and sunitinib and can be used as first-line therapy. All of the TKIs can be used as subsequent therapy.

Sunitinib

Sunitinib is an orally administered TKI that inhibits VEGF receptor (VEGFR) types 1 and 2, PDGF receptors (PDGFR) α and β, stem cell factor receptor (c-Kit), and FLT-3 and RET kinases. Motzer and colleagues^52,53 compared sunitinib 50 mg daily orally for 4 weeks with 2 weeks off to the then standard of care, interferon alfa 9 million units SC 3 times weekly. Sunitinib significantly increased the overall objective response rate (47% versus 12%; P < 0.001), PFS (11 versus 5 months; P < 0.001), and overall survival (26.4 versus 21.8 months; hazard ratio [HR], 0.821). The most common side effects are diarrhea, fatigue, nausea/vomiting, anorexia, hypertension, stomatitis, and hand-foot syndrome, occurring in more than 30% of patients. Often patients will require dose reductions or temporary discontinuations to tolerate therapy. Alternative dosing strategies (eg, 50 mg dose orally daily for 2 weeks alternating with 1-week free interval) have been attempted but not prospectively evaluated for efficacy.^69–71

Pazopanib

Pazopanib is an oral multi-kinase inhibitor of VEGFR types 1 and 2, PDGFR, and c-KIT. Results of a phase 3 trial comparing pazopanib (800 mg orally daily) to placebo favored the TKI, with a PFS of 9.2 months versus 4.2 months. A subset of treatment-naïve patients had a longer PFS of 11.1 versus 2.8 months and a response rate of 32% versus 4%.⁷² This led to a noninferiority phase 3 trial comparing pazopanib with sunitinib as first-line therapy.⁵⁰ In this study, PFS was similar (8.4 versus 9.5 months; HR 1.05), and overall safety and quality-of-life endpoints favored pazopanib. Much less fatigue, stomatitis, hand-foot syndrome, and thrombocytopenia occurred with pazopanib, whereas hair color changes, weight loss, alopecia, and elevations of LFT enzymes occurred more frequently with pazopanib. Hypertension is common with the administration of pazopanib as well.

Sorafenib

Sorafenib is an orally administered inhibitor of Raf, serine/threonine kinase, VEGFR, PDGFR, FLT-3, c-Kit, and RET. The pivotal phase 3 Treatment Approaches in Renal Cancer Global Evaluation Trial (TARGET) compared sorafenib (400 mg orally twice daily) with placebo in patients who had progressed on prior cytokine-based therapy.⁷³ A final analysis, which excluded patients who were allowed to cross over therapies, found improved overall survival times (14.3 versus 1.8 months, P = 0.029).⁵¹ Sorafenib is associated with lower rates of diarrhea, rash, fatigue, hand-foot syndrome, alopecia, hypertension, and nausea than sunitinib, although these agents have not been compared to one another.

Axitinib

Axitinib is an oral inhibitor of VEGFRs 1, 2, and 3. Results of the phase 3 AXIS trial comparing axitinib (5 mg orally twice daily) with sorafenib (400 mg orally twice daily) in patients receiving 1 prior systemic therapy showed axitinib was more active than sorafenib in improving ORR (19% versus 9%; P = 0.001) and PFS (6.7 versus 4.7 months; P < 0.001), although no difference in overall survival times was noted.⁷⁴ In a subsequent phase 3 trial comparing these drugs in the first-line setting, axitinib showed a nonsignificantly higher response rate and PFS. Despite this, the National Comprehensive Cancer Network guidelines consider axitinib an acceptable first-line therapy because activity with acceptable toxicity was demonstrated (Table 2).⁴⁶ The most common adverse effects of axitinib are diarrhea, hypertension, fatigue, decreased appetite, dysphonia, hypothyroidism, and upper abdominal pain.

CABOZANTINIB

Given that resistance eventually develops in most patients treated with standard treatments, including bevacizumab and TKIs, the need to evaluate the safety and efficacy of novel agents targeting VEGFR and overcoming this resistance is of vital importance. Cabozantinib is an oral small-molecule inhibitor of VEGFR, Met, and Axl, all tyrosine kinases implicated in metastatic RCC. Overexpression of Met and Axl, which occurs as a result of inactivation of the VHL gene, is associated with a poor prognosis in patients with RCC. In a

MTOR INHIBITORS

The mTOR inhibitors, temsirolimus and everolimus, are also approved for the treatment of metastatic or advanced RCC. These drugs block mTOR’s phosphorylation and subsequent translation of mRNA to inhibit cell proliferation, cell growth, and angiogenesis.⁷⁷ Temsirolimus can be used as first-line therapy for patients with a poor prognosis, and everolimus is appropriate as a subsequent therapy.

Temsirolimus is an intravenous prodrug of rapamycin. It was the first of the class to be approved for metastatic RCC for treatment-naïve patients with a poor prognosis (ie, at least 3 of 6 predictors of poor survival based on MSKCC model).⁵⁴ The pivotal ARCC trial compared temsirolimus (25 mg IV weekly) alone, interferon alfa (3 million units SC 3 times weekly) alone, or the combination (temsirolimus 15 mg IV weekly plus interferon alfa 6 million units SC 3 times weekly). In this trial, temsirolimus monotherapy produced a significantly longer overall survival time than interferon alfa alone (10.9 versus 7.3 months; P = 0.008) and improved PFS time when administered alone or in combination with interferon alfa (3.8 and 3.7 months, respectively, versus 1.9 months). Because no real efficacy advantage of the combination was demonstrated, temsirolimus is administered alone. The most common adverse effects of temsirolimus are asthenia, rash, anemia, nausea, anorexia, pain, and dyspnea. Additionally, hyperglycemia, hyper-cholesterolemia, and hyperlipidemia occur with these agents. Noninfectious pneumonitis is a rare but often fatal complication.

Everolimus is also an orally administered derivative of rapamycin that is approved for use after failure of VEGF-targeted therapies. The results of the landmark trial RECORD-1 demonstrated that everolimus (10 mg orally daily) is effective at prolonging PFS (4 versus 1.9 months; P < 0.001) when compared with best supportive care, a viable treatment option at the time of approval.⁷⁸ The most common adverse effects of everolimus are stomatitis, rash, fatigue, asthenia, and diarrhea. As with temsirolimus, elevations in glucose, lipids, and triglycerides and noninfectious pneumonitis can occur.

TKI + MTOR INHIBITOR

Lenvatinib is also a small molecule targeting multiple tyrosine kinases, primarily VEGF2. Combined with the mTOR inhibitor everolimus, it has been shown to be an effective regimen in patients with metastatic RCC who have failed other therapies. In a randomized phase 2 study involving patients with advanced or metastatic clear-cell RCC, patients were randomly assigned to receive either lenvatinib (24 mg/day), everolimus (10 mg/day), or lenvatinib plus everolimus (18 mg/day and 5 mg/day, respectively). Patients received the treatment continuously on a 28-day cycle until progression or inability to tolerate toxicity. Patients in the lenvatinib plus everolimus arm had median PFS of 14.6 months (95% CI 5.9 to 20.1) versus 5.5 months (95% CI 3.5 to 7.1) with everlolimus alone (HR 0.40 [95% CI 0.24 to 0.68]; P = 0.0005). PFS with levantinib alone was 7.4 months (95% CI 5.6 to 10.20; HR 0.66 [95% CI 0.30 to 1.10]; P = 0.12). In addition, PFS with levantinib alone was significantly prolonged in comparison with everolimus alone (HR 0.61 [95% CI 0.38 to 0.98]; P = 0.048). Grade 3 or 4 toxicity were less frequent in the everolimus only arm and the most common grade 3 or 4 toxicity in the lenvatinib plus everolimus arm was diarrhea. The results of this study show that the combination of lenvatinib plus everolimus is an acceptable second-line option for treatment of patients with advanced or metastatic RCC.⁵⁵

The patient is initially started on pazopanib and tolerates the medication well, with partial response to the treatment. However, on restaging scans he is noted to have small bowel perforation. Pazopanib is discontinued until the patient has a full recovery. He is then started on everolimus. Restaging scans done 3 months after starting everolimus demonstrate disease progression.

• What is the appropriate next step in treatment?

PD1 BLOCKADE

Programmed death 1 (PD-1) protein is a T-cell inhibitory receptor with 2 ligands, PD-L1 and PD-L2. PD-L1 is expressed on many tumors. Blocking the interaction between PD-1 and PD-L1 by anti-PD-1 humanized antibodies potentiates a robust immune response and has been a breakthrough in the field of cancer immunotherapy.⁷⁹ Previous studies have demonstrated that overexpression of PD-L1 leads to worse outcomes and poor prognosis in patients with RCC.⁸⁰ Nivolumab, a fully human IgG4 PD-1 immune checkpoint inhibitor, blocks the interaction between PD-1 and its ligands, PD-L1 and PD-L2. In a randomized, open-label, phase 3 study comparing nivolumab with everolimus in patients with RCC who had previously undergone treatment with other standard therapies, Motzer and colleagues⁸¹ demonstrated a longer overall survival time and fewer adverse effects with nivolumab. In this study, 821 patients with clear-cell RCC were randomly assigned to receive nivolumab (3 mg/kg of body weight IV every 2 weeks) or everolimus (10 mg orally once daily). The median overall survival time with nivolumab was 25 months versus 19.6 months with everolimus (P < 0.0148). Nineteen percent of patients receiving nivolumab experienced grade 3 or 4 toxicities, with fatigue being the most common adverse effect. Grade 3 or 4 toxicities were observed in 37% of patients treated with everolimus, with anemia being the most common. Based on the results of this trial, on November 23, 2015, the U.S. Food and Drug Administration approved nivolumab to treat patients with metastatic RCC who have received a prior antiangiogenic therapy.

CASE CONCLUSION

Both TKI and mTOR inhibitor therapy fail, and the patient is eligible for third-line therapy. Because of his previous GI perforation, other TKIs are not an option. The patient opts for enrollment in hospice due to declining performance status. For other patients in this situation with a good performance status, nivolumab would be a reasonable option.

FUTURE DIRECTIONS

With the approval of nivolumab, multiple treatment options are now available for patients with metastatic or unresectable RCC. Development of other PD-1 inhibitors and immunotherapies as well as multi-targeted TKIs will only serve to expand treatment options for these patients. Given the aggressive course and poor prognosis of non-clear cell renal cell tumors and those with sarcomatoid features, evaluation of systemic and targeted therapies for these subtypes should remain active areas of research and investigation.

INTRODUCTION

Soft tissue sarcomas (STSs) are rare adult tumors, with 3.4 new cases per 100,000 persons or 12,310 expected new cases in 2016.¹ Sarcomas are a heterogeneous collection of tumors that affect fat, muscle, nerve, nerve sheath, vascular, and connective tissues. There are more than 50 histological subtypes that comprise this diverse category of tumors. Treatment varies by stage, with limb-sparing surgery representing the mainstay of curative-intent treatment. Radiation and chemotherapy may also be considered depending on the size, grade, and location of the tumor. Survival rates have been stagnant until recently, with a disease-specific survival hovering around 65%.¹ Given the complexity of these cases, all patients ideally should be evaluated and treated by a multidisciplinary team at an institution with extensive experience treating STS.²

EPIDEMIOLOGY AND CLASSIFICATION

The most common STS subtypes are gastrointestinal stromal tumor (GIST), undifferentiate pleomorphic sarcoma (previously referred to as malignant fibrous histiocytoma), liposarcoma, leiomyosarcoma, synovial sarcoma, malignant peripheral nerve sheath tumor, rhabdomyosarcoma, and unclassified sarcoma.³ Liposarcoma is one of the most common subtypes, comprising 20% of all STSs; it is subdivided into well-differentiated/dedifferentiated liposarcomas, myxoid/round cell liposarcomas, and pleomorphic liposarcomas. Well-differentiated liposarcomas tend to occur in the retroperitoneum and limbs, while both myxoid and round cell as well as pleomorphic liposarcomas more commonly originate on the limbs. Histology varies based on subtype and ranges from mature-appearing adipocytes and fibroblasts to undifferentiated cells with minimal lipogenic differentiation.⁴

Leiomyosarcomas are smooth muscle tumors and are usually located in the retroperitoneum, but have also been associated with peripheral soft tissue and vasculature. Typical histology ranges from well-defined areas of spindle-shaped cells to poorly differentiated anaplastic spindle cells.^5,6 Synovial sarcomas are a distinct type of STS that can show epithelial differentiation and account for 5% of adult STSs. The extremities are the most common presenting location (90%).⁷

Rhabdomyosarcomas are skeletal muscle tumors and are further subdivided into embryonal, alveolar, and pleomorphic subtypes. Embryonal histology ranges from primitive mesenchymal-appearing cells to highly differentiated muscle cells. Alveolar rhabdomyosarcoma has the worst prognosis of the subtypes and consists of round cells with high nuclear-to-chromatin ratios that form “glandular-like” or “alveolar” spaces.⁸ Pleomorphic rhabdomyosarcomas are composed of rhabdomyoblasts that can affect many different locations, but most commonly present on the lower extremities.⁹

Malignant peripheral nerve sheath tumor (MPNST) comprises 5% to 10% of all STSs. These tumors are associated with neurofibromatosis type 1 (NF-1), with 25% to 50% of tumors occurring in NF-1 patients. Additionally, most patients have a truncating lesion in the NF1 gene on chromosome 17.¹⁰ Anghileri et al in their single institution analysis of 205 patients with MPNSTs found the 2 most common presenting sites were the trunk and extremities. Histologically, these tumors have dense fascicles of spindle cells.¹⁰

GISTs are the most common STS of the gastrointestinal (GI) tract. Previously, GISTs were classified as smooth muscle tumors and were not accounted for in the literature as a separate entity distinct from leiomyomas, leiomyoblastomas, and leiomyosarcomas.¹¹ GISTs are found throughout the GI tract: the most common sites are the stomach (60%) and small intestine (30%). Less common sites include duodenum (4%–5%), esophagus (1%), rectum (1%–2%), and appendix (< 0.2%).¹² GISTs can be spindle cell, epithelioid, or mesenchymal tumors. Immunohistochemically, GISTs are KIT (CD117) positive. Other cell markers that are also commonly positive include CD34 (60%–70%) and smooth muscle actin (SMA) (25%).¹¹ The majority of GISTs (80%) have an activating c-KIT gene mutation. The most common mutation site is exon 11, with less common c-KIT gene mutations also occurring at exon 9 or 13. Not all GISTs have KIT mutations. The second most common mutation is the PDGFRA mutation (5%–10% of GISTs).² A minority of GISTs are negative for both KIT and PDGFRA mutations. These tumors were previously called wild-type, but as the majority have either a succinate dehydrogenase (SDH) loss of function or loss of SDHB protein expression, they are now referred to as SDH-deficient GISTs.² GISTs vary in aggressiveness from incidental to aggressive. Typically, small intestine and rectal GISTs are more aggressive than gastric GISTs. Both size and mitotic rate help to predict the metastatic potential of the tumor. Tumors less than 2 cm in size and having a mitotic rate of less than 5 per 50 high-power fields (hpf) have the lowest risk of metastases, while tumors greater than 5 cm and with more than 5 mitoses per 50 hpf have the highest rates of metastases.¹²

Undifferentiated sarcomas have no specific features and typically consist of primitive mesenchymal cells.

CLINICAL EVALUATION

CASE PRESENTATION

Initial Presentation and History

A 55-year-old man presents to his primary care physician with a painless mass in his anterior thigh. The mass has been present for the past 3 months and he believes that it is enlarging. The patient has a history of well-controlled hypertension and hyperlipidemia. His medications include atorvastatin and hydrochlorothiazide. He has no known drug allergies. Family history is notable for diabetes and hypertension. He drinks 4 to 5 alcoholic drinks a week and he is a former smoker. He quit smoking in his 30s and only smoked intermittently prior to quitting. He denies any illicit drug use. He works as a high school principal. Currently, he feels well. His review of systems is otherwise noncontributory.

Physical Examination

On physical exam, he is afebrile with a blood pressure of 132/75 mm Hg, respiratory rate of 10 breaths/min, and oxygen saturation of 99% on room air. He is a well appearing, overweight male. His head and neck exam is unremarkable. Lung exam reveals clear breath sounds, and cardiac exam reveals a regular rate and rhythm. His abdomen is obese, soft, and without hepatosplenomegaly. There is a large, fixed mass on the anterior lateral aspect of his right thigh. He has no appreciable lymphadenopathy. His neurological exam is unremarkable.

• What are risk factors for sarcoma?

There are few known risk factors for sarcoma. Established risks factors include prior radiation therapy, chronic lymphedema, viruses, and genetic cancer syndromes including Li-Fraumeni syndrome, hereditary retinoblastoma, and NF-1. Other environmental exposures include phenoxyacetic acids and chlorophenols.¹⁴ The majority of cases are sporadic, with only a minority of patients having one of these known risk factors.¹⁵ Up to one third of sarcomas have a specific translocation and are driven by fusion oncogenes (Table 1).

A painless mass is the most typical presenting symptom. Size at presentation varies based on location, with extremity and head and neck locations typically presenting at smaller sizes than retroperitoneal tumors.¹⁴ Patients may experience pain and numbness as the mass enlarges and impinges on surrounding structures including nerves and vasculature. The vast majority of patients are without systemic symptoms.

• How is sarcoma staged?

The American Joint Committee on Cancer (AJCC) staging system is the most widely used staging system in the United States. The latest AJCC manual was updated in 2010 to include a 3-tiered grading system where the tumor is classified according to tumor size, lymph node involvement, metastases, and grade at time of diagnosis (Table 2 and Table 3). Additionally, tumor depth in relation to deep fascia is also taken into account, with superficial tumors being assigned a designation of “a” and deep tumors a designation of “b.”

Previously, 2 of the most widely used grading systems were the National Cancer Institute (NCI) and French Federation of Cancer Centers Sarcoma Group (FNCLCC) systems, both 3-tier grading systems. The main components that determine the NCI grade are the tumor’s histologic type and location and the amount of tumor necrosis. The FNCLCC system evaluation focuses on tumor differentiation, mitotic rate, and amount of tumor necrosis. A study that compared the NCI and FNCLCC grading systems found that FNCLCC was a better predictor of mortality and distant metastasis.¹⁶ Previously, the AJCC was a 4-tier grading system, but the 2010 version was updated to the 3-tier FNCLCC grading system. Additionally, the AJCC system has reclassified single lymph node disease as stage III as it confers better survival than metastatic disease.¹⁷ It is important that pathology be evaluated by a sarcoma specialist as disagreements with regard to histologic subtype and grade are common.^18,19

• What are the most important prognostic factors?

Prognostic factors include grade, size, and presence of metastases at presentation. Best survival is associated with low-grade, small tumors with no metastases at time of diagnosis.¹⁴

Imaging should be undertaken to help differentiate between benign and malignant lesions. Ideally, it should be undertaken before a biopsy is planned as the imaging can be used to plan biopsy as well as provide invaluable prognostic information. There are several imaging modalities that should be considered during the preliminary work-up and staging of STSs. Conventional imaging includes magnetic resonance imaging (MRI) of the original tumor site; computed tomography (CT) to evaluate for pulmonary metastases and, depending on location, liver metastases; and in the case of small, low-grade tumors, chest radiography. MRI is considered the test of choice for soft tissue masses and can help delineate benign masses such as hematomas, lipomas, and hemangiomas from sarcomas.²⁰ It is difficult to compare the accuracy of positron emission tomography (PET)/CT to CT and MRI because most studies have evaluated PET/CT in parallel with CT and MRI.²¹ Tateishi et al compared the accuracy of conventional imaging, PET/CT, and PET/CT combined with conventional imaging at determining the TNM staging for 117 patients. They found that conventional imaging correctly classified 77% of patients, PET alone correctly classified 70%, PET/CT correctly classified 83%, and PET/CT combined with conventional imaging correctly staged 87%.²²

• Which subtypes are most likely to metastasize?

Although the vast majority of sarcomas spread hematogenously, 3 have a propensity to spread lymphogenously: epithelioid sarcoma, rhabdomyosarcoma, and clear-cell sarcoma. Additionally, certain subtypes are more likely to metastasize: leiomyosarcomas, synovial sarcomas, neurogenic sarcomas, rhabdomyosarcomas, and epithelioid sarcomas.²³ Sarcomas metastasize to the lungs more frequently than to the liver. The metastatic pattern is defined primarily by sarcoma subtype and site of primary tumor. Sarcomas rarely metastasize to the brain (~1%).

MANAGEMENT

CASE CONTINUED

The patient undergoes an ultrasound to better visualize the mass. Given the heterogeneous character of the mass, he is referred for an MRI to evaluate the mass and a CT scan of the chest, abdomen, and pelvis to evaluate for distant metastases. MRI reveals a 5.1 cm × 4.6 cm heterogeneous mass invading the superficial fascia of the rectus femoris muscle. No suspicious lymph nodes or other masses are identified on imaging. The patient next undergoes an image-guided core needle biopsy. Pathology from that procedure is consistent with a stage III, T2bNxMx, grade 3, dedifferentiated liposarcoma.

• What is the best management approach for this patient?

SURGERY

Surgery is the mainstay of treatment for STS. Patients with the best prognosis are those who undergo complete resection with negative surgical margins.^24,25 Goal tumor-free margin is 1 to 3 cm.²⁶ Complete resection confers the best long-term survival. Both local and metastatic recurrence is higher in patients with incomplete resection and positive margins.^24,25 In a study that analyzed 2084 localized primary STSs, patients with negative margins had a local recurrence rate of 15% versus a rate of 28% in patients with positive margins. This translated into higher 5-year local recurrence-free survival for patients with negative surgical margins (82%) compared to patients with positive margins (65%).²⁷ Another study similarly found that patients with negative margins at referral to their institution who underwent postoperative radiation had high local control rates of 93% (95% confidence interval [CI] 87% to 97%) at 5, 10, and 15 years.²⁶ Although radiation improves local control, neither preoperative or postoperative radiation has been shown to improve progression-free or overall survival.²⁸ Other factors that are associated with risk of recurrence are tumor location, history of previous recurrence, age of patient, histopathology, tumor grade, and tumor size. Approximately 40% to 50% of patients with high-grade tumors (defined as size > 5 cm, deep location, and high grade) will develop distant metastases.²⁹

Zagars et al found that positive or uncertain resection margin had a relative risk of local recurrence of 2.0 (95% CI 1.3 to 3.1; P = 0.002), and presentation with locally recurrent disease (vs new tumor) had a relative risk of local recurrence of 2.0 (95% CI 1.2 to 3.4; P = 0.013).²⁶ Patients with STS of head and neck and deep trunk have higher recurrence rates than those with superficial trunk and extremity STS. A single-institution retrospective review demonstrated that patients with completely resectable retroperitoneal sarcomas have longer median survival (103 months) compared to patients with incompletely resected abdominal sarcomas (18 months).²⁵

CASE CONTINUED

The patient is referred for limb-sparing surgery after presentation at a multidisciplinary tumor board. Prior to undergoing resection of the tumor, he is also referred to radiation-oncology to discuss the risks and benefits of combination radiotherapy and surgery as opposed to surgical resection alone.

• What is the evidence for radiation therapy?

RADIATION THERAPY

Radiation therapy is used in the preoperative, intraoperative, and postoperative settings to reduce the risk of local recurrence. There are several options for radiation, including external beam radiation therapy (EBRT), intraoperative radiation, and brachytherapy. A newer strategy, intensity-modulated radiation therapy (IMRT), utilizes 3-dimensional modeling to reduce radiation dosages. Overall there are no differences in overall survival or local recurrence rates between preoperative and postoperative radiation in STS.²⁸

The rationale behind preoperative radiation is that it reduces seeding of tumor cells, especially at the time of surgery.³⁰ Additionally, for EBRT, preoperative radiation has smaller field sizes and lower radiation doses. It can also help to reduce the size of the tumor prior to resection. Intraoperative radiation is often paired with preoperative radiation as a boost dose given only to the area of residual tumor.

Suit et al reviewed patients treated at a single institution with limb-sparing surgery and different radiation strategies. Local control rates between preoperative and postoperative radiation groups were not statistically significant. Local recurrence was linked to grade and size of the tumor in both groups. The authors did note, however, that the preoperative radiation group tended to have larger tumor sizes at baseline compared to the patients who received postoperative radiation.³⁰ A study that compared 190 patients who received preoperative and postoperative EBRT or brachytherapy (primary end point was wound complications, and local control was a secondary end point) showed a trend towards greater local control with preoperative radiation; however, the preoperative radiation group had significantly more wound complications compared to the postoperative radiation group.³¹

Yang et al found that postoperative EBRT decreases rates of local recurrence compared to surgery alone in high-grade extremity sarcomas.³² However, there were no differences in rates of distant metastases and overall survival between the 2 treatment groups. Similarly, in patients with low-grade sarcoma, there were fewer local recurrences in those who received EBRT and surgery as compared to surgery alone.³² Another study that evaluated 164 patients who received either adjuvant brachytherapy or no further therapy after complete resection found that brachytherapy reduced local recurrence in high-grade sarcomas. No difference in local recurrence rates was found in patients with low-grade sarcomas, nor was a significant difference found in the rates of distant metastases and overall survival between the 2 treatment groups.³³ With regards to IMRT, a single institution cohort experience with 41 patients who received IMRT following limb-sparing surgery had similar local control rates when compared to historical controls.³⁴

CASE CONTINUED

After discussion of the risks and benefits of radiation therapy, the patient opts for preoperative radiation prior to resection of his liposarcoma. He receives 50 Gy of EBRT prior to undergoing resection. Resection results in R1 margin consistent with microscopic disease. He receives 16 Gy of EBRT as a boost after recovery from his resection.²

• What is the evidence for neoadjuvant and adjuvant chemotherapy for stage I tumors?

CHEMOTHERAPY

Localized Sarcoma

For localized sarcoma, limb-sparing resection with or without radiation forms the backbone of treatment. Studies have evaluated chemotherapy in both the neoadjuvant and adjuvant settings, with the vast majority of studies evaluating doxorubicin-based chemotherapy regimens in the adjuvant settings. Due to the rare nature of sarcomas, most studies are not sufficiently powered to detect significant benefit from chemotherapy. Several trials evaluating chemotherapy regimens in the neoadjuvant and adjuvant settings needed to be terminated prematurely due to inadequate enrollment into the study. ^35,36

For stage IA (T1a-Tb, N0, M0, low grade) tumors, no additional therapy is recommended after limb-sparing surgery with appropriate surgical margins. For stage IB (T2a-2b, N0, M0, low grade) tumors with insufficient margins, re-resection and radiation therapy should be considered, while for stage IIA (T1a-1b, N0, M0, G2-3) tumors preoperative or postoperative radiation therapy is recommended.² Studies have not found benefit of adjuvant chemotherapy in these low-grade, stage I tumors in terms of progression-free survival and overall survival.³⁷

• At what stage should chemotherapy be considered?

For stage IIb and stage III tumors, surgery and radiation therapy again form the backbone of therapy; however, neoadjuvant and adjuvant chemotherapy are also recommended as considerations. Anthracycline-based chemotherapy with either single-agent doxorubicin or doxorubicin and ifosfamide in combination are considered first-line chemotherapy agents in locally advanced STS.^2,29,37

Evidence regarding the efficacy of both neoadjuvant and adjuvant chemotherapy regimens in the setting of locally advanced high-grade STS has been mixed. The Sarcoma Meta-analysis Collaboration evaluated 14 trials of doxorubicin-based adjuvant chemotherapy and found a trend towards overall survival in the treatment groups that received chemotherapy.³⁷ All trials included in the meta-analysis compared patients with localized resectable soft-tissue sarcomas who were randomized to either adjuvant chemotherapy or no adjuvant chemotherapy after limb-sparing surgery with or without radiation therapy. None of the individual trials showed a significant benefit, and all trials had large confidence intervals; however, the meta-analysis showed significant benefit in the chemotherapy treatment groups with regard to local recurrence, distant recurrence, and progression-free survival. No significant difference in overall survival was found.³⁷ Pervais et al updated the Sarcoma Meta-analysis Collaboration’s 1997 meta-analysis with the inclusion of 4 new trials that evaluated doxorubicin combined with ifosfamide and found that both patients who received doxorubicin-based regimens or doxorubicin with ifosfamide had significant decreases in distant and overall recurrences. Only the trials that utilized doxorubicin and ifosfamide had an improved overall survival that was statistically significant (hazard ratio 0.56 [95% CI 0.36 to 0.85]; P = 0.01).²⁹ Although no significant heterogeneity was found among the trials included in either meta-analysis, a variety of sarcomas were included in each clinical trial evaluated. Given the extremely small number of each sarcoma subtype present in each trial, subgroup analysis is difficult and prone to inaccuracies. As a result, it is not known if certain histological subtypes are more or less responsive to chemotherapy.^37–39

One randomized controlled trial evaluated neoadjuvant chemotherapy in high-risk sarcomas defined as tumors greater than 8 cm or grade II/III tumors. This study evaluated doxorubicin and ifosfamide and found no significant difference in disease-free and overall survival in the neoadjuvant therapy group compared to the control group.³⁵ There remains controversy in the literature with regards to adjuvant chemotherapy. Many oncologists offer adjuvant chemotherapy to patients with certain stage III subtypes. Examples of subtypes that may be offered adjuvant therapy include myxoid liposarcomas, synovial sarcomas, and leiomyosarcomas.² With regards to how many cycles of chemotherapy should be considered, a noninferiority study compared 3 cycles of epirubicin and ifosfamide to 5 cycles of epirubicin and ifosfamide in patients with high-risk locally advanced adult STSs. Three cycles of preoperative epirubicin and ifosfamide was found to be noninferior to 5 cycles with regards to overall survival.³⁸

• What is this patient’s risk for recurrence?

The patient is at intermediate risk for recurrence. Numerous studies have demonstrated that tumor size, grade, and location are the most important factors to determine risk of recurrence, with larger size, higher grades, and deeper locations being associated with higher risk of recurrence. In an analysis of 1041 patients with STS of the extremities, high grade was the most important risk factor for distant metastases.³⁹ The highest risk of recurrence is within the first 2 years. Given that the patient’s initial tumor was located in the extremity, he is more likely to have a distant metastasis as his site of recurrence; individuals with retroperitoneal tumors and visceral tumors are more likely to recur locally.⁴⁰ For STSs of the extremity, distant metastases determine overall survival, whereas patients with retroperitoneal sarcomas can die from complications of local metastases.⁴¹ Once a patient develops distant metastases, the most important prognostic factor is the size of the tumor, with tumors larger than 10 cm having a relative risk of 1.5 (95% CI 1.0 to 2.0).³⁹

• What are the recommendations for surveillance?

Surveillance recommendations are based on the stage of the sarcoma. Stage I tumors are the least likely to recur either locally or distally. As a result, it is recommended that stage I tumors be followed with history and physical exam every 3 to 6 months for the first 2 to 3 years, and then annually after the first 2 to 3 years. Chest x-rays should be considered every 6 to 12 months.² For stage II–IV tumors, history and physical exam is recommended every 3 to 6 months for the first 2 to 3 years. Chest and distant metastases imaging should also be performed every 3 to 6 months during this time frame. For the next 2 years, history and physical exam and imaging are recommended every 6 months. After the first 4 to 5 years, annual follow-up is recommended.²

CASE CONTINUED

The patient does well for 1 year. With physical therapy, he regains most of the strength and coordination of the lower extremity. He is followed every 3 months with chest x-rays and a MRI of the thigh for the first year. On his fourth follow-up clinic visit, he describes increased dyspnea on exertion over the previous few weeks and is found to have multiple lung metastases in both lungs on chest x-ray. He undergoes further evaluation for metastases and is not found to have any other metastatic lesions. Bronchoscopy and biopsy of 1 of the lung nodules confirms recurrent dedifferentiated liposarcoma.

• Should this patient undergo metastectomy?

An analysis of 3149 patients with STS treated at Memorial Sloan-Kettering who developed lung metastases found that patients with pulmonary metastases have survival rates of 25%. The most important prognostic factor for survival was complete resection of all metastases.⁴² For stage IV disease, surgery is used only in certain instances. In instances where tumor is more localized or limited, removal of metastases or metastectomy can play a role in management.²

CASE CONTINUED

Because the patient’s metastases are limited to the lungs, he is referred for metastectomy. He undergoes wedge resection for definitive diagnosis but it is not possible to completely resect all of the metastases. He is thus referred to a medical oncologist to discuss his treatment options.

• What are treatment options for unresectable or metastatic disease?

Metastatic Disease

Unlike local and locally advanced disease, chemotherapy forms the backbone of treatment in stage IV disease. Doxorubicin and olaratumab or doxorubicin and ifosfamide in combination are considered first line in metastatic disease. Response rates for single-agent doxorubicin range from 16% to 27%, while phase 2 and phase 3 studies of doxorubicin and ifosfamide have found response rates ranging from 18% to 36%.⁴³ In addition, the effectiveness of doxorubicin and ifosfamide phase 2 and 3 trials varied. Edmonson et al found a tumor regression rate of 34% for doxorubicin and ifosfamide as compared to 20% for doxorubicin alone.⁴⁴ In comparison, Santoro et al found a response rate of 21.3% for doxorubicin alone and 25.2% for doxorubicin and ifosfamide.⁴⁵ Neither study found increased survival benefit for doxorubicin and ifosfamide when compared to doxorubicin alone. In a Cochrane review evaluating randomized trials that compared doxorubicin and combination chemotherapy regimens, response rates varied from 14% for doxorubicin in combination with streptomycin to 34% for doxorubicin and ifosfamide. Most trials did not show a significant benefit for combination therapies when compared to doxorubicin alone.⁴³ Mean survival with doxorubicin or doxorubicin and ifosfamide is 12 months. High rates of recurrence highlight the need for additional chemotherapy regimens.

The newest approved agent is olaratumab, a monoclonal antibody that binds platelet-derived growth factor receptor alpha and prevents receptor activation. A phase 1-b and phase 2 trial evaluated patients with locally advanced and metastatic STS and randomly assigned them to either olaratumab and doxorubicin or doxorubicin alone.⁴⁶ Progression-free survival for olaratumab/doxorubicin was 6.6 months (95% CI 4.1 to 8.3) compared to 4.1 months (95% CI 2.8 to 5.4) for doxorubicin alone. The objective response rate was 18.2% (95% CI 9.8 to 29.6) for olaratumab/doxorubicin compared to 7.5% (95% CI 2.5 to 6.6) for doxorubicin alone. Furthermore, the median overall survival for olaratumab plus doxorubicin was 26.5 months (95% CI 20.9 to 31.7) compared to 14.7 months for doxorubicin alone (95% CI 5.5 to 26.0). Impressively, this improved response was notable across histological types. Furthermore, patients who had previously been treated with more than 1 regimen and those who were treatment naïve had similar response rates.⁴⁶

• What are second-line treatment options?

Doxorubicin has been used in combination with several other agents including dacarbazine (DTIC) as well as DTIC and ifosfamide (MAID). Borden et al evaluated patients with metastatic STS and randomly assigned the patients to either doxorubicin or doxorubicin and DTIC. Combination therapy demonstrated better tumor response than doxorubicin alone: 30% complete or partial response for combination therapy and 18% for doxorubicin alone.⁴⁷ However, Omura et al

Several additional regimens have undergone evaluation in metastatic and recurrent STSs. Gemcitabine has been used both as a single agent and as part of combination therapy in many studies. Studies with gemcitabine in combination with either docetaxel or DTIC have been the most efficacious. In a phase 2 trial, patients with metastatic STS were randomly assigned to either gemcitabine alone or gemcitabine and docetaxel. Combination therapy had a higher response rate (16% versus 8%) and longer overall survival (17.9 months versus 11.5 months) than gemcitabine alone.⁵⁰ Furthermore, a phase 2 trial of gemcitabine and docetaxel in patients with unresectable leiomyosarcoma showed an overall response rate of 56%, with 3 complete and 15 partial responses among the 34 patients enrolled in the study.⁵¹

A phase 2 trial randomly assigned patients with unresectable or metastatic STS to either DTIC or combination gemcitabine and DTIC.⁵² Gemcitabine-DTIC had a superior progression-free survival at 3 months (56% [95% CI 43% to 69%]) as compared to DTIC alone (37% [95% CI 23.5% to 50%]). Furthermore, mean progression-free survival and overall survival were improved in the gemcitabine-DTIC group (4.2 months and 16.8 months) as compared to the DTIC group (2.0 months and 8.2 months).⁵² DTIC has a single-agent response rate of 16%, but has been shown to be particularly effective in the setting of leiomyosarcomas.⁴⁹

• Does response to treatment regimens differ by histologic subtype?

The majority of STS trials include many different histologic subtypes. Given the rarity of sarcomas as a whole, many trials have had difficulty recruiting adequate numbers of patients to have sufficient power to definitely determine if the treatment under investigation has clinical benefit. Furthermore, the patients recruited have been heterogeneous with regard to subtype. Many older studies hypothesized that the efficacy of chemotherapeutic agents vary based on histologic subtype; however, for most subtypes the number of individuals included in those trials was too low to evaluate efficacy based on subtype.

Some exceptions exist, however. For example, both gemcitabine-DTIC and gemcitabine-docetaxel have been found to be particularly effective in the treatment of leiomyosarcomas.^50,52 Additionally, a retrospective study found a 51% overall response rate for patients with myxoid liposarcomas treated with trabectedin.⁵³ Studies of patients with angiosarcoma treated with paclitaxel have demonstrated response rates of 43% and 53%.^54,55

• What are the newest approved and investigational agents?

A recently approved agent is trabectedin, a tris tetrahydroisoquinoline alkaloid isolated from ascidians that binds to the minor groove of DNA and causes disruptions in the cell cycle. Samuels et al reported data from a single-arm, open-label expanded access trial that evaluated patients with advanced metastatic sarcomas.⁵⁶ In this study, patients with liposarcomas and leiomyosarcomas had an objective response rate of 6.9% (95% CI 4.8 to 9.6) as compared to a rate of 5.9% (95% CI 4.4 to 7.8) for all assessable patients. Median survival was 11.9 months for all patients, with improved median survivals for liposarcoma and leiomyosarcomas of 16.2 months (95% CI 14.1 to 19.5) compared to 8.4 months (95% CI 7.1 to 10.7 months) for other subtypes.⁵⁶

Schöffski et al evaluated eribulin, a chemotherapeutic agent that affects microtubule dynamics, in a phase 2 trial of patients with progressive or high-grade STS with progression on previous chemotherapy. They found a median progression-free survival of 2.6 months (95% CI 1.7 to 6.2) for adipocytic sarcoma, 2.9 months (95% CI 2.4 to 4.6) for leiomyosarcoma, 2.6 months (95% CI 2.3 to 4.3) for synovial sarcoma, and 2.1 months (95% CI 1.4 to 2.9) for other sarcomas.⁵⁷

Van der Graaf and colleagues randomly assigned patients with metastatic nonadipocytic STS to pazopanib or placebo in a phase 3 trial. Pazopanib is a small-molecule endothelial growth factor inhibitor with activity against vascular endothelial growth factors 1, 2, and 3 as well as platelet-derived growth factors. Median progression-free survival was 4.6 months (95% CI 3.7 to 4.8) with pazopanib compared to 1.6 months (95% CI 0.9 to 1.8) with placebo.⁵⁸ Adipocytic sarcomas (liposarcomas) were excluded from the trial because phase 2 trials had found a lower rate of progression-free survival (26%) for them compared to other subtypes.

Toxicities were seen with each of the regimens studied and were common in the randomized trials, with higher rates of toxicities in the combination chemotherapy regimens. The most common toxicities are myelosuppression, nausea, and vomiting. In the doxorubicin trials, the most common toxicities were myelosuppression, nausea, and vomiting.⁴⁴

Ifosfamide both as an individual agent and in combination with doxorubicin has higher rates and higher grades of toxicity than doxorubicin alone. Myelosuppression is the most common toxicity associated with ifosfamide, and the most commonly affected cell line is leukocytes.⁴⁴ Combination doxorubicin and ifosfamide also had high rates of nausea and vomiting (95%) and alopecia (100%).³⁵

Neutropenia is the most common toxicity associated with gemcitabine and dacarbazine, while their most common nonhematologic toxicities are fatigue and nausea.^52,59 Trabectedin’s most common toxicities are nausea (29%), neutropenia (24%), and fatigue (23%). It has also been shown to cause increased alkaline phosphatase (20%) and alanine aminotransferase (19%) levels.⁵⁶ In a phase 2 study of eribulin, 50% of patients had neutropenia, and other toxicities included fatigue, alopecia, nausea, sensory neuropathy, and thrombocytopenia.⁵⁷ Pazopanib is generally well tolerated; the most common toxicities are fatigue (65%), diarrhea (58%), nausea (54%), and hypertension (41%).⁵⁸ Higher rates of neutropenia, mucositis, nausea, vomiting, diarrhea, and transfusion reactions were seen with olaratumab and doxorubicin compared to doxorubicin alone in phase 1b and 2 studies.⁴⁶

CASE CONCLUSION

Given his poor prognosis with unresectable metastatic undifferentiated liposarcoma, the patient considers a clinical trial prior to undergoing combined therapy with doxorubicin and ifosfamide. He tolerates therapy well with stable disease at 6 months.

CONCLUSION

STSs are a heterogeneous collection of rare tumors. Low-grade, localized tumors have the best prognosis, and patients who undergo complete resection have the best long-term survival. Due to the rarity of STSs, trials often have limited enrollment, and little progress has been made with regards to treatment and survival rates for metastatic and unresectable disease. All patients should be evaluated and treated at specialized sarcoma centers. This case highlights the need for continued research and clinical trials to improve overall survival of patients with sarcoma.

INTRODUCTION

Soft tissue sarcomas (STSs) are rare adult tumors, with 3.4 new cases per 100,000 persons or 12,310 expected new cases in 2016.¹ Sarcomas are a heterogeneous collection of tumors that affect fat, muscle, nerve, nerve sheath, vascular, and connective tissues. There are more than 50 histological subtypes that comprise this diverse category of tumors. Treatment varies by stage, with limb-sparing surgery representing the mainstay of curative-intent treatment. Radiation and chemotherapy may also be considered depending on the size, grade, and location of the tumor. Survival rates have been stagnant until recently, with a disease-specific survival hovering around 65%.¹ Given the complexity of these cases, all patients ideally should be evaluated and treated by a multidisciplinary team at an institution with extensive experience treating STS.²

EPIDEMIOLOGY AND CLASSIFICATION

The most common STS subtypes are gastrointestinal stromal tumor (GIST), undifferentiate pleomorphic sarcoma (previously referred to as malignant fibrous histiocytoma), liposarcoma, leiomyosarcoma, synovial sarcoma, malignant peripheral nerve sheath tumor, rhabdomyosarcoma, and unclassified sarcoma.³ Liposarcoma is one of the most common subtypes, comprising 20% of all STSs; it is subdivided into well-differentiated/dedifferentiated liposarcomas, myxoid/round cell liposarcomas, and pleomorphic liposarcomas. Well-differentiated liposarcomas tend to occur in the retroperitoneum and limbs, while both myxoid and round cell as well as pleomorphic liposarcomas more commonly originate on the limbs. Histology varies based on subtype and ranges from mature-appearing adipocytes and fibroblasts to undifferentiated cells with minimal lipogenic differentiation.⁴

Leiomyosarcomas are smooth muscle tumors and are usually located in the retroperitoneum, but have also been associated with peripheral soft tissue and vasculature. Typical histology ranges from well-defined areas of spindle-shaped cells to poorly differentiated anaplastic spindle cells.^5,6 Synovial sarcomas are a distinct type of STS that can show epithelial differentiation and account for 5% of adult STSs. The extremities are the most common presenting location (90%).⁷

Rhabdomyosarcomas are skeletal muscle tumors and are further subdivided into embryonal, alveolar, and pleomorphic subtypes. Embryonal histology ranges from primitive mesenchymal-appearing cells to highly differentiated muscle cells. Alveolar rhabdomyosarcoma has the worst prognosis of the subtypes and consists of round cells with high nuclear-to-chromatin ratios that form “glandular-like” or “alveolar” spaces.⁸ Pleomorphic rhabdomyosarcomas are composed of rhabdomyoblasts that can affect many different locations, but most commonly present on the lower extremities.⁹

Malignant peripheral nerve sheath tumor (MPNST) comprises 5% to 10% of all STSs. These tumors are associated with neurofibromatosis type 1 (NF-1), with 25% to 50% of tumors occurring in NF-1 patients. Additionally, most patients have a truncating lesion in the NF1 gene on chromosome 17.¹⁰ Anghileri et al in their single institution analysis of 205 patients with MPNSTs found the 2 most common presenting sites were the trunk and extremities. Histologically, these tumors have dense fascicles of spindle cells.¹⁰

GISTs are the most common STS of the gastrointestinal (GI) tract. Previously, GISTs were classified as smooth muscle tumors and were not accounted for in the literature as a separate entity distinct from leiomyomas, leiomyoblastomas, and leiomyosarcomas.¹¹ GISTs are found throughout the GI tract: the most common sites are the stomach (60%) and small intestine (30%). Less common sites include duodenum (4%–5%), esophagus (1%), rectum (1%–2%), and appendix (< 0.2%).¹² GISTs can be spindle cell, epithelioid, or mesenchymal tumors. Immunohistochemically, GISTs are KIT (CD117) positive. Other cell markers that are also commonly positive include CD34 (60%–70%) and smooth muscle actin (SMA) (25%).¹¹ The majority of GISTs (80%) have an activating c-KIT gene mutation. The most common mutation site is exon 11, with less common c-KIT gene mutations also occurring at exon 9 or 13. Not all GISTs have KIT mutations. The second most common mutation is the PDGFRA mutation (5%–10% of GISTs).² A minority of GISTs are negative for both KIT and PDGFRA mutations. These tumors were previously called wild-type, but as the majority have either a succinate dehydrogenase (SDH) loss of function or loss of SDHB protein expression, they are now referred to as SDH-deficient GISTs.² GISTs vary in aggressiveness from incidental to aggressive. Typically, small intestine and rectal GISTs are more aggressive than gastric GISTs. Both size and mitotic rate help to predict the metastatic potential of the tumor. Tumors less than 2 cm in size and having a mitotic rate of less than 5 per 50 high-power fields (hpf) have the lowest risk of metastases, while tumors greater than 5 cm and with more than 5 mitoses per 50 hpf have the highest rates of metastases.¹²

Undifferentiated sarcomas have no specific features and typically consist of primitive mesenchymal cells.

CLINICAL EVALUATION

CASE PRESENTATION

Initial Presentation and History

A 55-year-old man presents to his primary care physician with a painless mass in his anterior thigh. The mass has been present for the past 3 months and he believes that it is enlarging. The patient has a history of well-controlled hypertension and hyperlipidemia. His medications include atorvastatin and hydrochlorothiazide. He has no known drug allergies. Family history is notable for diabetes and hypertension. He drinks 4 to 5 alcoholic drinks a week and he is a former smoker. He quit smoking in his 30s and only smoked intermittently prior to quitting. He denies any illicit drug use. He works as a high school principal. Currently, he feels well. His review of systems is otherwise noncontributory.

Physical Examination

On physical exam, he is afebrile with a blood pressure of 132/75 mm Hg, respiratory rate of 10 breaths/min, and oxygen saturation of 99% on room air. He is a well appearing, overweight male. His head and neck exam is unremarkable. Lung exam reveals clear breath sounds, and cardiac exam reveals a regular rate and rhythm. His abdomen is obese, soft, and without hepatosplenomegaly. There is a large, fixed mass on the anterior lateral aspect of his right thigh. He has no appreciable lymphadenopathy. His neurological exam is unremarkable.

• What are risk factors for sarcoma?

There are few known risk factors for sarcoma. Established risks factors include prior radiation therapy, chronic lymphedema, viruses, and genetic cancer syndromes including Li-Fraumeni syndrome, hereditary retinoblastoma, and NF-1. Other environmental exposures include phenoxyacetic acids and chlorophenols.¹⁴ The majority of cases are sporadic, with only a minority of patients having one of these known risk factors.¹⁵ Up to one third of sarcomas have a specific translocation and are driven by fusion oncogenes (Table 1).

A painless mass is the most typical presenting symptom. Size at presentation varies based on location, with extremity and head and neck locations typically presenting at smaller sizes than retroperitoneal tumors.¹⁴ Patients may experience pain and numbness as the mass enlarges and impinges on surrounding structures including nerves and vasculature. The vast majority of patients are without systemic symptoms.

• How is sarcoma staged?

The American Joint Committee on Cancer (AJCC) staging system is the most widely used staging system in the United States. The latest AJCC manual was updated in 2010 to include a 3-tiered grading system where the tumor is classified according to tumor size, lymph node involvement, metastases, and grade at time of diagnosis (Table 2 and Table 3). Additionally, tumor depth in relation to deep fascia is also taken into account, with superficial tumors being assigned a designation of “a” and deep tumors a designation of “b.”

Previously, 2 of the most widely used grading systems were the National Cancer Institute (NCI) and French Federation of Cancer Centers Sarcoma Group (FNCLCC) systems, both 3-tier grading systems. The main components that determine the NCI grade are the tumor’s histologic type and location and the amount of tumor necrosis. The FNCLCC system evaluation focuses on tumor differentiation, mitotic rate, and amount of tumor necrosis. A study that compared the NCI and FNCLCC grading systems found that FNCLCC was a better predictor of mortality and distant metastasis.¹⁶ Previously, the AJCC was a 4-tier grading system, but the 2010 version was updated to the 3-tier FNCLCC grading system. Additionally, the AJCC system has reclassified single lymph node disease as stage III as it confers better survival than metastatic disease.¹⁷ It is important that pathology be evaluated by a sarcoma specialist as disagreements with regard to histologic subtype and grade are common.^18,19

• What are the most important prognostic factors?

Prognostic factors include grade, size, and presence of metastases at presentation. Best survival is associated with low-grade, small tumors with no metastases at time of diagnosis.¹⁴

Imaging should be undertaken to help differentiate between benign and malignant lesions. Ideally, it should be undertaken before a biopsy is planned as the imaging can be used to plan biopsy as well as provide invaluable prognostic information. There are several imaging modalities that should be considered during the preliminary work-up and staging of STSs. Conventional imaging includes magnetic resonance imaging (MRI) of the original tumor site; computed tomography (CT) to evaluate for pulmonary metastases and, depending on location, liver metastases; and in the case of small, low-grade tumors, chest radiography. MRI is considered the test of choice for soft tissue masses and can help delineate benign masses such as hematomas, lipomas, and hemangiomas from sarcomas.²⁰ It is difficult to compare the accuracy of positron emission tomography (PET)/CT to CT and MRI because most studies have evaluated PET/CT in parallel with CT and MRI.²¹ Tateishi et al compared the accuracy of conventional imaging, PET/CT, and PET/CT combined with conventional imaging at determining the TNM staging for 117 patients. They found that conventional imaging correctly classified 77% of patients, PET alone correctly classified 70%, PET/CT correctly classified 83%, and PET/CT combined with conventional imaging correctly staged 87%.²²

• Which subtypes are most likely to metastasize?

Although the vast majority of sarcomas spread hematogenously, 3 have a propensity to spread lymphogenously: epithelioid sarcoma, rhabdomyosarcoma, and clear-cell sarcoma. Additionally, certain subtypes are more likely to metastasize: leiomyosarcomas, synovial sarcomas, neurogenic sarcomas, rhabdomyosarcomas, and epithelioid sarcomas.²³ Sarcomas metastasize to the lungs more frequently than to the liver. The metastatic pattern is defined primarily by sarcoma subtype and site of primary tumor. Sarcomas rarely metastasize to the brain (~1%).

MANAGEMENT

CASE CONTINUED

The patient undergoes an ultrasound to better visualize the mass. Given the heterogeneous character of the mass, he is referred for an MRI to evaluate the mass and a CT scan of the chest, abdomen, and pelvis to evaluate for distant metastases. MRI reveals a 5.1 cm × 4.6 cm heterogeneous mass invading the superficial fascia of the rectus femoris muscle. No suspicious lymph nodes or other masses are identified on imaging. The patient next undergoes an image-guided core needle biopsy. Pathology from that procedure is consistent with a stage III, T2bNxMx, grade 3, dedifferentiated liposarcoma.

• What is the best management approach for this patient?

SURGERY

Surgery is the mainstay of treatment for STS. Patients with the best prognosis are those who undergo complete resection with negative surgical margins.^24,25 Goal tumor-free margin is 1 to 3 cm.²⁶ Complete resection confers the best long-term survival. Both local and metastatic recurrence is higher in patients with incomplete resection and positive margins.^24,25 In a study that analyzed 2084 localized primary STSs, patients with negative margins had a local recurrence rate of 15% versus a rate of 28% in patients with positive margins. This translated into higher 5-year local recurrence-free survival for patients with negative surgical margins (82%) compared to patients with positive margins (65%).²⁷ Another study similarly found that patients with negative margins at referral to their institution who underwent postoperative radiation had high local control rates of 93% (95% confidence interval [CI] 87% to 97%) at 5, 10, and 15 years.²⁶ Although radiation improves local control, neither preoperative or postoperative radiation has been shown to improve progression-free or overall survival.²⁸ Other factors that are associated with risk of recurrence are tumor location, history of previous recurrence, age of patient, histopathology, tumor grade, and tumor size. Approximately 40% to 50% of patients with high-grade tumors (defined as size > 5 cm, deep location, and high grade) will develop distant metastases.²⁹

Zagars et al found that positive or uncertain resection margin had a relative risk of local recurrence of 2.0 (95% CI 1.3 to 3.1; P = 0.002), and presentation with locally recurrent disease (vs new tumor) had a relative risk of local recurrence of 2.0 (95% CI 1.2 to 3.4; P = 0.013).²⁶ Patients with STS of head and neck and deep trunk have higher recurrence rates than those with superficial trunk and extremity STS. A single-institution retrospective review demonstrated that patients with completely resectable retroperitoneal sarcomas have longer median survival (103 months) compared to patients with incompletely resected abdominal sarcomas (18 months).²⁵

CASE CONTINUED

The patient is referred for limb-sparing surgery after presentation at a multidisciplinary tumor board. Prior to undergoing resection of the tumor, he is also referred to radiation-oncology to discuss the risks and benefits of combination radiotherapy and surgery as opposed to surgical resection alone.

• What is the evidence for radiation therapy?

RADIATION THERAPY

Radiation therapy is used in the preoperative, intraoperative, and postoperative settings to reduce the risk of local recurrence. There are several options for radiation, including external beam radiation therapy (EBRT), intraoperative radiation, and brachytherapy. A newer strategy, intensity-modulated radiation therapy (IMRT), utilizes 3-dimensional modeling to reduce radiation dosages. Overall there are no differences in overall survival or local recurrence rates between preoperative and postoperative radiation in STS.²⁸

The rationale behind preoperative radiation is that it reduces seeding of tumor cells, especially at the time of surgery.³⁰ Additionally, for EBRT, preoperative radiation has smaller field sizes and lower radiation doses. It can also help to reduce the size of the tumor prior to resection. Intraoperative radiation is often paired with preoperative radiation as a boost dose given only to the area of residual tumor.

Suit et al reviewed patients treated at a single institution with limb-sparing surgery and different radiation strategies. Local control rates between preoperative and postoperative radiation groups were not statistically significant. Local recurrence was linked to grade and size of the tumor in both groups. The authors did note, however, that the preoperative radiation group tended to have larger tumor sizes at baseline compared to the patients who received postoperative radiation.³⁰ A study that compared 190 patients who received preoperative and postoperative EBRT or brachytherapy (primary end point was wound complications, and local control was a secondary end point) showed a trend towards greater local control with preoperative radiation; however, the preoperative radiation group had significantly more wound complications compared to the postoperative radiation group.³¹

Yang et al found that postoperative EBRT decreases rates of local recurrence compared to surgery alone in high-grade extremity sarcomas.³² However, there were no differences in rates of distant metastases and overall survival between the 2 treatment groups. Similarly, in patients with low-grade sarcoma, there were fewer local recurrences in those who received EBRT and surgery as compared to surgery alone.³² Another study that evaluated 164 patients who received either adjuvant brachytherapy or no further therapy after complete resection found that brachytherapy reduced local recurrence in high-grade sarcomas. No difference in local recurrence rates was found in patients with low-grade sarcomas, nor was a significant difference found in the rates of distant metastases and overall survival between the 2 treatment groups.³³ With regards to IMRT, a single institution cohort experience with 41 patients who received IMRT following limb-sparing surgery had similar local control rates when compared to historical controls.³⁴

CASE CONTINUED

After discussion of the risks and benefits of radiation therapy, the patient opts for preoperative radiation prior to resection of his liposarcoma. He receives 50 Gy of EBRT prior to undergoing resection. Resection results in R1 margin consistent with microscopic disease. He receives 16 Gy of EBRT as a boost after recovery from his resection.²

• What is the evidence for neoadjuvant and adjuvant chemotherapy for stage I tumors?

CHEMOTHERAPY

Localized Sarcoma

For localized sarcoma, limb-sparing resection with or without radiation forms the backbone of treatment. Studies have evaluated chemotherapy in both the neoadjuvant and adjuvant settings, with the vast majority of studies evaluating doxorubicin-based chemotherapy regimens in the adjuvant settings. Due to the rare nature of sarcomas, most studies are not sufficiently powered to detect significant benefit from chemotherapy. Several trials evaluating chemotherapy regimens in the neoadjuvant and adjuvant settings needed to be terminated prematurely due to inadequate enrollment into the study. ^35,36

For stage IA (T1a-Tb, N0, M0, low grade) tumors, no additional therapy is recommended after limb-sparing surgery with appropriate surgical margins. For stage IB (T2a-2b, N0, M0, low grade) tumors with insufficient margins, re-resection and radiation therapy should be considered, while for stage IIA (T1a-1b, N0, M0, G2-3) tumors preoperative or postoperative radiation therapy is recommended.² Studies have not found benefit of adjuvant chemotherapy in these low-grade, stage I tumors in terms of progression-free survival and overall survival.³⁷

• At what stage should chemotherapy be considered?

For stage IIb and stage III tumors, surgery and radiation therapy again form the backbone of therapy; however, neoadjuvant and adjuvant chemotherapy are also recommended as considerations. Anthracycline-based chemotherapy with either single-agent doxorubicin or doxorubicin and ifosfamide in combination are considered first-line chemotherapy agents in locally advanced STS.^2,29,37

Evidence regarding the efficacy of both neoadjuvant and adjuvant chemotherapy regimens in the setting of locally advanced high-grade STS has been mixed. The Sarcoma Meta-analysis Collaboration evaluated 14 trials of doxorubicin-based adjuvant chemotherapy and found a trend towards overall survival in the treatment groups that received chemotherapy.³⁷ All trials included in the meta-analysis compared patients with localized resectable soft-tissue sarcomas who were randomized to either adjuvant chemotherapy or no adjuvant chemotherapy after limb-sparing surgery with or without radiation therapy. None of the individual trials showed a significant benefit, and all trials had large confidence intervals; however, the meta-analysis showed significant benefit in the chemotherapy treatment groups with regard to local recurrence, distant recurrence, and progression-free survival. No significant difference in overall survival was found.³⁷ Pervais et al updated the Sarcoma Meta-analysis Collaboration’s 1997 meta-analysis with the inclusion of 4 new trials that evaluated doxorubicin combined with ifosfamide and found that both patients who received doxorubicin-based regimens or doxorubicin with ifosfamide had significant decreases in distant and overall recurrences. Only the trials that utilized doxorubicin and ifosfamide had an improved overall survival that was statistically significant (hazard ratio 0.56 [95% CI 0.36 to 0.85]; P = 0.01).²⁹ Although no significant heterogeneity was found among the trials included in either meta-analysis, a variety of sarcomas were included in each clinical trial evaluated. Given the extremely small number of each sarcoma subtype present in each trial, subgroup analysis is difficult and prone to inaccuracies. As a result, it is not known if certain histological subtypes are more or less responsive to chemotherapy.^37–39

One randomized controlled trial evaluated neoadjuvant chemotherapy in high-risk sarcomas defined as tumors greater than 8 cm or grade II/III tumors. This study evaluated doxorubicin and ifosfamide and found no significant difference in disease-free and overall survival in the neoadjuvant therapy group compared to the control group.³⁵ There remains controversy in the literature with regards to adjuvant chemotherapy. Many oncologists offer adjuvant chemotherapy to patients with certain stage III subtypes. Examples of subtypes that may be offered adjuvant therapy include myxoid liposarcomas, synovial sarcomas, and leiomyosarcomas.² With regards to how many cycles of chemotherapy should be considered, a noninferiority study compared 3 cycles of epirubicin and ifosfamide to 5 cycles of epirubicin and ifosfamide in patients with high-risk locally advanced adult STSs. Three cycles of preoperative epirubicin and ifosfamide was found to be noninferior to 5 cycles with regards to overall survival.³⁸

• What is this patient’s risk for recurrence?

The patient is at intermediate risk for recurrence. Numerous studies have demonstrated that tumor size, grade, and location are the most important factors to determine risk of recurrence, with larger size, higher grades, and deeper locations being associated with higher risk of recurrence. In an analysis of 1041 patients with STS of the extremities, high grade was the most important risk factor for distant metastases.³⁹ The highest risk of recurrence is within the first 2 years. Given that the patient’s initial tumor was located in the extremity, he is more likely to have a distant metastasis as his site of recurrence; individuals with retroperitoneal tumors and visceral tumors are more likely to recur locally.⁴⁰ For STSs of the extremity, distant metastases determine overall survival, whereas patients with retroperitoneal sarcomas can die from complications of local metastases.⁴¹ Once a patient develops distant metastases, the most important prognostic factor is the size of the tumor, with tumors larger than 10 cm having a relative risk of 1.5 (95% CI 1.0 to 2.0).³⁹

• What are the recommendations for surveillance?

Surveillance recommendations are based on the stage of the sarcoma. Stage I tumors are the least likely to recur either locally or distally. As a result, it is recommended that stage I tumors be followed with history and physical exam every 3 to 6 months for the first 2 to 3 years, and then annually after the first 2 to 3 years. Chest x-rays should be considered every 6 to 12 months.² For stage II–IV tumors, history and physical exam is recommended every 3 to 6 months for the first 2 to 3 years. Chest and distant metastases imaging should also be performed every 3 to 6 months during this time frame. For the next 2 years, history and physical exam and imaging are recommended every 6 months. After the first 4 to 5 years, annual follow-up is recommended.²

CASE CONTINUED

The patient does well for 1 year. With physical therapy, he regains most of the strength and coordination of the lower extremity. He is followed every 3 months with chest x-rays and a MRI of the thigh for the first year. On his fourth follow-up clinic visit, he describes increased dyspnea on exertion over the previous few weeks and is found to have multiple lung metastases in both lungs on chest x-ray. He undergoes further evaluation for metastases and is not found to have any other metastatic lesions. Bronchoscopy and biopsy of 1 of the lung nodules confirms recurrent dedifferentiated liposarcoma.

• Should this patient undergo metastectomy?

An analysis of 3149 patients with STS treated at Memorial Sloan-Kettering who developed lung metastases found that patients with pulmonary metastases have survival rates of 25%. The most important prognostic factor for survival was complete resection of all metastases.⁴² For stage IV disease, surgery is used only in certain instances. In instances where tumor is more localized or limited, removal of metastases or metastectomy can play a role in management.²

CASE CONTINUED

Because the patient’s metastases are limited to the lungs, he is referred for metastectomy. He undergoes wedge resection for definitive diagnosis but it is not possible to completely resect all of the metastases. He is thus referred to a medical oncologist to discuss his treatment options.

• What are treatment options for unresectable or metastatic disease?

Metastatic Disease

Unlike local and locally advanced disease, chemotherapy forms the backbone of treatment in stage IV disease. Doxorubicin and olaratumab or doxorubicin and ifosfamide in combination are considered first line in metastatic disease. Response rates for single-agent doxorubicin range from 16% to 27%, while phase 2 and phase 3 studies of doxorubicin and ifosfamide have found response rates ranging from 18% to 36%.⁴³ In addition, the effectiveness of doxorubicin and ifosfamide phase 2 and 3 trials varied. Edmonson et al found a tumor regression rate of 34% for doxorubicin and ifosfamide as compared to 20% for doxorubicin alone.⁴⁴ In comparison, Santoro et al found a response rate of 21.3% for doxorubicin alone and 25.2% for doxorubicin and ifosfamide.⁴⁵ Neither study found increased survival benefit for doxorubicin and ifosfamide when compared to doxorubicin alone. In a Cochrane review evaluating randomized trials that compared doxorubicin and combination chemotherapy regimens, response rates varied from 14% for doxorubicin in combination with streptomycin to 34% for doxorubicin and ifosfamide. Most trials did not show a significant benefit for combination therapies when compared to doxorubicin alone.⁴³ Mean survival with doxorubicin or doxorubicin and ifosfamide is 12 months. High rates of recurrence highlight the need for additional chemotherapy regimens.

The newest approved agent is olaratumab, a monoclonal antibody that binds platelet-derived growth factor receptor alpha and prevents receptor activation. A phase 1-b and phase 2 trial evaluated patients with locally advanced and metastatic STS and randomly assigned them to either olaratumab and doxorubicin or doxorubicin alone.⁴⁶ Progression-free survival for olaratumab/doxorubicin was 6.6 months (95% CI 4.1 to 8.3) compared to 4.1 months (95% CI 2.8 to 5.4) for doxorubicin alone. The objective response rate was 18.2% (95% CI 9.8 to 29.6) for olaratumab/doxorubicin compared to 7.5% (95% CI 2.5 to 6.6) for doxorubicin alone. Furthermore, the median overall survival for olaratumab plus doxorubicin was 26.5 months (95% CI 20.9 to 31.7) compared to 14.7 months for doxorubicin alone (95% CI 5.5 to 26.0). Impressively, this improved response was notable across histological types. Furthermore, patients who had previously been treated with more than 1 regimen and those who were treatment naïve had similar response rates.⁴⁶

• What are second-line treatment options?

Doxorubicin has been used in combination with several other agents including dacarbazine (DTIC) as well as DTIC and ifosfamide (MAID). Borden et al evaluated patients with metastatic STS and randomly assigned the patients to either doxorubicin or doxorubicin and DTIC. Combination therapy demonstrated better tumor response than doxorubicin alone: 30% complete or partial response for combination therapy and 18% for doxorubicin alone.⁴⁷ However, Omura et al

Several additional regimens have undergone evaluation in metastatic and recurrent STSs. Gemcitabine has been used both as a single agent and as part of combination therapy in many studies. Studies with gemcitabine in combination with either docetaxel or DTIC have been the most efficacious. In a phase 2 trial, patients with metastatic STS were randomly assigned to either gemcitabine alone or gemcitabine and docetaxel. Combination therapy had a higher response rate (16% versus 8%) and longer overall survival (17.9 months versus 11.5 months) than gemcitabine alone.⁵⁰ Furthermore, a phase 2 trial of gemcitabine and docetaxel in patients with unresectable leiomyosarcoma showed an overall response rate of 56%, with 3 complete and 15 partial responses among the 34 patients enrolled in the study.⁵¹

A phase 2 trial randomly assigned patients with unresectable or metastatic STS to either DTIC or combination gemcitabine and DTIC.⁵² Gemcitabine-DTIC had a superior progression-free survival at 3 months (56% [95% CI 43% to 69%]) as compared to DTIC alone (37% [95% CI 23.5% to 50%]). Furthermore, mean progression-free survival and overall survival were improved in the gemcitabine-DTIC group (4.2 months and 16.8 months) as compared to the DTIC group (2.0 months and 8.2 months).⁵² DTIC has a single-agent response rate of 16%, but has been shown to be particularly effective in the setting of leiomyosarcomas.⁴⁹

• Does response to treatment regimens differ by histologic subtype?

The majority of STS trials include many different histologic subtypes. Given the rarity of sarcomas as a whole, many trials have had difficulty recruiting adequate numbers of patients to have sufficient power to definitely determine if the treatment under investigation has clinical benefit. Furthermore, the patients recruited have been heterogeneous with regard to subtype. Many older studies hypothesized that the efficacy of chemotherapeutic agents vary based on histologic subtype; however, for most subtypes the number of individuals included in those trials was too low to evaluate efficacy based on subtype.

Some exceptions exist, however. For example, both gemcitabine-DTIC and gemcitabine-docetaxel have been found to be particularly effective in the treatment of leiomyosarcomas.^50,52 Additionally, a retrospective study found a 51% overall response rate for patients with myxoid liposarcomas treated with trabectedin.⁵³ Studies of patients with angiosarcoma treated with paclitaxel have demonstrated response rates of 43% and 53%.^54,55

• What are the newest approved and investigational agents?

A recently approved agent is trabectedin, a tris tetrahydroisoquinoline alkaloid isolated from ascidians that binds to the minor groove of DNA and causes disruptions in the cell cycle. Samuels et al reported data from a single-arm, open-label expanded access trial that evaluated patients with advanced metastatic sarcomas.⁵⁶ In this study, patients with liposarcomas and leiomyosarcomas had an objective response rate of 6.9% (95% CI 4.8 to 9.6) as compared to a rate of 5.9% (95% CI 4.4 to 7.8) for all assessable patients. Median survival was 11.9 months for all patients, with improved median survivals for liposarcoma and leiomyosarcomas of 16.2 months (95% CI 14.1 to 19.5) compared to 8.4 months (95% CI 7.1 to 10.7 months) for other subtypes.⁵⁶

Schöffski et al evaluated eribulin, a chemotherapeutic agent that affects microtubule dynamics, in a phase 2 trial of patients with progressive or high-grade STS with progression on previous chemotherapy. They found a median progression-free survival of 2.6 months (95% CI 1.7 to 6.2) for adipocytic sarcoma, 2.9 months (95% CI 2.4 to 4.6) for leiomyosarcoma, 2.6 months (95% CI 2.3 to 4.3) for synovial sarcoma, and 2.1 months (95% CI 1.4 to 2.9) for other sarcomas.⁵⁷

Van der Graaf and colleagues randomly assigned patients with metastatic nonadipocytic STS to pazopanib or placebo in a phase 3 trial. Pazopanib is a small-molecule endothelial growth factor inhibitor with activity against vascular endothelial growth factors 1, 2, and 3 as well as platelet-derived growth factors. Median progression-free survival was 4.6 months (95% CI 3.7 to 4.8) with pazopanib compared to 1.6 months (95% CI 0.9 to 1.8) with placebo.⁵⁸ Adipocytic sarcomas (liposarcomas) were excluded from the trial because phase 2 trials had found a lower rate of progression-free survival (26%) for them compared to other subtypes.

Toxicities were seen with each of the regimens studied and were common in the randomized trials, with higher rates of toxicities in the combination chemotherapy regimens. The most common toxicities are myelosuppression, nausea, and vomiting. In the doxorubicin trials, the most common toxicities were myelosuppression, nausea, and vomiting.⁴⁴

Ifosfamide both as an individual agent and in combination with doxorubicin has higher rates and higher grades of toxicity than doxorubicin alone. Myelosuppression is the most common toxicity associated with ifosfamide, and the most commonly affected cell line is leukocytes.⁴⁴ Combination doxorubicin and ifosfamide also had high rates of nausea and vomiting (95%) and alopecia (100%).³⁵

Neutropenia is the most common toxicity associated with gemcitabine and dacarbazine, while their most common nonhematologic toxicities are fatigue and nausea.^52,59 Trabectedin’s most common toxicities are nausea (29%), neutropenia (24%), and fatigue (23%). It has also been shown to cause increased alkaline phosphatase (20%) and alanine aminotransferase (19%) levels.⁵⁶ In a phase 2 study of eribulin, 50% of patients had neutropenia, and other toxicities included fatigue, alopecia, nausea, sensory neuropathy, and thrombocytopenia.⁵⁷ Pazopanib is generally well tolerated; the most common toxicities are fatigue (65%), diarrhea (58%), nausea (54%), and hypertension (41%).⁵⁸ Higher rates of neutropenia, mucositis, nausea, vomiting, diarrhea, and transfusion reactions were seen with olaratumab and doxorubicin compared to doxorubicin alone in phase 1b and 2 studies.⁴⁶

CASE CONCLUSION

Given his poor prognosis with unresectable metastatic undifferentiated liposarcoma, the patient considers a clinical trial prior to undergoing combined therapy with doxorubicin and ifosfamide. He tolerates therapy well with stable disease at 6 months.

CONCLUSION

STSs are a heterogeneous collection of rare tumors. Low-grade, localized tumors have the best prognosis, and patients who undergo complete resection have the best long-term survival. Due to the rarity of STSs, trials often have limited enrollment, and little progress has been made with regards to treatment and survival rates for metastatic and unresectable disease. All patients should be evaluated and treated at specialized sarcoma centers. This case highlights the need for continued research and clinical trials to improve overall survival of patients with sarcoma.

INTRODUCTION

Soft tissue sarcomas (STSs) are rare adult tumors, with 3.4 new cases per 100,000 persons or 12,310 expected new cases in 2016.¹ Sarcomas are a heterogeneous collection of tumors that affect fat, muscle, nerve, nerve sheath, vascular, and connective tissues. There are more than 50 histological subtypes that comprise this diverse category of tumors. Treatment varies by stage, with limb-sparing surgery representing the mainstay of curative-intent treatment. Radiation and chemotherapy may also be considered depending on the size, grade, and location of the tumor. Survival rates have been stagnant until recently, with a disease-specific survival hovering around 65%.¹ Given the complexity of these cases, all patients ideally should be evaluated and treated by a multidisciplinary team at an institution with extensive experience treating STS.²

EPIDEMIOLOGY AND CLASSIFICATION

The most common STS subtypes are gastrointestinal stromal tumor (GIST), undifferentiate pleomorphic sarcoma (previously referred to as malignant fibrous histiocytoma), liposarcoma, leiomyosarcoma, synovial sarcoma, malignant peripheral nerve sheath tumor, rhabdomyosarcoma, and unclassified sarcoma.³ Liposarcoma is one of the most common subtypes, comprising 20% of all STSs; it is subdivided into well-differentiated/dedifferentiated liposarcomas, myxoid/round cell liposarcomas, and pleomorphic liposarcomas. Well-differentiated liposarcomas tend to occur in the retroperitoneum and limbs, while both myxoid and round cell as well as pleomorphic liposarcomas more commonly originate on the limbs. Histology varies based on subtype and ranges from mature-appearing adipocytes and fibroblasts to undifferentiated cells with minimal lipogenic differentiation.⁴

Leiomyosarcomas are smooth muscle tumors and are usually located in the retroperitoneum, but have also been associated with peripheral soft tissue and vasculature. Typical histology ranges from well-defined areas of spindle-shaped cells to poorly differentiated anaplastic spindle cells.^5,6 Synovial sarcomas are a distinct type of STS that can show epithelial differentiation and account for 5% of adult STSs. The extremities are the most common presenting location (90%).⁷

Rhabdomyosarcomas are skeletal muscle tumors and are further subdivided into embryonal, alveolar, and pleomorphic subtypes. Embryonal histology ranges from primitive mesenchymal-appearing cells to highly differentiated muscle cells. Alveolar rhabdomyosarcoma has the worst prognosis of the subtypes and consists of round cells with high nuclear-to-chromatin ratios that form “glandular-like” or “alveolar” spaces.⁸ Pleomorphic rhabdomyosarcomas are composed of rhabdomyoblasts that can affect many different locations, but most commonly present on the lower extremities.⁹

Malignant peripheral nerve sheath tumor (MPNST) comprises 5% to 10% of all STSs. These tumors are associated with neurofibromatosis type 1 (NF-1), with 25% to 50% of tumors occurring in NF-1 patients. Additionally, most patients have a truncating lesion in the NF1 gene on chromosome 17.¹⁰ Anghileri et al in their single institution analysis of 205 patients with MPNSTs found the 2 most common presenting sites were the trunk and extremities. Histologically, these tumors have dense fascicles of spindle cells.¹⁰

GISTs are the most common STS of the gastrointestinal (GI) tract. Previously, GISTs were classified as smooth muscle tumors and were not accounted for in the literature as a separate entity distinct from leiomyomas, leiomyoblastomas, and leiomyosarcomas.¹¹ GISTs are found throughout the GI tract: the most common sites are the stomach (60%) and small intestine (30%). Less common sites include duodenum (4%–5%), esophagus (1%), rectum (1%–2%), and appendix (< 0.2%).¹² GISTs can be spindle cell, epithelioid, or mesenchymal tumors. Immunohistochemically, GISTs are KIT (CD117) positive. Other cell markers that are also commonly positive include CD34 (60%–70%) and smooth muscle actin (SMA) (25%).¹¹ The majority of GISTs (80%) have an activating c-KIT gene mutation. The most common mutation site is exon 11, with less common c-KIT gene mutations also occurring at exon 9 or 13. Not all GISTs have KIT mutations. The second most common mutation is the PDGFRA mutation (5%–10% of GISTs).² A minority of GISTs are negative for both KIT and PDGFRA mutations. These tumors were previously called wild-type, but as the majority have either a succinate dehydrogenase (SDH) loss of function or loss of SDHB protein expression, they are now referred to as SDH-deficient GISTs.² GISTs vary in aggressiveness from incidental to aggressive. Typically, small intestine and rectal GISTs are more aggressive than gastric GISTs. Both size and mitotic rate help to predict the metastatic potential of the tumor. Tumors less than 2 cm in size and having a mitotic rate of less than 5 per 50 high-power fields (hpf) have the lowest risk of metastases, while tumors greater than 5 cm and with more than 5 mitoses per 50 hpf have the highest rates of metastases.¹²

Undifferentiated sarcomas have no specific features and typically consist of primitive mesenchymal cells.

CLINICAL EVALUATION

CASE PRESENTATION

Initial Presentation and History

A 55-year-old man presents to his primary care physician with a painless mass in his anterior thigh. The mass has been present for the past 3 months and he believes that it is enlarging. The patient has a history of well-controlled hypertension and hyperlipidemia. His medications include atorvastatin and hydrochlorothiazide. He has no known drug allergies. Family history is notable for diabetes and hypertension. He drinks 4 to 5 alcoholic drinks a week and he is a former smoker. He quit smoking in his 30s and only smoked intermittently prior to quitting. He denies any illicit drug use. He works as a high school principal. Currently, he feels well. His review of systems is otherwise noncontributory.

Physical Examination

On physical exam, he is afebrile with a blood pressure of 132/75 mm Hg, respiratory rate of 10 breaths/min, and oxygen saturation of 99% on room air. He is a well appearing, overweight male. His head and neck exam is unremarkable. Lung exam reveals clear breath sounds, and cardiac exam reveals a regular rate and rhythm. His abdomen is obese, soft, and without hepatosplenomegaly. There is a large, fixed mass on the anterior lateral aspect of his right thigh. He has no appreciable lymphadenopathy. His neurological exam is unremarkable.

• What are risk factors for sarcoma?

There are few known risk factors for sarcoma. Established risks factors include prior radiation therapy, chronic lymphedema, viruses, and genetic cancer syndromes including Li-Fraumeni syndrome, hereditary retinoblastoma, and NF-1. Other environmental exposures include phenoxyacetic acids and chlorophenols.¹⁴ The majority of cases are sporadic, with only a minority of patients having one of these known risk factors.¹⁵ Up to one third of sarcomas have a specific translocation and are driven by fusion oncogenes (Table 1).

A painless mass is the most typical presenting symptom. Size at presentation varies based on location, with extremity and head and neck locations typically presenting at smaller sizes than retroperitoneal tumors.¹⁴ Patients may experience pain and numbness as the mass enlarges and impinges on surrounding structures including nerves and vasculature. The vast majority of patients are without systemic symptoms.

• How is sarcoma staged?

The American Joint Committee on Cancer (AJCC) staging system is the most widely used staging system in the United States. The latest AJCC manual was updated in 2010 to include a 3-tiered grading system where the tumor is classified according to tumor size, lymph node involvement, metastases, and grade at time of diagnosis (Table 2 and Table 3). Additionally, tumor depth in relation to deep fascia is also taken into account, with superficial tumors being assigned a designation of “a” and deep tumors a designation of “b.”

Previously, 2 of the most widely used grading systems were the National Cancer Institute (NCI) and French Federation of Cancer Centers Sarcoma Group (FNCLCC) systems, both 3-tier grading systems. The main components that determine the NCI grade are the tumor’s histologic type and location and the amount of tumor necrosis. The FNCLCC system evaluation focuses on tumor differentiation, mitotic rate, and amount of tumor necrosis. A study that compared the NCI and FNCLCC grading systems found that FNCLCC was a better predictor of mortality and distant metastasis.¹⁶ Previously, the AJCC was a 4-tier grading system, but the 2010 version was updated to the 3-tier FNCLCC grading system. Additionally, the AJCC system has reclassified single lymph node disease as stage III as it confers better survival than metastatic disease.¹⁷ It is important that pathology be evaluated by a sarcoma specialist as disagreements with regard to histologic subtype and grade are common.^18,19

• What are the most important prognostic factors?

Prognostic factors include grade, size, and presence of metastases at presentation. Best survival is associated with low-grade, small tumors with no metastases at time of diagnosis.¹⁴

Imaging should be undertaken to help differentiate between benign and malignant lesions. Ideally, it should be undertaken before a biopsy is planned as the imaging can be used to plan biopsy as well as provide invaluable prognostic information. There are several imaging modalities that should be considered during the preliminary work-up and staging of STSs. Conventional imaging includes magnetic resonance imaging (MRI) of the original tumor site; computed tomography (CT) to evaluate for pulmonary metastases and, depending on location, liver metastases; and in the case of small, low-grade tumors, chest radiography. MRI is considered the test of choice for soft tissue masses and can help delineate benign masses such as hematomas, lipomas, and hemangiomas from sarcomas.²⁰ It is difficult to compare the accuracy of positron emission tomography (PET)/CT to CT and MRI because most studies have evaluated PET/CT in parallel with CT and MRI.²¹ Tateishi et al compared the accuracy of conventional imaging, PET/CT, and PET/CT combined with conventional imaging at determining the TNM staging for 117 patients. They found that conventional imaging correctly classified 77% of patients, PET alone correctly classified 70%, PET/CT correctly classified 83%, and PET/CT combined with conventional imaging correctly staged 87%.²²

• Which subtypes are most likely to metastasize?

Although the vast majority of sarcomas spread hematogenously, 3 have a propensity to spread lymphogenously: epithelioid sarcoma, rhabdomyosarcoma, and clear-cell sarcoma. Additionally, certain subtypes are more likely to metastasize: leiomyosarcomas, synovial sarcomas, neurogenic sarcomas, rhabdomyosarcomas, and epithelioid sarcomas.²³ Sarcomas metastasize to the lungs more frequently than to the liver. The metastatic pattern is defined primarily by sarcoma subtype and site of primary tumor. Sarcomas rarely metastasize to the brain (~1%).

MANAGEMENT

CASE CONTINUED

The patient undergoes an ultrasound to better visualize the mass. Given the heterogeneous character of the mass, he is referred for an MRI to evaluate the mass and a CT scan of the chest, abdomen, and pelvis to evaluate for distant metastases. MRI reveals a 5.1 cm × 4.6 cm heterogeneous mass invading the superficial fascia of the rectus femoris muscle. No suspicious lymph nodes or other masses are identified on imaging. The patient next undergoes an image-guided core needle biopsy. Pathology from that procedure is consistent with a stage III, T2bNxMx, grade 3, dedifferentiated liposarcoma.

• What is the best management approach for this patient?

SURGERY

Surgery is the mainstay of treatment for STS. Patients with the best prognosis are those who undergo complete resection with negative surgical margins.^24,25 Goal tumor-free margin is 1 to 3 cm.²⁶ Complete resection confers the best long-term survival. Both local and metastatic recurrence is higher in patients with incomplete resection and positive margins.^24,25 In a study that analyzed 2084 localized primary STSs, patients with negative margins had a local recurrence rate of 15% versus a rate of 28% in patients with positive margins. This translated into higher 5-year local recurrence-free survival for patients with negative surgical margins (82%) compared to patients with positive margins (65%).²⁷ Another study similarly found that patients with negative margins at referral to their institution who underwent postoperative radiation had high local control rates of 93% (95% confidence interval [CI] 87% to 97%) at 5, 10, and 15 years.²⁶ Although radiation improves local control, neither preoperative or postoperative radiation has been shown to improve progression-free or overall survival.²⁸ Other factors that are associated with risk of recurrence are tumor location, history of previous recurrence, age of patient, histopathology, tumor grade, and tumor size. Approximately 40% to 50% of patients with high-grade tumors (defined as size > 5 cm, deep location, and high grade) will develop distant metastases.²⁹

Zagars et al found that positive or uncertain resection margin had a relative risk of local recurrence of 2.0 (95% CI 1.3 to 3.1; P = 0.002), and presentation with locally recurrent disease (vs new tumor) had a relative risk of local recurrence of 2.0 (95% CI 1.2 to 3.4; P = 0.013).²⁶ Patients with STS of head and neck and deep trunk have higher recurrence rates than those with superficial trunk and extremity STS. A single-institution retrospective review demonstrated that patients with completely resectable retroperitoneal sarcomas have longer median survival (103 months) compared to patients with incompletely resected abdominal sarcomas (18 months).²⁵

CASE CONTINUED

The patient is referred for limb-sparing surgery after presentation at a multidisciplinary tumor board. Prior to undergoing resection of the tumor, he is also referred to radiation-oncology to discuss the risks and benefits of combination radiotherapy and surgery as opposed to surgical resection alone.

• What is the evidence for radiation therapy?

RADIATION THERAPY

Radiation therapy is used in the preoperative, intraoperative, and postoperative settings to reduce the risk of local recurrence. There are several options for radiation, including external beam radiation therapy (EBRT), intraoperative radiation, and brachytherapy. A newer strategy, intensity-modulated radiation therapy (IMRT), utilizes 3-dimensional modeling to reduce radiation dosages. Overall there are no differences in overall survival or local recurrence rates between preoperative and postoperative radiation in STS.²⁸

The rationale behind preoperative radiation is that it reduces seeding of tumor cells, especially at the time of surgery.³⁰ Additionally, for EBRT, preoperative radiation has smaller field sizes and lower radiation doses. It can also help to reduce the size of the tumor prior to resection. Intraoperative radiation is often paired with preoperative radiation as a boost dose given only to the area of residual tumor.

Suit et al reviewed patients treated at a single institution with limb-sparing surgery and different radiation strategies. Local control rates between preoperative and postoperative radiation groups were not statistically significant. Local recurrence was linked to grade and size of the tumor in both groups. The authors did note, however, that the preoperative radiation group tended to have larger tumor sizes at baseline compared to the patients who received postoperative radiation.³⁰ A study that compared 190 patients who received preoperative and postoperative EBRT or brachytherapy (primary end point was wound complications, and local control was a secondary end point) showed a trend towards greater local control with preoperative radiation; however, the preoperative radiation group had significantly more wound complications compared to the postoperative radiation group.³¹

Yang et al found that postoperative EBRT decreases rates of local recurrence compared to surgery alone in high-grade extremity sarcomas.³² However, there were no differences in rates of distant metastases and overall survival between the 2 treatment groups. Similarly, in patients with low-grade sarcoma, there were fewer local recurrences in those who received EBRT and surgery as compared to surgery alone.³² Another study that evaluated 164 patients who received either adjuvant brachytherapy or no further therapy after complete resection found that brachytherapy reduced local recurrence in high-grade sarcomas. No difference in local recurrence rates was found in patients with low-grade sarcomas, nor was a significant difference found in the rates of distant metastases and overall survival between the 2 treatment groups.³³ With regards to IMRT, a single institution cohort experience with 41 patients who received IMRT following limb-sparing surgery had similar local control rates when compared to historical controls.³⁴

CASE CONTINUED

After discussion of the risks and benefits of radiation therapy, the patient opts for preoperative radiation prior to resection of his liposarcoma. He receives 50 Gy of EBRT prior to undergoing resection. Resection results in R1 margin consistent with microscopic disease. He receives 16 Gy of EBRT as a boost after recovery from his resection.²

• What is the evidence for neoadjuvant and adjuvant chemotherapy for stage I tumors?

CHEMOTHERAPY

Localized Sarcoma

For localized sarcoma, limb-sparing resection with or without radiation forms the backbone of treatment. Studies have evaluated chemotherapy in both the neoadjuvant and adjuvant settings, with the vast majority of studies evaluating doxorubicin-based chemotherapy regimens in the adjuvant settings. Due to the rare nature of sarcomas, most studies are not sufficiently powered to detect significant benefit from chemotherapy. Several trials evaluating chemotherapy regimens in the neoadjuvant and adjuvant settings needed to be terminated prematurely due to inadequate enrollment into the study. ^35,36

For stage IA (T1a-Tb, N0, M0, low grade) tumors, no additional therapy is recommended after limb-sparing surgery with appropriate surgical margins. For stage IB (T2a-2b, N0, M0, low grade) tumors with insufficient margins, re-resection and radiation therapy should be considered, while for stage IIA (T1a-1b, N0, M0, G2-3) tumors preoperative or postoperative radiation therapy is recommended.² Studies have not found benefit of adjuvant chemotherapy in these low-grade, stage I tumors in terms of progression-free survival and overall survival.³⁷

• At what stage should chemotherapy be considered?

For stage IIb and stage III tumors, surgery and radiation therapy again form the backbone of therapy; however, neoadjuvant and adjuvant chemotherapy are also recommended as considerations. Anthracycline-based chemotherapy with either single-agent doxorubicin or doxorubicin and ifosfamide in combination are considered first-line chemotherapy agents in locally advanced STS.^2,29,37

Evidence regarding the efficacy of both neoadjuvant and adjuvant chemotherapy regimens in the setting of locally advanced high-grade STS has been mixed. The Sarcoma Meta-analysis Collaboration evaluated 14 trials of doxorubicin-based adjuvant chemotherapy and found a trend towards overall survival in the treatment groups that received chemotherapy.³⁷ All trials included in the meta-analysis compared patients with localized resectable soft-tissue sarcomas who were randomized to either adjuvant chemotherapy or no adjuvant chemotherapy after limb-sparing surgery with or without radiation therapy. None of the individual trials showed a significant benefit, and all trials had large confidence intervals; however, the meta-analysis showed significant benefit in the chemotherapy treatment groups with regard to local recurrence, distant recurrence, and progression-free survival. No significant difference in overall survival was found.³⁷ Pervais et al updated the Sarcoma Meta-analysis Collaboration’s 1997 meta-analysis with the inclusion of 4 new trials that evaluated doxorubicin combined with ifosfamide and found that both patients who received doxorubicin-based regimens or doxorubicin with ifosfamide had significant decreases in distant and overall recurrences. Only the trials that utilized doxorubicin and ifosfamide had an improved overall survival that was statistically significant (hazard ratio 0.56 [95% CI 0.36 to 0.85]; P = 0.01).²⁹ Although no significant heterogeneity was found among the trials included in either meta-analysis, a variety of sarcomas were included in each clinical trial evaluated. Given the extremely small number of each sarcoma subtype present in each trial, subgroup analysis is difficult and prone to inaccuracies. As a result, it is not known if certain histological subtypes are more or less responsive to chemotherapy.^37–39

One randomized controlled trial evaluated neoadjuvant chemotherapy in high-risk sarcomas defined as tumors greater than 8 cm or grade II/III tumors. This study evaluated doxorubicin and ifosfamide and found no significant difference in disease-free and overall survival in the neoadjuvant therapy group compared to the control group.³⁵ There remains controversy in the literature with regards to adjuvant chemotherapy. Many oncologists offer adjuvant chemotherapy to patients with certain stage III subtypes. Examples of subtypes that may be offered adjuvant therapy include myxoid liposarcomas, synovial sarcomas, and leiomyosarcomas.² With regards to how many cycles of chemotherapy should be considered, a noninferiority study compared 3 cycles of epirubicin and ifosfamide to 5 cycles of epirubicin and ifosfamide in patients with high-risk locally advanced adult STSs. Three cycles of preoperative epirubicin and ifosfamide was found to be noninferior to 5 cycles with regards to overall survival.³⁸

• What is this patient’s risk for recurrence?

The patient is at intermediate risk for recurrence. Numerous studies have demonstrated that tumor size, grade, and location are the most important factors to determine risk of recurrence, with larger size, higher grades, and deeper locations being associated with higher risk of recurrence. In an analysis of 1041 patients with STS of the extremities, high grade was the most important risk factor for distant metastases.³⁹ The highest risk of recurrence is within the first 2 years. Given that the patient’s initial tumor was located in the extremity, he is more likely to have a distant metastasis as his site of recurrence; individuals with retroperitoneal tumors and visceral tumors are more likely to recur locally.⁴⁰ For STSs of the extremity, distant metastases determine overall survival, whereas patients with retroperitoneal sarcomas can die from complications of local metastases.⁴¹ Once a patient develops distant metastases, the most important prognostic factor is the size of the tumor, with tumors larger than 10 cm having a relative risk of 1.5 (95% CI 1.0 to 2.0).³⁹

• What are the recommendations for surveillance?

Surveillance recommendations are based on the stage of the sarcoma. Stage I tumors are the least likely to recur either locally or distally. As a result, it is recommended that stage I tumors be followed with history and physical exam every 3 to 6 months for the first 2 to 3 years, and then annually after the first 2 to 3 years. Chest x-rays should be considered every 6 to 12 months.² For stage II–IV tumors, history and physical exam is recommended every 3 to 6 months for the first 2 to 3 years. Chest and distant metastases imaging should also be performed every 3 to 6 months during this time frame. For the next 2 years, history and physical exam and imaging are recommended every 6 months. After the first 4 to 5 years, annual follow-up is recommended.²

CASE CONTINUED

The patient does well for 1 year. With physical therapy, he regains most of the strength and coordination of the lower extremity. He is followed every 3 months with chest x-rays and a MRI of the thigh for the first year. On his fourth follow-up clinic visit, he describes increased dyspnea on exertion over the previous few weeks and is found to have multiple lung metastases in both lungs on chest x-ray. He undergoes further evaluation for metastases and is not found to have any other metastatic lesions. Bronchoscopy and biopsy of 1 of the lung nodules confirms recurrent dedifferentiated liposarcoma.

• Should this patient undergo metastectomy?

An analysis of 3149 patients with STS treated at Memorial Sloan-Kettering who developed lung metastases found that patients with pulmonary metastases have survival rates of 25%. The most important prognostic factor for survival was complete resection of all metastases.⁴² For stage IV disease, surgery is used only in certain instances. In instances where tumor is more localized or limited, removal of metastases or metastectomy can play a role in management.²

CASE CONTINUED

Because the patient’s metastases are limited to the lungs, he is referred for metastectomy. He undergoes wedge resection for definitive diagnosis but it is not possible to completely resect all of the metastases. He is thus referred to a medical oncologist to discuss his treatment options.

• What are treatment options for unresectable or metastatic disease?

Metastatic Disease

Unlike local and locally advanced disease, chemotherapy forms the backbone of treatment in stage IV disease. Doxorubicin and olaratumab or doxorubicin and ifosfamide in combination are considered first line in metastatic disease. Response rates for single-agent doxorubicin range from 16% to 27%, while phase 2 and phase 3 studies of doxorubicin and ifosfamide have found response rates ranging from 18% to 36%.⁴³ In addition, the effectiveness of doxorubicin and ifosfamide phase 2 and 3 trials varied. Edmonson et al found a tumor regression rate of 34% for doxorubicin and ifosfamide as compared to 20% for doxorubicin alone.⁴⁴ In comparison, Santoro et al found a response rate of 21.3% for doxorubicin alone and 25.2% for doxorubicin and ifosfamide.⁴⁵ Neither study found increased survival benefit for doxorubicin and ifosfamide when compared to doxorubicin alone. In a Cochrane review evaluating randomized trials that compared doxorubicin and combination chemotherapy regimens, response rates varied from 14% for doxorubicin in combination with streptomycin to 34% for doxorubicin and ifosfamide. Most trials did not show a significant benefit for combination therapies when compared to doxorubicin alone.⁴³ Mean survival with doxorubicin or doxorubicin and ifosfamide is 12 months. High rates of recurrence highlight the need for additional chemotherapy regimens.

The newest approved agent is olaratumab, a monoclonal antibody that binds platelet-derived growth factor receptor alpha and prevents receptor activation. A phase 1-b and phase 2 trial evaluated patients with locally advanced and metastatic STS and randomly assigned them to either olaratumab and doxorubicin or doxorubicin alone.⁴⁶ Progression-free survival for olaratumab/doxorubicin was 6.6 months (95% CI 4.1 to 8.3) compared to 4.1 months (95% CI 2.8 to 5.4) for doxorubicin alone. The objective response rate was 18.2% (95% CI 9.8 to 29.6) for olaratumab/doxorubicin compared to 7.5% (95% CI 2.5 to 6.6) for doxorubicin alone. Furthermore, the median overall survival for olaratumab plus doxorubicin was 26.5 months (95% CI 20.9 to 31.7) compared to 14.7 months for doxorubicin alone (95% CI 5.5 to 26.0). Impressively, this improved response was notable across histological types. Furthermore, patients who had previously been treated with more than 1 regimen and those who were treatment naïve had similar response rates.⁴⁶

• What are second-line treatment options?

Doxorubicin has been used in combination with several other agents including dacarbazine (DTIC) as well as DTIC and ifosfamide (MAID). Borden et al evaluated patients with metastatic STS and randomly assigned the patients to either doxorubicin or doxorubicin and DTIC. Combination therapy demonstrated better tumor response than doxorubicin alone: 30% complete or partial response for combination therapy and 18% for doxorubicin alone.⁴⁷ However, Omura et al

Several additional regimens have undergone evaluation in metastatic and recurrent STSs. Gemcitabine has been used both as a single agent and as part of combination therapy in many studies. Studies with gemcitabine in combination with either docetaxel or DTIC have been the most efficacious. In a phase 2 trial, patients with metastatic STS were randomly assigned to either gemcitabine alone or gemcitabine and docetaxel. Combination therapy had a higher response rate (16% versus 8%) and longer overall survival (17.9 months versus 11.5 months) than gemcitabine alone.⁵⁰ Furthermore, a phase 2 trial of gemcitabine and docetaxel in patients with unresectable leiomyosarcoma showed an overall response rate of 56%, with 3 complete and 15 partial responses among the 34 patients enrolled in the study.⁵¹

A phase 2 trial randomly assigned patients with unresectable or metastatic STS to either DTIC or combination gemcitabine and DTIC.⁵² Gemcitabine-DTIC had a superior progression-free survival at 3 months (56% [95% CI 43% to 69%]) as compared to DTIC alone (37% [95% CI 23.5% to 50%]). Furthermore, mean progression-free survival and overall survival were improved in the gemcitabine-DTIC group (4.2 months and 16.8 months) as compared to the DTIC group (2.0 months and 8.2 months).⁵² DTIC has a single-agent response rate of 16%, but has been shown to be particularly effective in the setting of leiomyosarcomas.⁴⁹

• Does response to treatment regimens differ by histologic subtype?

The majority of STS trials include many different histologic subtypes. Given the rarity of sarcomas as a whole, many trials have had difficulty recruiting adequate numbers of patients to have sufficient power to definitely determine if the treatment under investigation has clinical benefit. Furthermore, the patients recruited have been heterogeneous with regard to subtype. Many older studies hypothesized that the efficacy of chemotherapeutic agents vary based on histologic subtype; however, for most subtypes the number of individuals included in those trials was too low to evaluate efficacy based on subtype.

Some exceptions exist, however. For example, both gemcitabine-DTIC and gemcitabine-docetaxel have been found to be particularly effective in the treatment of leiomyosarcomas.^50,52 Additionally, a retrospective study found a 51% overall response rate for patients with myxoid liposarcomas treated with trabectedin.⁵³ Studies of patients with angiosarcoma treated with paclitaxel have demonstrated response rates of 43% and 53%.^54,55

• What are the newest approved and investigational agents?

A recently approved agent is trabectedin, a tris tetrahydroisoquinoline alkaloid isolated from ascidians that binds to the minor groove of DNA and causes disruptions in the cell cycle. Samuels et al reported data from a single-arm, open-label expanded access trial that evaluated patients with advanced metastatic sarcomas.⁵⁶ In this study, patients with liposarcomas and leiomyosarcomas had an objective response rate of 6.9% (95% CI 4.8 to 9.6) as compared to a rate of 5.9% (95% CI 4.4 to 7.8) for all assessable patients. Median survival was 11.9 months for all patients, with improved median survivals for liposarcoma and leiomyosarcomas of 16.2 months (95% CI 14.1 to 19.5) compared to 8.4 months (95% CI 7.1 to 10.7 months) for other subtypes.⁵⁶

Schöffski et al evaluated eribulin, a chemotherapeutic agent that affects microtubule dynamics, in a phase 2 trial of patients with progressive or high-grade STS with progression on previous chemotherapy. They found a median progression-free survival of 2.6 months (95% CI 1.7 to 6.2) for adipocytic sarcoma, 2.9 months (95% CI 2.4 to 4.6) for leiomyosarcoma, 2.6 months (95% CI 2.3 to 4.3) for synovial sarcoma, and 2.1 months (95% CI 1.4 to 2.9) for other sarcomas.⁵⁷

Van der Graaf and colleagues randomly assigned patients with metastatic nonadipocytic STS to pazopanib or placebo in a phase 3 trial. Pazopanib is a small-molecule endothelial growth factor inhibitor with activity against vascular endothelial growth factors 1, 2, and 3 as well as platelet-derived growth factors. Median progression-free survival was 4.6 months (95% CI 3.7 to 4.8) with pazopanib compared to 1.6 months (95% CI 0.9 to 1.8) with placebo.⁵⁸ Adipocytic sarcomas (liposarcomas) were excluded from the trial because phase 2 trials had found a lower rate of progression-free survival (26%) for them compared to other subtypes.

Toxicities were seen with each of the regimens studied and were common in the randomized trials, with higher rates of toxicities in the combination chemotherapy regimens. The most common toxicities are myelosuppression, nausea, and vomiting. In the doxorubicin trials, the most common toxicities were myelosuppression, nausea, and vomiting.⁴⁴

Ifosfamide both as an individual agent and in combination with doxorubicin has higher rates and higher grades of toxicity than doxorubicin alone. Myelosuppression is the most common toxicity associated with ifosfamide, and the most commonly affected cell line is leukocytes.⁴⁴ Combination doxorubicin and ifosfamide also had high rates of nausea and vomiting (95%) and alopecia (100%).³⁵

Neutropenia is the most common toxicity associated with gemcitabine and dacarbazine, while their most common nonhematologic toxicities are fatigue and nausea.^52,59 Trabectedin’s most common toxicities are nausea (29%), neutropenia (24%), and fatigue (23%). It has also been shown to cause increased alkaline phosphatase (20%) and alanine aminotransferase (19%) levels.⁵⁶ In a phase 2 study of eribulin, 50% of patients had neutropenia, and other toxicities included fatigue, alopecia, nausea, sensory neuropathy, and thrombocytopenia.⁵⁷ Pazopanib is generally well tolerated; the most common toxicities are fatigue (65%), diarrhea (58%), nausea (54%), and hypertension (41%).⁵⁸ Higher rates of neutropenia, mucositis, nausea, vomiting, diarrhea, and transfusion reactions were seen with olaratumab and doxorubicin compared to doxorubicin alone in phase 1b and 2 studies.⁴⁶

CASE CONCLUSION

Given his poor prognosis with unresectable metastatic undifferentiated liposarcoma, the patient considers a clinical trial prior to undergoing combined therapy with doxorubicin and ifosfamide. He tolerates therapy well with stable disease at 6 months.

CONCLUSION

STSs are a heterogeneous collection of rare tumors. Low-grade, localized tumors have the best prognosis, and patients who undergo complete resection have the best long-term survival. Due to the rarity of STSs, trials often have limited enrollment, and little progress has been made with regards to treatment and survival rates for metastatic and unresectable disease. All patients should be evaluated and treated at specialized sarcoma centers. This case highlights the need for continued research and clinical trials to improve overall survival of patients with sarcoma.

INTRODUCTION

Sickle cell disease is the most common inherited blood disorder in the world. It affects more than 100,000 individuals in the United States, and millions more worldwide.¹ Sickle cell disease is most commonly found in individuals of African heritage, but the disease also occurs in Hispanics and people of Middle Eastern and subcontinent Indian heritage.² The distribution of the sickle hemoglobin (hemoglobin S [HbS]) allele overlaps with the distribution of malaria; HbS carriers, or individuals with sickle cell trait, have protection against malaria,³ and are not considered to have sickle cell disease.

Sickle cell disease is a severe monogenic disorder marked by significant morbidity and mortality, affecting every organ in the body.⁴ The term sickle cell disease refers to all genotypes that cause sickling; the most common are the homozygous hemoglobin SS (HbSS) and compound heterozygotes hemoglobin SC (HbSC), hemoglobin S–β⁰-thalassemia (HbSβ⁰),and hemoglobin S–β⁺-thalassemia(HbSβ⁺), although HbS and several rarer hemoglobin variants such as HbSO(Arab) and HbSD(Punjab) can also cause sickle cell disease.The term sickle cell anemia refers exclusively to the most severe genotypes, HbSS and HbSβ⁰.⁵ Common sickling genotypes along with their relative clinical severity are shown in Table 1.^6–11

This article reviews the pathophysiology of sickle cell disease, common clinical complications, and available therapies. A complex case which illustrates diagnostic and management challenges is presented as well.

PATHOPHYSIOLOGY

HbS is the result of a substitution of valine for glutamic acid in the sixth amino acid of the β-globin chain.¹² The change from a hydrophilic to a hydrophobic amino acid causes the hemoglobin molecules to stack, or polymerize, when deoxygenated. This rigid rod of hemoglobin distorts the cell, producing the characteristic crescent or sickle shape that gives the disease its name.¹³ Polymerization of hemoglobin within the cell is promoted by dehydration, which increases the concentration of HbS.^13,14 Polymerization occurs when hemoglobin is in the deoxygenated state.¹³

The sickle red blood cell is abnormal; it is rigid and dense, and lacks the deformability needed to navigate the microvasculature.¹⁵ Blockages of blood flow result in painful vaso-occlusion that is the hallmark of the disease, and that also can cause damage to the spleen, kidneys, and liver.¹⁶ The sickle red cell is also fragile, with a lifespan of only 20 days compared to the 120-day lifespan of a normal red blood cell.¹³ Frequent hemolysis results in anemia and the release of free hemoglobin, which both scavenges nitric oxide and impairs the production of more nitric oxide, which is essential for vasodilatation.¹⁷ This contributes to vascular dysfunction and an increased risk for stroke.¹⁸ If untreated, the natural course of sickle cell anemia is mortality in early childhood in most cases.¹⁹ Common chronic and acute sickle cell disease–related complications and recommended therapies, based on 2014 National Institutes of Health guidelines, are shown in Table 2 and Table 3.²⁰

One of the most challenging aspects of sickle cell disease is its clinical variability. While in general, HbSS and HbSβ⁰ are the most severe genotypes, there are patients with HbSC and HbSb⁺ who have significant sickle-cell–related complications, and may have a more severe clinical course than a HbSS patient.²¹ A great deal of this clinical variability cannot be explained, but some can be attributed to endogenous fetal hemoglobin (HbF) levels.^22–24 The importance of HbF levels in sickle cell disease was first noted by a pediatrician in the 1940s.²⁵ She observed that sickle cell disease complications in children under the age of 1 were rare, and attributed it to the presence of HbF.²⁵ HbF levels decline more slowly in individuals with hemoglobinopathies, reaching their nadir after the age of 5 rather than within 6 months of birth in individuals without hemoglobinopathies.²⁶ HbF levels remain elevated lifelong in most sickle cell disease patients, especially those with the HbSS and HbSβ⁰ genotypes. Levels of HbF vary widely between individuals, from zero to 20% to 30%, with a median of 10%.^26–28 Individuals who produce more HbF have a milder course, in general.²⁴ An association between the 4 β-globin haplotypes and HbF levels has been reported in the past,^27,29 but more sophisticated next-generation sequencing has revealed causal variants in BCL11A and HBS1L-MYB that contribute approximately 50% of the observed variability in HbF levels.^30–33

Co-inheritance of α-thalassemia also modifies disease course; less available α-globin chains results in a lower hemoglobin concentration within the cell. Paradoxically, this results in a higher overall hemoglobin level, as there is a reduction in polymerization, and therefore sickling due to lower HbS concentrations in the cell. Patients therefore are less anemic, reducing the risk of stroke in childhood,^34,35 but blood viscosity may be higher, resulting in more frequent pain crises and increased risk³⁶ of avascular necrosis.^34,35,37 It is often helpful to think of sickle cell patients as falling into 1 of 2 groups: high hemolysis/low hemoglobin and high viscosity/high hemoglobin. Individuals with high rates of hemolysis are at greater risk for stroke, pulmonary hypertension, and acute chest syndrome (ACS). Higher rates of hemolysis result in higher levels of free hemoglobin, which scavenges nitric oxide. This leads to the vascular damage and dysfunction that contributes to the associated clinical complications. This phenotype is most commonly seen in HbSS and HbSβ⁰.³⁸ High hemoglobin/high viscosity phenotypes are most often found in HbSC patients and in sickle cell anemia with α-thalassemia coinheritance.^39–42

TREATMENT OPTIONS

In high-resource countries with newborn screening, the initiation of penicillin prophylaxis has dramatically altered the natural history of the disease, allowing the majority of patients to reach adulthood.⁴³ Penicillin prophylaxis is usually discontinued at age 5 years; however, individuals who have undergone surgical splenectomy or have had pneumococcal sepsis on penicillin prophylaxis may remain on penicillin to age 18 or beyond.²⁰

Another advance in sickle cell care is screening for stroke risk through transcranial Doppler ultrasound (TCD).^44–47 This screening tool has reduced the incidence of childhood stroke from 10% by age 11 to 1%. TCDs typically cannot be performed after the age of 16 due to changes in the skull. Individuals found to have abnormal (elevated) TCD velocities are placed on chronic transfusion therapy for primary stroke prevention. They may remain on monthly chronic transfusions, with the goal of suppressing the percentage of HbS to 30% to 50% indefinitely. A clinical trial (STOPII) designed to determine if pediatric sickle cell disease patients on chronic transfusion therapy for primary stroke prevention could be safely taken off transfusion therapy was discontinued early due to an excess of strokes and conversion to abnormal TCD velocities in the untransfused arm.⁴⁴ Individuals who have experienced an ischemic stroke have a 70% risk of another stroke, and must remain on chronic transfusion therapy indefinitely. Chronic transfusion reduces their stroke risk to 13%.

The only widely used pharmacologic therapy for sickle cell disease is hydroxyurea.^12,48–50 A significant portion of the benefit of hydroxyurea stems from its induction of HbF.⁵¹ HbF does not sickle, and it interrupts the polymerization of HbS in the cell, if present in high enough concentrations.⁵⁰ The level of HbF needed to achieve clinical improvement is not known, but in vitro assays suggest 20% HbF is needed to prevent sickling.^52,53 However, endogenous and hydroxyurea-induced HbF is not distributed evenly through the red cells, so sickling is possible regardless of the level of HbF induced.^54,55 Hydroxyurea likely has other disease-modifying effects as well, including reduction of white blood cell count and reticulocyte count and reduction of red cell adhesion to the endothelium.^56–58 Clinical criteria for initiation of hydroxyurea in adult sickle cell disease patients are shown in Table 4.²⁰ Hydroxyurea is given daily and is dosed to maximum tolerated dose for the individual by following the absolute neutrophil count (ANC). The goal ANC is between 2000 and 4000/µL. At times, absolute reticulocyte count (ARC) can be dose-limiting; goal ARC is greater than 70,000/µL.⁵⁹ Platelet counts may be reduced as well, especially in HbSC patients.^60,61

The only curative therapy for sickle cell disease is hematopoietic stem cell transplant.⁶² Transplant use is limited by availability of matched sibling donors,⁶² and even at experienced centers transplant carries a small risk for mortality, graft rejection, and graft-versus-host disease. Furthermore, consensus on disease complications for which transplant is recommended is also lacking.^63–65 Clinical trials of gene therapy for sickle cell disease and thalassemia are ongoing.⁶⁶

COMPLICATIONS AND DISEASE-SPECIFIC THERAPIES

CASE PRESENTATION

A 26-year-old African-American man who works as a school bus driver presents to an academic center’s emergency department complaining of pain in his left leg, similar to prior pain events. He is described as having sickle cell trait, although no hemoglobin profile is available in his chart. He describes the pain as dull and aching, 10/10 in intensity. A complete blood count (CBC) is obtained; it reveals a hemoglobin of 14.5 g/dL, white blood cell (WBC) count of 5600/µL, and platelet count of 344,000/µL. His CBC is also notable for a mean corpuscular volume (MCV) of 72 fL, a mean corpuscular hemoglobin concentration (MCHC) of 37 g/dL, and a red blood cell distribution width (RDW) of 12. Slide review of a peripheral blood smear shows 2+ target cells (Figure).

The patient is given 6 mg of morphine, which provides some relief of his pain, and is discharged with a prescription for hydrocodone bitartrate/acetaminophen 5/325 mg. The diagnosis given is musculoskeletal pain, and he is instructed to follow-up with a primary care physician. His past medical history is significant for 4 or 5 visits to the emergency department per year in the past 4 years. Prior to 4 years ago, he rarely required medical attention.

• What laboratory and clinical features might lead you to question the diagnosis of sickle cell trait in this patient?

The patient’s hemoglobin is within normal range, which is consistent with sickle cell trait; however, he is microcytic, with a normal RDW. It is possible to be mildly microcytic in the early stages of iron deficiency, prior to the development of anemia, but the RDW would typically be elevated, demonstrating the presence of newer, smaller cells produced under conditions of iron deficiency.⁶⁷ It is also possible that his microcytosis with a normal RDW could represent sickle cell trait with co-inheritance of β-thalassemia. Up to 30% of African Americans have β-thalassemia,² and 1 in 10 have sickle cell trait.⁶⁸ However, a high MCHC, indicating the presence of dense cells, and target cells noted on slide review are most consistent with HbSC.⁹ HbSC patients, especially males, can have hemoglobin levels in the normal range.⁴ The biggest inconsistency with the diagnosis of sickle cell trait is his history of frequent pain events. Individuals with sickle cell trait rarely present with pain crises, except under extreme conditions of dehydration or high altitude.⁶⁸ Sickle cell trait is generally regarded as a benign condition, although a study of U.S. military recruits found a 30-fold higher risk of sudden death during basic training in persons with sickle cell trait.⁶⁹ Additional sickle cell trait–related complications include hematuria, risk of splenic sequestration or infarct under extreme conditions and high altitude, and a rare and usually fatal renal malignancy, renal medullary carcinoma, which is vanishingly rare in individuals without sickle cell trait.^70,71 Although the patient reported having sickle cell trait, this diagnosis should have been verified with a hemoglobin panel, given his atypical presentation.²⁰

In sickle cell disease, vaso-occlusive pain events can be common, often beginning in early childhood.¹⁷ This disease complication accounts for 95% of all adult sickle cell disease hospitalizations.⁷² There is a great deal of variability in pain symptoms between individuals, and within individuals at various times in their lives:⁷³ 30% have no pain events, 50% have occasional events, and 20% have monthly or more frequent events that require hospitalization.⁷⁴ The frequency and severity of pain events are modulated by HbF levels, β-thalassemia status, genotypes, therapies like hydroxyurea, or in rare cases, chronic transfusion therapy.²³ Personal factors, such as psychosocial stressors, also contribute to the frequency of pain events.⁷⁵ Pain event triggers include exposure to cold water, windy or cold weather, temperature changes, and extreme temperatures.^76–79 Patient age also contributes to pain event frequency. Many patients see an increase in pain event frequency in their late 20s, and a marked decrease in their 40s.^23,73 More than 3 pain events per year is associated with reduced life expectancy.²³

Acute management of pain episodes involves nonsteroidal anti-inflammatory drugs, oral opioids, and when hospitalization is required, intravenous opioids, often delivered via patient-controlled analgesia (PCA) pumps.⁷⁹ As sickle cell disease patients become teenagers and young adults, some experience an increased frequency of pain episodes, with fewer pain-free days, or a failure to return to baseline before the next pain crisis occurs.^80,81 This is characteristic of emerging chronic pain.⁸² Chronic pain is a significant problem in adult patients with sickle cell disease, with up to 85% reporting pain on most days.^72,80 The development of chronic pain may be reduced by early and aggressive treatment of acute pain events, as well as use of hydroxyurea to reduce the number of pain events. Many adult sickle cell patients with chronic pain are treated with daily opioids.²⁰ Given the significant side effects of chronic opioid use—sedation, respiratory depression, itching, nausea, and impairment of function and quality of life—non-opioid therapies are under investigation.⁸³ Many chronic pain patients have symptoms of neuropathic pain, and may benefit from neuropathic agents like gabapentin, both to reduce opioid use and to more effectively treat chronic neuropathic pain, which is known to respond poorly to opioids.^84–86

• Is the patient’s peripheral blood smear consistent with a diagnosis of sickle cell trait?

Several target cells are visible, which is not typical of sickle cell trait, but may be seen in HbSC or thalassemia. The finding of an intracellular crystal is pathognomonic for HbSC or HbCC. HbC polymerizes in high oxygen conditions, opposite of HbS, which polymerizes in low oxygen conditions.⁹

CASE CONTINUED

The patient’s family history is significant for a sister who died at age 3 from sickle cell–related complications, and a sister with sickle cell trait who had a cholecystectomy for gallstones at age 22. His father died at age 38 due to unknown causes. The sickle cell trait status of his parents is unknown. His mother is alive, and has hypertension.

• Is the medical history of this patient’s family members consistent with sickle cell trait?

It is unlikely that sickle cell trait would result in early death in childhood, or in gallstones at age 22. Gallstones in early adulthood is a common presentation for HbSC patients not diagnosed by newborn screening.⁸⁷ Any hemolytic condition can lead to the formation of hemoglobin-containing pigmented gallstones, biliary sludge, and obstruction of the gallbladder. In the presence of right-sided abdominal pain, a serum bilirubin level of more than 4 mg/dL should lead to measurement of direct bilirubin; if greater than 10% of total, imaging of the gallbladder should be obtained. In sickle cell disease, 30% of patients will have gallstones by 18 years of age. The low hemolysis/high viscosity phenotype patients are typically older at diagnosis. Co-inheritance of Gilbert syndrome and sickle cell disease is not uncommon, and can result in formation of gallstones at a young age; Gilbert syndrome alone typically results in gallstones in mid-life.⁸⁸

Two months later, the patient presents again to the emergency department with the same complaint of leg pain, as well as abdominal pain. His hemoglobin is 12.5 g/dL, and his platelet count is 134,000/µL. His pain is not improved with 3 doses of morphine 6 mg intravenously, and he is admitted to the medicine service. A hemoglobin profile is obtained, revealing 52% HbS, 45% HbC, and 1.5% HbF, consistent with HbSC. In sickle cell trait, the hemoglobin profile is 60% HbA and 40% HbS (available α-globin prefers to pair with a normal β-globin, so the ratio of HbA to HbS is 60:40, not 50:50).

On the second hospital day, the patient’s hemoglobin drops to 7.2 g/dL and his platelet count decreases to 44,000/µL. His abdomen is distended and diffusely tender. The internist transfuses him with 2 units of packed red blood cells (PRBC), after which his hemoglobin increases to 11 g/dL, while his platelet count increases to 112,000/µL. Following the transfusion, his abdominal pain resolves, as does his anemia and thrombocytopenia.

• What caused this patient’s anemia and thrombocytopenia?

High on the differential diagnosis is a splenic sequestration. Acute splenic sequestration occurs when red cells are trapped in the splenic sinuses. Massive splenic enlargement may occur over several hours.^89,90 Unrecognized splenic sequestration has a high mortality rate from severe anemia and splenic rupture.⁹⁰ Splenic sequestration must be ruled out in a sickle cell patient with abdominal pain accompanied by dropping platelet and red cell counts, especially in milder subtypes that often have splenic function preserved into adolescence and adulthood. Sickle cell anemia patients usually become functionally asplenic in early childhood.^89,91,92 The rise in hemoglobin, more than would be expected from 2 units of PRBC, plus the improvement in platelet count without a platelet transfusion observed in the case patient strongly supports the diagnosis of splenic sequestration.

Splenic sequestration can occur in any sickle cell patient whose spleen has not fibrosed. Splenic sequestration in adulthood is not uncommon in HbSC patients, who often have preserved splenic function into adulthood.^93–95

Clinical signs of splenic sequestration include a rapid drop in hemoglobin, rise in reticulocyte count, a tender, enlarged spleen, and, in severe cases, hypovolemia.^89,93 It is treated with prompt blood transfusion, but care must be taken not to overtransfuse the patient, as the spleen can trap several grams of hemoglobin, which may be released upon transfusion, potentially causing life-threatening hyperviscosity.⁸⁹ Hemoglobin levels must be checked following transfusion in suspected splenic sequestration, and “mini transfusions” of 5 mL/kg are recommended in sickle cell disease patients who are hemodynamically stable.²⁰

Hepatic sequestration may also occur, but it is much less common than splenic sequestration.⁹⁶ Other conditions on the differential diagnosis include thrombotic thrombocytopenic purpura, which would be unlikely to respond to a transfusion. ACS can cause a drop in hemoglobin, and is treated with simple or exchange transfusions.⁹⁷ ACS is less likely without respiratory symptoms or oxygen requirement, and usually is not associated with thrombocytopenia. Sepsis may also cause anemia and thrombocytopenia, but again would not likely respond to a simple transfusion. The patient’s response to transfusion is consistent with a sequestering event, not a destructive event as in the case of sepsis.

CASE CONTINUED

Imaging reveals a grossly enlarged spleen, which is having a mass effect on the left kidney. The patient is started on hydroxyurea therapy at 500 mg 3 times daily. Discharge instructions include following up with his primary care physician, continuing hydroxyurea therapy, and receiving yearly dilated eye exams to evaluate for proliferative sickle retinopathy.

• Are these discharge instructions complete?

Splenic sequestration has a 50% recurrence rate.⁹⁸ In very young children, watchful waiting or chronic transfusion may be implemented to preserve the immunologic function of the spleen and reduce the risk of sepsis.⁸⁹ Splenectomy after a single episode of sequestration in adults is a matter of debate, with experts advising both watchful waiting⁹⁹ and splenectomy after recovery from the first sequestering event.¹⁰⁰ The patient should have been informed of the risk for recurrence, and the signs and symptoms of splenic sequestration as well as the need for emergency medical attention should have been discussed. Splenic sequestration may be milder in adults than in children, but fatal sequestrations have been reported.^95,101–103

Proliferative sickle cell retinopathy is a high viscosity/high hemoglobin complication that may occur more frequently in HbSC than HbSS, with an incidence of 33% in HbSC.^42,104 Spontaneous regression of retinopathy occurs in approximately 32% of eyes, and laser or scatter photocoagulation is an effective intervention.¹⁰⁵

• Would the patient need to be transfused prior to splenectomy?

Preoperative transfusion therapy is standard of care for HbSS patients undergoing general anesthesia. The TRAP study found that simple “top off” transfusion to a hemoglobin of 10 g/dL was as effective at preventing postoperative sickle cell–related complications as exchange transfusion to HbS of 30% or less, and had fewer transfusion-related complications like alloimmunization.¹⁰⁶ There is little data regarding preoperative transfusions in HbSC disease. A retrospective study suggests that HbSC patients undergoing abdominal surgeries should be transfused.¹⁰⁷ The higher hemoglobin level of the typical HbSC patient necessitates exchange transfusion to avoid hyperviscosity.

• Is hydroxyurea therapy indicated in this patient?

• Has it been dosed appropriately?

If the patient had the HbSS subtype, hydroxyurea would be clearly indicated, given his frequent pain events.²⁰ HbSC patients may be placed on hydroxyurea on a case-by-case basis, but evidence for its efficacy in this sickle cell subtype is lacking.¹⁰⁸ Large clinical trials like the Multi-Center Study of Hydroxyurea (MSH) that established the safety and efficacy of hydroxyurea in sickle cell anemia excluded HbSC and HbSβ⁺ patients.¹⁰⁹ These mild to moderate subtypes produce less HbF at baseline, and typically have a minimal to modest rise in HbF on hydroxyurea.¹¹⁰ In sickle cell anemia, hydroxyurea is titrated to maximum tolerated dose, defined as an ANC of 2000 to 4000/µL and an ARC of 70,000/µL or higher.⁵³ Because of their lower levels of chronic inflammation and lower reticulocyte counts due to higher hemoglobin levels, many HbSC and HbSβ⁺ patients have values in that range before initiating hydroxyurea therapy.⁹ Cytopenias, particularly of platelets in HbSC, occur at low doses of hydroxyurea.¹¹¹

Of note, although the half-life of hydroxyurea would suggest that 3 times daily dosing is indicated, daily dosing has been found to have equal response and is preferred. Another concern is the monitoring of this myelosuppressive medication. This patient has repeatedly failed to obtain a primary care physician or a hematologist, and hydroxyurea requires laboratory monitoring at least every 2 months, especially in a HbSC patient with a very large spleen who is at significant risk for thrombocytopenia and neutropenia.⁹

CASE CONTINUED

A week after discharge from his admission for abdominal pain diagnosed as splenic sequestration, the patient presents again to the emergency department with abdominal pain which he reports is his typical sickle cell pain. Hemoglobin is 13.8 g/dL, platelet count is 388,000/µL, and alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels are both 10 times their prior value. Creatinine is 1.2 mg/dL (0.75 mg/dL on his prior admission), and total bilirubin is 3 mg/dL, with 0.3 mg/dL direct bilirubin. He undergoes an ultrasound exam of his gallbladder, which reveals sludge and a possible gallstone. There is no evidence of cholecystitis. General surgery performs a laparoscopic cholecystectomy.

• Was this cholecystectomy necessary?

In patients with sickle cell disease, symptomatic gallstones and gallbladder sludge should be observed; recurrent abdominal pain without a significant change in bilirubin may not be due to gallstones or sludge, and therefore may not be relieved by cholecystectomy.^112,113 In sickle cell disease, 40% of patients with gallbladder sludge do not develop gallstones.⁸⁷ The patient’s bilirubin level was at baseline, and there was no increase in the direct (conjugated) fraction. Watchful waiting would have been appropriate, with cholecystectomy being performed if he experienced recurrent symptoms associated with fatty foods accompanied by an elevation in direct bilirubin.

More concerning and deserving of investigation was his elevated liver enzymes. Patients with sickle cell disease may experience recurrent ischemia and reperfusion injuries in the liver, which is called right upper quadrant syndrome. On autopsy of 70 sickle cell patients, 91% had hepatomegaly and 34% had focal necrosis.¹¹⁴ AST is often elevated in sickle cell disease, as it is affected by hemolysis. In this patient, both AST and ALT are elevated, consistent with a hepatocellular disorder. His abdominal pain and ALT rise may be a sign of a hepatic crisis.¹¹⁵ Rapid resolution of ALT elevation in a matter of days suggests a vaso-occlusive, inflammatory event that is self- limiting. Prolonged AST elevation requires further investigation, with consideration of autoimmune hepatitis, viral hepatitis, or iron overload. Iron overload is unlikely in this patient given his lifetime history of only 1 transfusion. Hepatic iron overload typically occurs in sickle cell disease after a minimum of 10 transfusions.¹¹⁵

CASE CONTINUED

The patient is discharged on the day after the procedure, with instructions to continue his hydroxyurea.

• Should the patient resume hydroxyurea therapy?

Hydroxyurea is hepatically cleared and thus it should be held until his liver function tests normalize.¹⁰⁶

CASE CONTINUED

Two months later, the patient presents to the emergency department with abdominal pain that moves to his left leg. A CBC is obtained, showing a hemoglobin of 11.8 g/dL and a platelet count of 144,000/µL. He is given 2 doses of morphine 6 mg intravenously, and reports that his leg pain is now a 4/10. He is discharged home with a prescription for hydrocodone/acetaminophen.

• Is the emergency department evaluation sufficient?

This patient remains at high risk for splenic sequestration,⁹³ with a hemoglobin 2 g lower than it was 2 months ago and platelets less than half. This decline could be consistent with early splenic sequestration.²⁰ Additionally, he had elevated liver function tests on a recent admission, as well as rising creatinine, without evidence of resolution. It is not appropriate to discharge him without checking a chemistry and liver panel, and abdominal imaging should be considered. The best plan would be to admit him for observation, given his risk for splenic sequestration, and consult surgery for an elective splenectomy if he has a second episode of splenic sequestration 2 months after the first.¹⁰⁰ His abdominal pain that migrates to his left leg could be due to his massive splenomegaly compressing his left kidney, as noted on imaging during his recent admission for splenic sequestration

CASE CONTINUED

An hour after discharge from the emergency department, EMS is called to his home for intractable pain. He is found lying on the floor, and reports excruciating left leg pain. He is brought to the closest hospital, a community hospital that he has not visited previously. There, he is admitted for hydration and pain control and placed on hydromorphone 2 mg every 4 hours as needed for pain. His hemoglobin is 10.8 g/dL, and platelets are 121,000/µL. A chemistry panel is remarkable for a creatinine level of 1.5 mg/dL and a potassium level of 3.2 mEq/L. Liver function tests are not obtained. After 3 doses of hydromorphone, he falls asleep. He is not in a monitored bed, and intravenous fluids, while ordered, are not started. At 6:30 AM the day after admission, he cannot be aroused on a routine vital sign check; he has an SpO₂ of 60%, a blood pressure of 80/60 mm Hg, and heart rate of 148 beats/min. A rapid response is called, and naloxone is administered along with oxygen by face mask and several fluid boluses. His systolic blood pressure increases to 100 mm Hg from a low of 70 mm Hg. His SpO₂ increases to 92%, and he is arousable and alert, although he reports 10/10 leg pain. His abdomen is noted to be distended and tender.

• What may have contributed to his clinical condition?

The patient is opioid tolerant and has received equivalent doses of opioids in the past without excess sedation. He may have liver dysfunction making him unable to metabolize opioids effectively. His hemoglobin and platelets continue to decline, raising concern for splenic sequestration versus sepsis. Failure to place him on a monitor allowed his hypoxia to continue for an unknown amount of time, placing him at high risk for developing ACS. Lack of intravenous hydration while he was too sedated to drink likely exacerbated his sickling.

At 9:20 AM, a CBC is obtained and reveals a hemoglobin of 4.8 g/dL and a platelet count of 44,000/µL. Two units of stat O negative blood are administered, and preparations are made to administer an exchange transfusion. A liver panel is obtained 3 hours later, which reveals an AST level of 1200 U/L and an ALT level of 1050 U/L. His bilirubin is 10 mg/dL, and his lactate dehydrogenase level is 1800 U/L. His urine is dark and is positive for bilirubin and ketones. He is transferred to the intensive care unit. A chest X-ray shows pulmonary congestion. Hematology/oncology is consulted.

He receives a 7-unit red blood cell exchange, which reduces his HbS to 11%. He continues to be hypotensive, and requires norepinephrine to support his blood pressure. Antibiotic therapy is started. His creatinine concentration rises to 2.3 mg/dL, potassium is 7.8 mEq/L, and bicarbonate is 12 mEq/L. He is placed on hemodialysis.

Computed tomography of the chest and abdomen reveals lower posterior lung infiltrates and a grossly enlarged spleen. He requires intubation. He is given a diagnosis of ACS in addition to kidney failure, liver failure, and “sickle crisis.” He continues to require daily to twice daily transfusions to maintain a hemoglobin of 7 to 9 g/dL, and his abdominal distension increases. As his condition worsens, surgery is consulted to discuss a liver transplant. He is deemed to not be a surgical candidate, and he passes away 6 days after entering the hospital. The immediate cause of death is listed as vaso-occlusive crisis, with ACS and sickle crisis listed as contributors.

• Are the causes of death accurate and complete?

If vaso-occlusive crisis is used to indicate a pain event, it is not an accurate cause of death. Pain is one of the most distressing complications of sickle cell disease, and frequent pain events are associated with early mortality,^4,80 but they are not in themselves fatal. ACS is the number one cause of death in sickle cell disease,⁴ and it likely contributed to this patient’s death. Sickle crisis is a vague term that should not be used in this context. Causes of death should include splenic sequestration and multisystem organ failure. Multisystem organ failure in sickle cell disease often responds to aggressive transfusion therapy, which this patient received.^116–118

CONCLUSION

Sickle cell disease is a complex chronic disease that impacts almost every organ system in the body. Clinicians may be inclined to attribute most pain in a patient with sickle cell disease to a simple vaso-occlusive crisis, treat them for this, and not investigate further. As the case presented here demonstrates, failure to identify the actual life-threatening process occurring in a patient with sickle cell disease presenting with pain can result in preventable early mortality. Clinicians must approach a sickle cell patient reporting pain in a thoughtful manner, and consider a complete differential diagnosis, including both sickle cell disease complications and those unrelated to sickle cell disease. Knowledge of the disease courses of the different sickle cell genotypes is essential, and must go beyond a superficial hierarchy of severity, but rather include an understanding of the complications each genotype is most prone to, and at what ages. Complete laboratory assessment, including a comprehensive metabolic panel, should be performed on all admitted patients, not just a complete blood count. Treating pain with high-dose opioids, while appropriate in an uncomplicated pain crisis, can lead to ACS or even respiratory failure in a patient with uninvestigated liver and kidney dysfunction. The most important lesson to remember is that even the sickle cell disease patient who has been given the unfortunate and pejorative label of “frequent flyer” by some providers has the potential for rapid deterioration into multisystem organ failure and death.

INTRODUCTION

Sickle cell disease is the most common inherited blood disorder in the world. It affects more than 100,000 individuals in the United States, and millions more worldwide.¹ Sickle cell disease is most commonly found in individuals of African heritage, but the disease also occurs in Hispanics and people of Middle Eastern and subcontinent Indian heritage.² The distribution of the sickle hemoglobin (hemoglobin S [HbS]) allele overlaps with the distribution of malaria; HbS carriers, or individuals with sickle cell trait, have protection against malaria,³ and are not considered to have sickle cell disease.

Sickle cell disease is a severe monogenic disorder marked by significant morbidity and mortality, affecting every organ in the body.⁴ The term sickle cell disease refers to all genotypes that cause sickling; the most common are the homozygous hemoglobin SS (HbSS) and compound heterozygotes hemoglobin SC (HbSC), hemoglobin S–β⁰-thalassemia (HbSβ⁰),and hemoglobin S–β⁺-thalassemia(HbSβ⁺), although HbS and several rarer hemoglobin variants such as HbSO(Arab) and HbSD(Punjab) can also cause sickle cell disease.The term sickle cell anemia refers exclusively to the most severe genotypes, HbSS and HbSβ⁰.⁵ Common sickling genotypes along with their relative clinical severity are shown in Table 1.^6–11

This article reviews the pathophysiology of sickle cell disease, common clinical complications, and available therapies. A complex case which illustrates diagnostic and management challenges is presented as well.

PATHOPHYSIOLOGY

HbS is the result of a substitution of valine for glutamic acid in the sixth amino acid of the β-globin chain.¹² The change from a hydrophilic to a hydrophobic amino acid causes the hemoglobin molecules to stack, or polymerize, when deoxygenated. This rigid rod of hemoglobin distorts the cell, producing the characteristic crescent or sickle shape that gives the disease its name.¹³ Polymerization of hemoglobin within the cell is promoted by dehydration, which increases the concentration of HbS.^13,14 Polymerization occurs when hemoglobin is in the deoxygenated state.¹³

The sickle red blood cell is abnormal; it is rigid and dense, and lacks the deformability needed to navigate the microvasculature.¹⁵ Blockages of blood flow result in painful vaso-occlusion that is the hallmark of the disease, and that also can cause damage to the spleen, kidneys, and liver.¹⁶ The sickle red cell is also fragile, with a lifespan of only 20 days compared to the 120-day lifespan of a normal red blood cell.¹³ Frequent hemolysis results in anemia and the release of free hemoglobin, which both scavenges nitric oxide and impairs the production of more nitric oxide, which is essential for vasodilatation.¹⁷ This contributes to vascular dysfunction and an increased risk for stroke.¹⁸ If untreated, the natural course of sickle cell anemia is mortality in early childhood in most cases.¹⁹ Common chronic and acute sickle cell disease–related complications and recommended therapies, based on 2014 National Institutes of Health guidelines, are shown in Table 2 and Table 3.²⁰

One of the most challenging aspects of sickle cell disease is its clinical variability. While in general, HbSS and HbSβ⁰ are the most severe genotypes, there are patients with HbSC and HbSb⁺ who have significant sickle-cell–related complications, and may have a more severe clinical course than a HbSS patient.²¹ A great deal of this clinical variability cannot be explained, but some can be attributed to endogenous fetal hemoglobin (HbF) levels.^22–24 The importance of HbF levels in sickle cell disease was first noted by a pediatrician in the 1940s.²⁵ She observed that sickle cell disease complications in children under the age of 1 were rare, and attributed it to the presence of HbF.²⁵ HbF levels decline more slowly in individuals with hemoglobinopathies, reaching their nadir after the age of 5 rather than within 6 months of birth in individuals without hemoglobinopathies.²⁶ HbF levels remain elevated lifelong in most sickle cell disease patients, especially those with the HbSS and HbSβ⁰ genotypes. Levels of HbF vary widely between individuals, from zero to 20% to 30%, with a median of 10%.^26–28 Individuals who produce more HbF have a milder course, in general.²⁴ An association between the 4 β-globin haplotypes and HbF levels has been reported in the past,^27,29 but more sophisticated next-generation sequencing has revealed causal variants in BCL11A and HBS1L-MYB that contribute approximately 50% of the observed variability in HbF levels.^30–33

Co-inheritance of α-thalassemia also modifies disease course; less available α-globin chains results in a lower hemoglobin concentration within the cell. Paradoxically, this results in a higher overall hemoglobin level, as there is a reduction in polymerization, and therefore sickling due to lower HbS concentrations in the cell. Patients therefore are less anemic, reducing the risk of stroke in childhood,^34,35 but blood viscosity may be higher, resulting in more frequent pain crises and increased risk³⁶ of avascular necrosis.^34,35,37 It is often helpful to think of sickle cell patients as falling into 1 of 2 groups: high hemolysis/low hemoglobin and high viscosity/high hemoglobin. Individuals with high rates of hemolysis are at greater risk for stroke, pulmonary hypertension, and acute chest syndrome (ACS). Higher rates of hemolysis result in higher levels of free hemoglobin, which scavenges nitric oxide. This leads to the vascular damage and dysfunction that contributes to the associated clinical complications. This phenotype is most commonly seen in HbSS and HbSβ⁰.³⁸ High hemoglobin/high viscosity phenotypes are most often found in HbSC patients and in sickle cell anemia with α-thalassemia coinheritance.^39–42

TREATMENT OPTIONS

In high-resource countries with newborn screening, the initiation of penicillin prophylaxis has dramatically altered the natural history of the disease, allowing the majority of patients to reach adulthood.⁴³ Penicillin prophylaxis is usually discontinued at age 5 years; however, individuals who have undergone surgical splenectomy or have had pneumococcal sepsis on penicillin prophylaxis may remain on penicillin to age 18 or beyond.²⁰

Another advance in sickle cell care is screening for stroke risk through transcranial Doppler ultrasound (TCD).^44–47 This screening tool has reduced the incidence of childhood stroke from 10% by age 11 to 1%. TCDs typically cannot be performed after the age of 16 due to changes in the skull. Individuals found to have abnormal (elevated) TCD velocities are placed on chronic transfusion therapy for primary stroke prevention. They may remain on monthly chronic transfusions, with the goal of suppressing the percentage of HbS to 30% to 50% indefinitely. A clinical trial (STOPII) designed to determine if pediatric sickle cell disease patients on chronic transfusion therapy for primary stroke prevention could be safely taken off transfusion therapy was discontinued early due to an excess of strokes and conversion to abnormal TCD velocities in the untransfused arm.⁴⁴ Individuals who have experienced an ischemic stroke have a 70% risk of another stroke, and must remain on chronic transfusion therapy indefinitely. Chronic transfusion reduces their stroke risk to 13%.

The only widely used pharmacologic therapy for sickle cell disease is hydroxyurea.^12,48–50 A significant portion of the benefit of hydroxyurea stems from its induction of HbF.⁵¹ HbF does not sickle, and it interrupts the polymerization of HbS in the cell, if present in high enough concentrations.⁵⁰ The level of HbF needed to achieve clinical improvement is not known, but in vitro assays suggest 20% HbF is needed to prevent sickling.^52,53 However, endogenous and hydroxyurea-induced HbF is not distributed evenly through the red cells, so sickling is possible regardless of the level of HbF induced.^54,55 Hydroxyurea likely has other disease-modifying effects as well, including reduction of white blood cell count and reticulocyte count and reduction of red cell adhesion to the endothelium.^56–58 Clinical criteria for initiation of hydroxyurea in adult sickle cell disease patients are shown in Table 4.²⁰ Hydroxyurea is given daily and is dosed to maximum tolerated dose for the individual by following the absolute neutrophil count (ANC). The goal ANC is between 2000 and 4000/µL. At times, absolute reticulocyte count (ARC) can be dose-limiting; goal ARC is greater than 70,000/µL.⁵⁹ Platelet counts may be reduced as well, especially in HbSC patients.^60,61

The only curative therapy for sickle cell disease is hematopoietic stem cell transplant.⁶² Transplant use is limited by availability of matched sibling donors,⁶² and even at experienced centers transplant carries a small risk for mortality, graft rejection, and graft-versus-host disease. Furthermore, consensus on disease complications for which transplant is recommended is also lacking.^63–65 Clinical trials of gene therapy for sickle cell disease and thalassemia are ongoing.⁶⁶

COMPLICATIONS AND DISEASE-SPECIFIC THERAPIES

CASE PRESENTATION

A 26-year-old African-American man who works as a school bus driver presents to an academic center’s emergency department complaining of pain in his left leg, similar to prior pain events. He is described as having sickle cell trait, although no hemoglobin profile is available in his chart. He describes the pain as dull and aching, 10/10 in intensity. A complete blood count (CBC) is obtained; it reveals a hemoglobin of 14.5 g/dL, white blood cell (WBC) count of 5600/µL, and platelet count of 344,000/µL. His CBC is also notable for a mean corpuscular volume (MCV) of 72 fL, a mean corpuscular hemoglobin concentration (MCHC) of 37 g/dL, and a red blood cell distribution width (RDW) of 12. Slide review of a peripheral blood smear shows 2+ target cells (Figure).

The patient is given 6 mg of morphine, which provides some relief of his pain, and is discharged with a prescription for hydrocodone bitartrate/acetaminophen 5/325 mg. The diagnosis given is musculoskeletal pain, and he is instructed to follow-up with a primary care physician. His past medical history is significant for 4 or 5 visits to the emergency department per year in the past 4 years. Prior to 4 years ago, he rarely required medical attention.

• What laboratory and clinical features might lead you to question the diagnosis of sickle cell trait in this patient?

The patient’s hemoglobin is within normal range, which is consistent with sickle cell trait; however, he is microcytic, with a normal RDW. It is possible to be mildly microcytic in the early stages of iron deficiency, prior to the development of anemia, but the RDW would typically be elevated, demonstrating the presence of newer, smaller cells produced under conditions of iron deficiency.⁶⁷ It is also possible that his microcytosis with a normal RDW could represent sickle cell trait with co-inheritance of β-thalassemia. Up to 30% of African Americans have β-thalassemia,² and 1 in 10 have sickle cell trait.⁶⁸ However, a high MCHC, indicating the presence of dense cells, and target cells noted on slide review are most consistent with HbSC.⁹ HbSC patients, especially males, can have hemoglobin levels in the normal range.⁴ The biggest inconsistency with the diagnosis of sickle cell trait is his history of frequent pain events. Individuals with sickle cell trait rarely present with pain crises, except under extreme conditions of dehydration or high altitude.⁶⁸ Sickle cell trait is generally regarded as a benign condition, although a study of U.S. military recruits found a 30-fold higher risk of sudden death during basic training in persons with sickle cell trait.⁶⁹ Additional sickle cell trait–related complications include hematuria, risk of splenic sequestration or infarct under extreme conditions and high altitude, and a rare and usually fatal renal malignancy, renal medullary carcinoma, which is vanishingly rare in individuals without sickle cell trait.^70,71 Although the patient reported having sickle cell trait, this diagnosis should have been verified with a hemoglobin panel, given his atypical presentation.²⁰

In sickle cell disease, vaso-occlusive pain events can be common, often beginning in early childhood.¹⁷ This disease complication accounts for 95% of all adult sickle cell disease hospitalizations.⁷² There is a great deal of variability in pain symptoms between individuals, and within individuals at various times in their lives:⁷³ 30% have no pain events, 50% have occasional events, and 20% have monthly or more frequent events that require hospitalization.⁷⁴ The frequency and severity of pain events are modulated by HbF levels, β-thalassemia status, genotypes, therapies like hydroxyurea, or in rare cases, chronic transfusion therapy.²³ Personal factors, such as psychosocial stressors, also contribute to the frequency of pain events.⁷⁵ Pain event triggers include exposure to cold water, windy or cold weather, temperature changes, and extreme temperatures.^76–79 Patient age also contributes to pain event frequency. Many patients see an increase in pain event frequency in their late 20s, and a marked decrease in their 40s.^23,73 More than 3 pain events per year is associated with reduced life expectancy.²³

Acute management of pain episodes involves nonsteroidal anti-inflammatory drugs, oral opioids, and when hospitalization is required, intravenous opioids, often delivered via patient-controlled analgesia (PCA) pumps.⁷⁹ As sickle cell disease patients become teenagers and young adults, some experience an increased frequency of pain episodes, with fewer pain-free days, or a failure to return to baseline before the next pain crisis occurs.^80,81 This is characteristic of emerging chronic pain.⁸² Chronic pain is a significant problem in adult patients with sickle cell disease, with up to 85% reporting pain on most days.^72,80 The development of chronic pain may be reduced by early and aggressive treatment of acute pain events, as well as use of hydroxyurea to reduce the number of pain events. Many adult sickle cell patients with chronic pain are treated with daily opioids.²⁰ Given the significant side effects of chronic opioid use—sedation, respiratory depression, itching, nausea, and impairment of function and quality of life—non-opioid therapies are under investigation.⁸³ Many chronic pain patients have symptoms of neuropathic pain, and may benefit from neuropathic agents like gabapentin, both to reduce opioid use and to more effectively treat chronic neuropathic pain, which is known to respond poorly to opioids.^84–86

• Is the patient’s peripheral blood smear consistent with a diagnosis of sickle cell trait?

Several target cells are visible, which is not typical of sickle cell trait, but may be seen in HbSC or thalassemia. The finding of an intracellular crystal is pathognomonic for HbSC or HbCC. HbC polymerizes in high oxygen conditions, opposite of HbS, which polymerizes in low oxygen conditions.⁹

CASE CONTINUED

The patient’s family history is significant for a sister who died at age 3 from sickle cell–related complications, and a sister with sickle cell trait who had a cholecystectomy for gallstones at age 22. His father died at age 38 due to unknown causes. The sickle cell trait status of his parents is unknown. His mother is alive, and has hypertension.

• Is the medical history of this patient’s family members consistent with sickle cell trait?

It is unlikely that sickle cell trait would result in early death in childhood, or in gallstones at age 22. Gallstones in early adulthood is a common presentation for HbSC patients not diagnosed by newborn screening.⁸⁷ Any hemolytic condition can lead to the formation of hemoglobin-containing pigmented gallstones, biliary sludge, and obstruction of the gallbladder. In the presence of right-sided abdominal pain, a serum bilirubin level of more than 4 mg/dL should lead to measurement of direct bilirubin; if greater than 10% of total, imaging of the gallbladder should be obtained. In sickle cell disease, 30% of patients will have gallstones by 18 years of age. The low hemolysis/high viscosity phenotype patients are typically older at diagnosis. Co-inheritance of Gilbert syndrome and sickle cell disease is not uncommon, and can result in formation of gallstones at a young age; Gilbert syndrome alone typically results in gallstones in mid-life.⁸⁸

Two months later, the patient presents again to the emergency department with the same complaint of leg pain, as well as abdominal pain. His hemoglobin is 12.5 g/dL, and his platelet count is 134,000/µL. His pain is not improved with 3 doses of morphine 6 mg intravenously, and he is admitted to the medicine service. A hemoglobin profile is obtained, revealing 52% HbS, 45% HbC, and 1.5% HbF, consistent with HbSC. In sickle cell trait, the hemoglobin profile is 60% HbA and 40% HbS (available α-globin prefers to pair with a normal β-globin, so the ratio of HbA to HbS is 60:40, not 50:50).

On the second hospital day, the patient’s hemoglobin drops to 7.2 g/dL and his platelet count decreases to 44,000/µL. His abdomen is distended and diffusely tender. The internist transfuses him with 2 units of packed red blood cells (PRBC), after which his hemoglobin increases to 11 g/dL, while his platelet count increases to 112,000/µL. Following the transfusion, his abdominal pain resolves, as does his anemia and thrombocytopenia.

• What caused this patient’s anemia and thrombocytopenia?

High on the differential diagnosis is a splenic sequestration. Acute splenic sequestration occurs when red cells are trapped in the splenic sinuses. Massive splenic enlargement may occur over several hours.^89,90 Unrecognized splenic sequestration has a high mortality rate from severe anemia and splenic rupture.⁹⁰ Splenic sequestration must be ruled out in a sickle cell patient with abdominal pain accompanied by dropping platelet and red cell counts, especially in milder subtypes that often have splenic function preserved into adolescence and adulthood. Sickle cell anemia patients usually become functionally asplenic in early childhood.^89,91,92 The rise in hemoglobin, more than would be expected from 2 units of PRBC, plus the improvement in platelet count without a platelet transfusion observed in the case patient strongly supports the diagnosis of splenic sequestration.

Splenic sequestration can occur in any sickle cell patient whose spleen has not fibrosed. Splenic sequestration in adulthood is not uncommon in HbSC patients, who often have preserved splenic function into adulthood.^93–95

Clinical signs of splenic sequestration include a rapid drop in hemoglobin, rise in reticulocyte count, a tender, enlarged spleen, and, in severe cases, hypovolemia.^89,93 It is treated with prompt blood transfusion, but care must be taken not to overtransfuse the patient, as the spleen can trap several grams of hemoglobin, which may be released upon transfusion, potentially causing life-threatening hyperviscosity.⁸⁹ Hemoglobin levels must be checked following transfusion in suspected splenic sequestration, and “mini transfusions” of 5 mL/kg are recommended in sickle cell disease patients who are hemodynamically stable.²⁰

Hepatic sequestration may also occur, but it is much less common than splenic sequestration.⁹⁶ Other conditions on the differential diagnosis include thrombotic thrombocytopenic purpura, which would be unlikely to respond to a transfusion. ACS can cause a drop in hemoglobin, and is treated with simple or exchange transfusions.⁹⁷ ACS is less likely without respiratory symptoms or oxygen requirement, and usually is not associated with thrombocytopenia. Sepsis may also cause anemia and thrombocytopenia, but again would not likely respond to a simple transfusion. The patient’s response to transfusion is consistent with a sequestering event, not a destructive event as in the case of sepsis.

CASE CONTINUED

Imaging reveals a grossly enlarged spleen, which is having a mass effect on the left kidney. The patient is started on hydroxyurea therapy at 500 mg 3 times daily. Discharge instructions include following up with his primary care physician, continuing hydroxyurea therapy, and receiving yearly dilated eye exams to evaluate for proliferative sickle retinopathy.

• Are these discharge instructions complete?

Splenic sequestration has a 50% recurrence rate.⁹⁸ In very young children, watchful waiting or chronic transfusion may be implemented to preserve the immunologic function of the spleen and reduce the risk of sepsis.⁸⁹ Splenectomy after a single episode of sequestration in adults is a matter of debate, with experts advising both watchful waiting⁹⁹ and splenectomy after recovery from the first sequestering event.¹⁰⁰ The patient should have been informed of the risk for recurrence, and the signs and symptoms of splenic sequestration as well as the need for emergency medical attention should have been discussed. Splenic sequestration may be milder in adults than in children, but fatal sequestrations have been reported.^95,101–103

Proliferative sickle cell retinopathy is a high viscosity/high hemoglobin complication that may occur more frequently in HbSC than HbSS, with an incidence of 33% in HbSC.^42,104 Spontaneous regression of retinopathy occurs in approximately 32% of eyes, and laser or scatter photocoagulation is an effective intervention.¹⁰⁵

• Would the patient need to be transfused prior to splenectomy?

Preoperative transfusion therapy is standard of care for HbSS patients undergoing general anesthesia. The TRAP study found that simple “top off” transfusion to a hemoglobin of 10 g/dL was as effective at preventing postoperative sickle cell–related complications as exchange transfusion to HbS of 30% or less, and had fewer transfusion-related complications like alloimmunization.¹⁰⁶ There is little data regarding preoperative transfusions in HbSC disease. A retrospective study suggests that HbSC patients undergoing abdominal surgeries should be transfused.¹⁰⁷ The higher hemoglobin level of the typical HbSC patient necessitates exchange transfusion to avoid hyperviscosity.

• Is hydroxyurea therapy indicated in this patient?

• Has it been dosed appropriately?

If the patient had the HbSS subtype, hydroxyurea would be clearly indicated, given his frequent pain events.²⁰ HbSC patients may be placed on hydroxyurea on a case-by-case basis, but evidence for its efficacy in this sickle cell subtype is lacking.¹⁰⁸ Large clinical trials like the Multi-Center Study of Hydroxyurea (MSH) that established the safety and efficacy of hydroxyurea in sickle cell anemia excluded HbSC and HbSβ⁺ patients.¹⁰⁹ These mild to moderate subtypes produce less HbF at baseline, and typically have a minimal to modest rise in HbF on hydroxyurea.¹¹⁰ In sickle cell anemia, hydroxyurea is titrated to maximum tolerated dose, defined as an ANC of 2000 to 4000/µL and an ARC of 70,000/µL or higher.⁵³ Because of their lower levels of chronic inflammation and lower reticulocyte counts due to higher hemoglobin levels, many HbSC and HbSβ⁺ patients have values in that range before initiating hydroxyurea therapy.⁹ Cytopenias, particularly of platelets in HbSC, occur at low doses of hydroxyurea.¹¹¹

Of note, although the half-life of hydroxyurea would suggest that 3 times daily dosing is indicated, daily dosing has been found to have equal response and is preferred. Another concern is the monitoring of this myelosuppressive medication. This patient has repeatedly failed to obtain a primary care physician or a hematologist, and hydroxyurea requires laboratory monitoring at least every 2 months, especially in a HbSC patient with a very large spleen who is at significant risk for thrombocytopenia and neutropenia.⁹

CASE CONTINUED

A week after discharge from his admission for abdominal pain diagnosed as splenic sequestration, the patient presents again to the emergency department with abdominal pain which he reports is his typical sickle cell pain. Hemoglobin is 13.8 g/dL, platelet count is 388,000/µL, and alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels are both 10 times their prior value. Creatinine is 1.2 mg/dL (0.75 mg/dL on his prior admission), and total bilirubin is 3 mg/dL, with 0.3 mg/dL direct bilirubin. He undergoes an ultrasound exam of his gallbladder, which reveals sludge and a possible gallstone. There is no evidence of cholecystitis. General surgery performs a laparoscopic cholecystectomy.

• Was this cholecystectomy necessary?

In patients with sickle cell disease, symptomatic gallstones and gallbladder sludge should be observed; recurrent abdominal pain without a significant change in bilirubin may not be due to gallstones or sludge, and therefore may not be relieved by cholecystectomy.^112,113 In sickle cell disease, 40% of patients with gallbladder sludge do not develop gallstones.⁸⁷ The patient’s bilirubin level was at baseline, and there was no increase in the direct (conjugated) fraction. Watchful waiting would have been appropriate, with cholecystectomy being performed if he experienced recurrent symptoms associated with fatty foods accompanied by an elevation in direct bilirubin.

More concerning and deserving of investigation was his elevated liver enzymes. Patients with sickle cell disease may experience recurrent ischemia and reperfusion injuries in the liver, which is called right upper quadrant syndrome. On autopsy of 70 sickle cell patients, 91% had hepatomegaly and 34% had focal necrosis.¹¹⁴ AST is often elevated in sickle cell disease, as it is affected by hemolysis. In this patient, both AST and ALT are elevated, consistent with a hepatocellular disorder. His abdominal pain and ALT rise may be a sign of a hepatic crisis.¹¹⁵ Rapid resolution of ALT elevation in a matter of days suggests a vaso-occlusive, inflammatory event that is self- limiting. Prolonged AST elevation requires further investigation, with consideration of autoimmune hepatitis, viral hepatitis, or iron overload. Iron overload is unlikely in this patient given his lifetime history of only 1 transfusion. Hepatic iron overload typically occurs in sickle cell disease after a minimum of 10 transfusions.¹¹⁵

CASE CONTINUED

The patient is discharged on the day after the procedure, with instructions to continue his hydroxyurea.

• Should the patient resume hydroxyurea therapy?

Hydroxyurea is hepatically cleared and thus it should be held until his liver function tests normalize.¹⁰⁶

CASE CONTINUED

Two months later, the patient presents to the emergency department with abdominal pain that moves to his left leg. A CBC is obtained, showing a hemoglobin of 11.8 g/dL and a platelet count of 144,000/µL. He is given 2 doses of morphine 6 mg intravenously, and reports that his leg pain is now a 4/10. He is discharged home with a prescription for hydrocodone/acetaminophen.

• Is the emergency department evaluation sufficient?

This patient remains at high risk for splenic sequestration,⁹³ with a hemoglobin 2 g lower than it was 2 months ago and platelets less than half. This decline could be consistent with early splenic sequestration.²⁰ Additionally, he had elevated liver function tests on a recent admission, as well as rising creatinine, without evidence of resolution. It is not appropriate to discharge him without checking a chemistry and liver panel, and abdominal imaging should be considered. The best plan would be to admit him for observation, given his risk for splenic sequestration, and consult surgery for an elective splenectomy if he has a second episode of splenic sequestration 2 months after the first.¹⁰⁰ His abdominal pain that migrates to his left leg could be due to his massive splenomegaly compressing his left kidney, as noted on imaging during his recent admission for splenic sequestration

CASE CONTINUED

An hour after discharge from the emergency department, EMS is called to his home for intractable pain. He is found lying on the floor, and reports excruciating left leg pain. He is brought to the closest hospital, a community hospital that he has not visited previously. There, he is admitted for hydration and pain control and placed on hydromorphone 2 mg every 4 hours as needed for pain. His hemoglobin is 10.8 g/dL, and platelets are 121,000/µL. A chemistry panel is remarkable for a creatinine level of 1.5 mg/dL and a potassium level of 3.2 mEq/L. Liver function tests are not obtained. After 3 doses of hydromorphone, he falls asleep. He is not in a monitored bed, and intravenous fluids, while ordered, are not started. At 6:30 AM the day after admission, he cannot be aroused on a routine vital sign check; he has an SpO₂ of 60%, a blood pressure of 80/60 mm Hg, and heart rate of 148 beats/min. A rapid response is called, and naloxone is administered along with oxygen by face mask and several fluid boluses. His systolic blood pressure increases to 100 mm Hg from a low of 70 mm Hg. His SpO₂ increases to 92%, and he is arousable and alert, although he reports 10/10 leg pain. His abdomen is noted to be distended and tender.

• What may have contributed to his clinical condition?

The patient is opioid tolerant and has received equivalent doses of opioids in the past without excess sedation. He may have liver dysfunction making him unable to metabolize opioids effectively. His hemoglobin and platelets continue to decline, raising concern for splenic sequestration versus sepsis. Failure to place him on a monitor allowed his hypoxia to continue for an unknown amount of time, placing him at high risk for developing ACS. Lack of intravenous hydration while he was too sedated to drink likely exacerbated his sickling.

At 9:20 AM, a CBC is obtained and reveals a hemoglobin of 4.8 g/dL and a platelet count of 44,000/µL. Two units of stat O negative blood are administered, and preparations are made to administer an exchange transfusion. A liver panel is obtained 3 hours later, which reveals an AST level of 1200 U/L and an ALT level of 1050 U/L. His bilirubin is 10 mg/dL, and his lactate dehydrogenase level is 1800 U/L. His urine is dark and is positive for bilirubin and ketones. He is transferred to the intensive care unit. A chest X-ray shows pulmonary congestion. Hematology/oncology is consulted.

He receives a 7-unit red blood cell exchange, which reduces his HbS to 11%. He continues to be hypotensive, and requires norepinephrine to support his blood pressure. Antibiotic therapy is started. His creatinine concentration rises to 2.3 mg/dL, potassium is 7.8 mEq/L, and bicarbonate is 12 mEq/L. He is placed on hemodialysis.

Computed tomography of the chest and abdomen reveals lower posterior lung infiltrates and a grossly enlarged spleen. He requires intubation. He is given a diagnosis of ACS in addition to kidney failure, liver failure, and “sickle crisis.” He continues to require daily to twice daily transfusions to maintain a hemoglobin of 7 to 9 g/dL, and his abdominal distension increases. As his condition worsens, surgery is consulted to discuss a liver transplant. He is deemed to not be a surgical candidate, and he passes away 6 days after entering the hospital. The immediate cause of death is listed as vaso-occlusive crisis, with ACS and sickle crisis listed as contributors.

• Are the causes of death accurate and complete?

If vaso-occlusive crisis is used to indicate a pain event, it is not an accurate cause of death. Pain is one of the most distressing complications of sickle cell disease, and frequent pain events are associated with early mortality,^4,80 but they are not in themselves fatal. ACS is the number one cause of death in sickle cell disease,⁴ and it likely contributed to this patient’s death. Sickle crisis is a vague term that should not be used in this context. Causes of death should include splenic sequestration and multisystem organ failure. Multisystem organ failure in sickle cell disease often responds to aggressive transfusion therapy, which this patient received.^116–118

CONCLUSION

Sickle cell disease is a complex chronic disease that impacts almost every organ system in the body. Clinicians may be inclined to attribute most pain in a patient with sickle cell disease to a simple vaso-occlusive crisis, treat them for this, and not investigate further. As the case presented here demonstrates, failure to identify the actual life-threatening process occurring in a patient with sickle cell disease presenting with pain can result in preventable early mortality. Clinicians must approach a sickle cell patient reporting pain in a thoughtful manner, and consider a complete differential diagnosis, including both sickle cell disease complications and those unrelated to sickle cell disease. Knowledge of the disease courses of the different sickle cell genotypes is essential, and must go beyond a superficial hierarchy of severity, but rather include an understanding of the complications each genotype is most prone to, and at what ages. Complete laboratory assessment, including a comprehensive metabolic panel, should be performed on all admitted patients, not just a complete blood count. Treating pain with high-dose opioids, while appropriate in an uncomplicated pain crisis, can lead to ACS or even respiratory failure in a patient with uninvestigated liver and kidney dysfunction. The most important lesson to remember is that even the sickle cell disease patient who has been given the unfortunate and pejorative label of “frequent flyer” by some providers has the potential for rapid deterioration into multisystem organ failure and death.

INTRODUCTION

Sickle cell disease is the most common inherited blood disorder in the world. It affects more than 100,000 individuals in the United States, and millions more worldwide.¹ Sickle cell disease is most commonly found in individuals of African heritage, but the disease also occurs in Hispanics and people of Middle Eastern and subcontinent Indian heritage.² The distribution of the sickle hemoglobin (hemoglobin S [HbS]) allele overlaps with the distribution of malaria; HbS carriers, or individuals with sickle cell trait, have protection against malaria,³ and are not considered to have sickle cell disease.

Sickle cell disease is a severe monogenic disorder marked by significant morbidity and mortality, affecting every organ in the body.⁴ The term sickle cell disease refers to all genotypes that cause sickling; the most common are the homozygous hemoglobin SS (HbSS) and compound heterozygotes hemoglobin SC (HbSC), hemoglobin S–β⁰-thalassemia (HbSβ⁰),and hemoglobin S–β⁺-thalassemia(HbSβ⁺), although HbS and several rarer hemoglobin variants such as HbSO(Arab) and HbSD(Punjab) can also cause sickle cell disease.The term sickle cell anemia refers exclusively to the most severe genotypes, HbSS and HbSβ⁰.⁵ Common sickling genotypes along with their relative clinical severity are shown in Table 1.^6–11

This article reviews the pathophysiology of sickle cell disease, common clinical complications, and available therapies. A complex case which illustrates diagnostic and management challenges is presented as well.

PATHOPHYSIOLOGY

HbS is the result of a substitution of valine for glutamic acid in the sixth amino acid of the β-globin chain.¹² The change from a hydrophilic to a hydrophobic amino acid causes the hemoglobin molecules to stack, or polymerize, when deoxygenated. This rigid rod of hemoglobin distorts the cell, producing the characteristic crescent or sickle shape that gives the disease its name.¹³ Polymerization of hemoglobin within the cell is promoted by dehydration, which increases the concentration of HbS.^13,14 Polymerization occurs when hemoglobin is in the deoxygenated state.¹³

The sickle red blood cell is abnormal; it is rigid and dense, and lacks the deformability needed to navigate the microvasculature.¹⁵ Blockages of blood flow result in painful vaso-occlusion that is the hallmark of the disease, and that also can cause damage to the spleen, kidneys, and liver.¹⁶ The sickle red cell is also fragile, with a lifespan of only 20 days compared to the 120-day lifespan of a normal red blood cell.¹³ Frequent hemolysis results in anemia and the release of free hemoglobin, which both scavenges nitric oxide and impairs the production of more nitric oxide, which is essential for vasodilatation.¹⁷ This contributes to vascular dysfunction and an increased risk for stroke.¹⁸ If untreated, the natural course of sickle cell anemia is mortality in early childhood in most cases.¹⁹ Common chronic and acute sickle cell disease–related complications and recommended therapies, based on 2014 National Institutes of Health guidelines, are shown in Table 2 and Table 3.²⁰

One of the most challenging aspects of sickle cell disease is its clinical variability. While in general, HbSS and HbSβ⁰ are the most severe genotypes, there are patients with HbSC and HbSb⁺ who have significant sickle-cell–related complications, and may have a more severe clinical course than a HbSS patient.²¹ A great deal of this clinical variability cannot be explained, but some can be attributed to endogenous fetal hemoglobin (HbF) levels.^22–24 The importance of HbF levels in sickle cell disease was first noted by a pediatrician in the 1940s.²⁵ She observed that sickle cell disease complications in children under the age of 1 were rare, and attributed it to the presence of HbF.²⁵ HbF levels decline more slowly in individuals with hemoglobinopathies, reaching their nadir after the age of 5 rather than within 6 months of birth in individuals without hemoglobinopathies.²⁶ HbF levels remain elevated lifelong in most sickle cell disease patients, especially those with the HbSS and HbSβ⁰ genotypes. Levels of HbF vary widely between individuals, from zero to 20% to 30%, with a median of 10%.^26–28 Individuals who produce more HbF have a milder course, in general.²⁴ An association between the 4 β-globin haplotypes and HbF levels has been reported in the past,^27,29 but more sophisticated next-generation sequencing has revealed causal variants in BCL11A and HBS1L-MYB that contribute approximately 50% of the observed variability in HbF levels.^30–33

Co-inheritance of α-thalassemia also modifies disease course; less available α-globin chains results in a lower hemoglobin concentration within the cell. Paradoxically, this results in a higher overall hemoglobin level, as there is a reduction in polymerization, and therefore sickling due to lower HbS concentrations in the cell. Patients therefore are less anemic, reducing the risk of stroke in childhood,^34,35 but blood viscosity may be higher, resulting in more frequent pain crises and increased risk³⁶ of avascular necrosis.^34,35,37 It is often helpful to think of sickle cell patients as falling into 1 of 2 groups: high hemolysis/low hemoglobin and high viscosity/high hemoglobin. Individuals with high rates of hemolysis are at greater risk for stroke, pulmonary hypertension, and acute chest syndrome (ACS). Higher rates of hemolysis result in higher levels of free hemoglobin, which scavenges nitric oxide. This leads to the vascular damage and dysfunction that contributes to the associated clinical complications. This phenotype is most commonly seen in HbSS and HbSβ⁰.³⁸ High hemoglobin/high viscosity phenotypes are most often found in HbSC patients and in sickle cell anemia with α-thalassemia coinheritance.^39–42

TREATMENT OPTIONS

In high-resource countries with newborn screening, the initiation of penicillin prophylaxis has dramatically altered the natural history of the disease, allowing the majority of patients to reach adulthood.⁴³ Penicillin prophylaxis is usually discontinued at age 5 years; however, individuals who have undergone surgical splenectomy or have had pneumococcal sepsis on penicillin prophylaxis may remain on penicillin to age 18 or beyond.²⁰

Another advance in sickle cell care is screening for stroke risk through transcranial Doppler ultrasound (TCD).^44–47 This screening tool has reduced the incidence of childhood stroke from 10% by age 11 to 1%. TCDs typically cannot be performed after the age of 16 due to changes in the skull. Individuals found to have abnormal (elevated) TCD velocities are placed on chronic transfusion therapy for primary stroke prevention. They may remain on monthly chronic transfusions, with the goal of suppressing the percentage of HbS to 30% to 50% indefinitely. A clinical trial (STOPII) designed to determine if pediatric sickle cell disease patients on chronic transfusion therapy for primary stroke prevention could be safely taken off transfusion therapy was discontinued early due to an excess of strokes and conversion to abnormal TCD velocities in the untransfused arm.⁴⁴ Individuals who have experienced an ischemic stroke have a 70% risk of another stroke, and must remain on chronic transfusion therapy indefinitely. Chronic transfusion reduces their stroke risk to 13%.

The only widely used pharmacologic therapy for sickle cell disease is hydroxyurea.^12,48–50 A significant portion of the benefit of hydroxyurea stems from its induction of HbF.⁵¹ HbF does not sickle, and it interrupts the polymerization of HbS in the cell, if present in high enough concentrations.⁵⁰ The level of HbF needed to achieve clinical improvement is not known, but in vitro assays suggest 20% HbF is needed to prevent sickling.^52,53 However, endogenous and hydroxyurea-induced HbF is not distributed evenly through the red cells, so sickling is possible regardless of the level of HbF induced.^54,55 Hydroxyurea likely has other disease-modifying effects as well, including reduction of white blood cell count and reticulocyte count and reduction of red cell adhesion to the endothelium.^56–58 Clinical criteria for initiation of hydroxyurea in adult sickle cell disease patients are shown in Table 4.²⁰ Hydroxyurea is given daily and is dosed to maximum tolerated dose for the individual by following the absolute neutrophil count (ANC). The goal ANC is between 2000 and 4000/µL. At times, absolute reticulocyte count (ARC) can be dose-limiting; goal ARC is greater than 70,000/µL.⁵⁹ Platelet counts may be reduced as well, especially in HbSC patients.^60,61

The only curative therapy for sickle cell disease is hematopoietic stem cell transplant.⁶² Transplant use is limited by availability of matched sibling donors,⁶² and even at experienced centers transplant carries a small risk for mortality, graft rejection, and graft-versus-host disease. Furthermore, consensus on disease complications for which transplant is recommended is also lacking.^63–65 Clinical trials of gene therapy for sickle cell disease and thalassemia are ongoing.⁶⁶

COMPLICATIONS AND DISEASE-SPECIFIC THERAPIES

CASE PRESENTATION

A 26-year-old African-American man who works as a school bus driver presents to an academic center’s emergency department complaining of pain in his left leg, similar to prior pain events. He is described as having sickle cell trait, although no hemoglobin profile is available in his chart. He describes the pain as dull and aching, 10/10 in intensity. A complete blood count (CBC) is obtained; it reveals a hemoglobin of 14.5 g/dL, white blood cell (WBC) count of 5600/µL, and platelet count of 344,000/µL. His CBC is also notable for a mean corpuscular volume (MCV) of 72 fL, a mean corpuscular hemoglobin concentration (MCHC) of 37 g/dL, and a red blood cell distribution width (RDW) of 12. Slide review of a peripheral blood smear shows 2+ target cells (Figure).

The patient is given 6 mg of morphine, which provides some relief of his pain, and is discharged with a prescription for hydrocodone bitartrate/acetaminophen 5/325 mg. The diagnosis given is musculoskeletal pain, and he is instructed to follow-up with a primary care physician. His past medical history is significant for 4 or 5 visits to the emergency department per year in the past 4 years. Prior to 4 years ago, he rarely required medical attention.

• What laboratory and clinical features might lead you to question the diagnosis of sickle cell trait in this patient?

The patient’s hemoglobin is within normal range, which is consistent with sickle cell trait; however, he is microcytic, with a normal RDW. It is possible to be mildly microcytic in the early stages of iron deficiency, prior to the development of anemia, but the RDW would typically be elevated, demonstrating the presence of newer, smaller cells produced under conditions of iron deficiency.⁶⁷ It is also possible that his microcytosis with a normal RDW could represent sickle cell trait with co-inheritance of β-thalassemia. Up to 30% of African Americans have β-thalassemia,² and 1 in 10 have sickle cell trait.⁶⁸ However, a high MCHC, indicating the presence of dense cells, and target cells noted on slide review are most consistent with HbSC.⁹ HbSC patients, especially males, can have hemoglobin levels in the normal range.⁴ The biggest inconsistency with the diagnosis of sickle cell trait is his history of frequent pain events. Individuals with sickle cell trait rarely present with pain crises, except under extreme conditions of dehydration or high altitude.⁶⁸ Sickle cell trait is generally regarded as a benign condition, although a study of U.S. military recruits found a 30-fold higher risk of sudden death during basic training in persons with sickle cell trait.⁶⁹ Additional sickle cell trait–related complications include hematuria, risk of splenic sequestration or infarct under extreme conditions and high altitude, and a rare and usually fatal renal malignancy, renal medullary carcinoma, which is vanishingly rare in individuals without sickle cell trait.^70,71 Although the patient reported having sickle cell trait, this diagnosis should have been verified with a hemoglobin panel, given his atypical presentation.²⁰

In sickle cell disease, vaso-occlusive pain events can be common, often beginning in early childhood.¹⁷ This disease complication accounts for 95% of all adult sickle cell disease hospitalizations.⁷² There is a great deal of variability in pain symptoms between individuals, and within individuals at various times in their lives:⁷³ 30% have no pain events, 50% have occasional events, and 20% have monthly or more frequent events that require hospitalization.⁷⁴ The frequency and severity of pain events are modulated by HbF levels, β-thalassemia status, genotypes, therapies like hydroxyurea, or in rare cases, chronic transfusion therapy.²³ Personal factors, such as psychosocial stressors, also contribute to the frequency of pain events.⁷⁵ Pain event triggers include exposure to cold water, windy or cold weather, temperature changes, and extreme temperatures.^76–79 Patient age also contributes to pain event frequency. Many patients see an increase in pain event frequency in their late 20s, and a marked decrease in their 40s.^23,73 More than 3 pain events per year is associated with reduced life expectancy.²³

Acute management of pain episodes involves nonsteroidal anti-inflammatory drugs, oral opioids, and when hospitalization is required, intravenous opioids, often delivered via patient-controlled analgesia (PCA) pumps.⁷⁹ As sickle cell disease patients become teenagers and young adults, some experience an increased frequency of pain episodes, with fewer pain-free days, or a failure to return to baseline before the next pain crisis occurs.^80,81 This is characteristic of emerging chronic pain.⁸² Chronic pain is a significant problem in adult patients with sickle cell disease, with up to 85% reporting pain on most days.^72,80 The development of chronic pain may be reduced by early and aggressive treatment of acute pain events, as well as use of hydroxyurea to reduce the number of pain events. Many adult sickle cell patients with chronic pain are treated with daily opioids.²⁰ Given the significant side effects of chronic opioid use—sedation, respiratory depression, itching, nausea, and impairment of function and quality of life—non-opioid therapies are under investigation.⁸³ Many chronic pain patients have symptoms of neuropathic pain, and may benefit from neuropathic agents like gabapentin, both to reduce opioid use and to more effectively treat chronic neuropathic pain, which is known to respond poorly to opioids.^84–86

• Is the patient’s peripheral blood smear consistent with a diagnosis of sickle cell trait?

Several target cells are visible, which is not typical of sickle cell trait, but may be seen in HbSC or thalassemia. The finding of an intracellular crystal is pathognomonic for HbSC or HbCC. HbC polymerizes in high oxygen conditions, opposite of HbS, which polymerizes in low oxygen conditions.⁹

CASE CONTINUED

The patient’s family history is significant for a sister who died at age 3 from sickle cell–related complications, and a sister with sickle cell trait who had a cholecystectomy for gallstones at age 22. His father died at age 38 due to unknown causes. The sickle cell trait status of his parents is unknown. His mother is alive, and has hypertension.

• Is the medical history of this patient’s family members consistent with sickle cell trait?

It is unlikely that sickle cell trait would result in early death in childhood, or in gallstones at age 22. Gallstones in early adulthood is a common presentation for HbSC patients not diagnosed by newborn screening.⁸⁷ Any hemolytic condition can lead to the formation of hemoglobin-containing pigmented gallstones, biliary sludge, and obstruction of the gallbladder. In the presence of right-sided abdominal pain, a serum bilirubin level of more than 4 mg/dL should lead to measurement of direct bilirubin; if greater than 10% of total, imaging of the gallbladder should be obtained. In sickle cell disease, 30% of patients will have gallstones by 18 years of age. The low hemolysis/high viscosity phenotype patients are typically older at diagnosis. Co-inheritance of Gilbert syndrome and sickle cell disease is not uncommon, and can result in formation of gallstones at a young age; Gilbert syndrome alone typically results in gallstones in mid-life.⁸⁸

Two months later, the patient presents again to the emergency department with the same complaint of leg pain, as well as abdominal pain. His hemoglobin is 12.5 g/dL, and his platelet count is 134,000/µL. His pain is not improved with 3 doses of morphine 6 mg intravenously, and he is admitted to the medicine service. A hemoglobin profile is obtained, revealing 52% HbS, 45% HbC, and 1.5% HbF, consistent with HbSC. In sickle cell trait, the hemoglobin profile is 60% HbA and 40% HbS (available α-globin prefers to pair with a normal β-globin, so the ratio of HbA to HbS is 60:40, not 50:50).

On the second hospital day, the patient’s hemoglobin drops to 7.2 g/dL and his platelet count decreases to 44,000/µL. His abdomen is distended and diffusely tender. The internist transfuses him with 2 units of packed red blood cells (PRBC), after which his hemoglobin increases to 11 g/dL, while his platelet count increases to 112,000/µL. Following the transfusion, his abdominal pain resolves, as does his anemia and thrombocytopenia.

• What caused this patient’s anemia and thrombocytopenia?

High on the differential diagnosis is a splenic sequestration. Acute splenic sequestration occurs when red cells are trapped in the splenic sinuses. Massive splenic enlargement may occur over several hours.^89,90 Unrecognized splenic sequestration has a high mortality rate from severe anemia and splenic rupture.⁹⁰ Splenic sequestration must be ruled out in a sickle cell patient with abdominal pain accompanied by dropping platelet and red cell counts, especially in milder subtypes that often have splenic function preserved into adolescence and adulthood. Sickle cell anemia patients usually become functionally asplenic in early childhood.^89,91,92 The rise in hemoglobin, more than would be expected from 2 units of PRBC, plus the improvement in platelet count without a platelet transfusion observed in the case patient strongly supports the diagnosis of splenic sequestration.

Splenic sequestration can occur in any sickle cell patient whose spleen has not fibrosed. Splenic sequestration in adulthood is not uncommon in HbSC patients, who often have preserved splenic function into adulthood.^93–95

Clinical signs of splenic sequestration include a rapid drop in hemoglobin, rise in reticulocyte count, a tender, enlarged spleen, and, in severe cases, hypovolemia.^89,93 It is treated with prompt blood transfusion, but care must be taken not to overtransfuse the patient, as the spleen can trap several grams of hemoglobin, which may be released upon transfusion, potentially causing life-threatening hyperviscosity.⁸⁹ Hemoglobin levels must be checked following transfusion in suspected splenic sequestration, and “mini transfusions” of 5 mL/kg are recommended in sickle cell disease patients who are hemodynamically stable.²⁰

Hepatic sequestration may also occur, but it is much less common than splenic sequestration.⁹⁶ Other conditions on the differential diagnosis include thrombotic thrombocytopenic purpura, which would be unlikely to respond to a transfusion. ACS can cause a drop in hemoglobin, and is treated with simple or exchange transfusions.⁹⁷ ACS is less likely without respiratory symptoms or oxygen requirement, and usually is not associated with thrombocytopenia. Sepsis may also cause anemia and thrombocytopenia, but again would not likely respond to a simple transfusion. The patient’s response to transfusion is consistent with a sequestering event, not a destructive event as in the case of sepsis.

CASE CONTINUED

Imaging reveals a grossly enlarged spleen, which is having a mass effect on the left kidney. The patient is started on hydroxyurea therapy at 500 mg 3 times daily. Discharge instructions include following up with his primary care physician, continuing hydroxyurea therapy, and receiving yearly dilated eye exams to evaluate for proliferative sickle retinopathy.

• Are these discharge instructions complete?

Splenic sequestration has a 50% recurrence rate.⁹⁸ In very young children, watchful waiting or chronic transfusion may be implemented to preserve the immunologic function of the spleen and reduce the risk of sepsis.⁸⁹ Splenectomy after a single episode of sequestration in adults is a matter of debate, with experts advising both watchful waiting⁹⁹ and splenectomy after recovery from the first sequestering event.¹⁰⁰ The patient should have been informed of the risk for recurrence, and the signs and symptoms of splenic sequestration as well as the need for emergency medical attention should have been discussed. Splenic sequestration may be milder in adults than in children, but fatal sequestrations have been reported.^95,101–103

Proliferative sickle cell retinopathy is a high viscosity/high hemoglobin complication that may occur more frequently in HbSC than HbSS, with an incidence of 33% in HbSC.^42,104 Spontaneous regression of retinopathy occurs in approximately 32% of eyes, and laser or scatter photocoagulation is an effective intervention.¹⁰⁵

• Would the patient need to be transfused prior to splenectomy?

Preoperative transfusion therapy is standard of care for HbSS patients undergoing general anesthesia. The TRAP study found that simple “top off” transfusion to a hemoglobin of 10 g/dL was as effective at preventing postoperative sickle cell–related complications as exchange transfusion to HbS of 30% or less, and had fewer transfusion-related complications like alloimmunization.¹⁰⁶ There is little data regarding preoperative transfusions in HbSC disease. A retrospective study suggests that HbSC patients undergoing abdominal surgeries should be transfused.¹⁰⁷ The higher hemoglobin level of the typical HbSC patient necessitates exchange transfusion to avoid hyperviscosity.

• Is hydroxyurea therapy indicated in this patient?

• Has it been dosed appropriately?

If the patient had the HbSS subtype, hydroxyurea would be clearly indicated, given his frequent pain events.²⁰ HbSC patients may be placed on hydroxyurea on a case-by-case basis, but evidence for its efficacy in this sickle cell subtype is lacking.¹⁰⁸ Large clinical trials like the Multi-Center Study of Hydroxyurea (MSH) that established the safety and efficacy of hydroxyurea in sickle cell anemia excluded HbSC and HbSβ⁺ patients.¹⁰⁹ These mild to moderate subtypes produce less HbF at baseline, and typically have a minimal to modest rise in HbF on hydroxyurea.¹¹⁰ In sickle cell anemia, hydroxyurea is titrated to maximum tolerated dose, defined as an ANC of 2000 to 4000/µL and an ARC of 70,000/µL or higher.⁵³ Because of their lower levels of chronic inflammation and lower reticulocyte counts due to higher hemoglobin levels, many HbSC and HbSβ⁺ patients have values in that range before initiating hydroxyurea therapy.⁹ Cytopenias, particularly of platelets in HbSC, occur at low doses of hydroxyurea.¹¹¹

Of note, although the half-life of hydroxyurea would suggest that 3 times daily dosing is indicated, daily dosing has been found to have equal response and is preferred. Another concern is the monitoring of this myelosuppressive medication. This patient has repeatedly failed to obtain a primary care physician or a hematologist, and hydroxyurea requires laboratory monitoring at least every 2 months, especially in a HbSC patient with a very large spleen who is at significant risk for thrombocytopenia and neutropenia.⁹

CASE CONTINUED

A week after discharge from his admission for abdominal pain diagnosed as splenic sequestration, the patient presents again to the emergency department with abdominal pain which he reports is his typical sickle cell pain. Hemoglobin is 13.8 g/dL, platelet count is 388,000/µL, and alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels are both 10 times their prior value. Creatinine is 1.2 mg/dL (0.75 mg/dL on his prior admission), and total bilirubin is 3 mg/dL, with 0.3 mg/dL direct bilirubin. He undergoes an ultrasound exam of his gallbladder, which reveals sludge and a possible gallstone. There is no evidence of cholecystitis. General surgery performs a laparoscopic cholecystectomy.

• Was this cholecystectomy necessary?

In patients with sickle cell disease, symptomatic gallstones and gallbladder sludge should be observed; recurrent abdominal pain without a significant change in bilirubin may not be due to gallstones or sludge, and therefore may not be relieved by cholecystectomy.^112,113 In sickle cell disease, 40% of patients with gallbladder sludge do not develop gallstones.⁸⁷ The patient’s bilirubin level was at baseline, and there was no increase in the direct (conjugated) fraction. Watchful waiting would have been appropriate, with cholecystectomy being performed if he experienced recurrent symptoms associated with fatty foods accompanied by an elevation in direct bilirubin.

More concerning and deserving of investigation was his elevated liver enzymes. Patients with sickle cell disease may experience recurrent ischemia and reperfusion injuries in the liver, which is called right upper quadrant syndrome. On autopsy of 70 sickle cell patients, 91% had hepatomegaly and 34% had focal necrosis.¹¹⁴ AST is often elevated in sickle cell disease, as it is affected by hemolysis. In this patient, both AST and ALT are elevated, consistent with a hepatocellular disorder. His abdominal pain and ALT rise may be a sign of a hepatic crisis.¹¹⁵ Rapid resolution of ALT elevation in a matter of days suggests a vaso-occlusive, inflammatory event that is self- limiting. Prolonged AST elevation requires further investigation, with consideration of autoimmune hepatitis, viral hepatitis, or iron overload. Iron overload is unlikely in this patient given his lifetime history of only 1 transfusion. Hepatic iron overload typically occurs in sickle cell disease after a minimum of 10 transfusions.¹¹⁵

CASE CONTINUED

The patient is discharged on the day after the procedure, with instructions to continue his hydroxyurea.

• Should the patient resume hydroxyurea therapy?

Hydroxyurea is hepatically cleared and thus it should be held until his liver function tests normalize.¹⁰⁶

CASE CONTINUED

Two months later, the patient presents to the emergency department with abdominal pain that moves to his left leg. A CBC is obtained, showing a hemoglobin of 11.8 g/dL and a platelet count of 144,000/µL. He is given 2 doses of morphine 6 mg intravenously, and reports that his leg pain is now a 4/10. He is discharged home with a prescription for hydrocodone/acetaminophen.

• Is the emergency department evaluation sufficient?

This patient remains at high risk for splenic sequestration,⁹³ with a hemoglobin 2 g lower than it was 2 months ago and platelets less than half. This decline could be consistent with early splenic sequestration.²⁰ Additionally, he had elevated liver function tests on a recent admission, as well as rising creatinine, without evidence of resolution. It is not appropriate to discharge him without checking a chemistry and liver panel, and abdominal imaging should be considered. The best plan would be to admit him for observation, given his risk for splenic sequestration, and consult surgery for an elective splenectomy if he has a second episode of splenic sequestration 2 months after the first.¹⁰⁰ His abdominal pain that migrates to his left leg could be due to his massive splenomegaly compressing his left kidney, as noted on imaging during his recent admission for splenic sequestration

CASE CONTINUED

An hour after discharge from the emergency department, EMS is called to his home for intractable pain. He is found lying on the floor, and reports excruciating left leg pain. He is brought to the closest hospital, a community hospital that he has not visited previously. There, he is admitted for hydration and pain control and placed on hydromorphone 2 mg every 4 hours as needed for pain. His hemoglobin is 10.8 g/dL, and platelets are 121,000/µL. A chemistry panel is remarkable for a creatinine level of 1.5 mg/dL and a potassium level of 3.2 mEq/L. Liver function tests are not obtained. After 3 doses of hydromorphone, he falls asleep. He is not in a monitored bed, and intravenous fluids, while ordered, are not started. At 6:30 AM the day after admission, he cannot be aroused on a routine vital sign check; he has an SpO₂ of 60%, a blood pressure of 80/60 mm Hg, and heart rate of 148 beats/min. A rapid response is called, and naloxone is administered along with oxygen by face mask and several fluid boluses. His systolic blood pressure increases to 100 mm Hg from a low of 70 mm Hg. His SpO₂ increases to 92%, and he is arousable and alert, although he reports 10/10 leg pain. His abdomen is noted to be distended and tender.

• What may have contributed to his clinical condition?

The patient is opioid tolerant and has received equivalent doses of opioids in the past without excess sedation. He may have liver dysfunction making him unable to metabolize opioids effectively. His hemoglobin and platelets continue to decline, raising concern for splenic sequestration versus sepsis. Failure to place him on a monitor allowed his hypoxia to continue for an unknown amount of time, placing him at high risk for developing ACS. Lack of intravenous hydration while he was too sedated to drink likely exacerbated his sickling.

At 9:20 AM, a CBC is obtained and reveals a hemoglobin of 4.8 g/dL and a platelet count of 44,000/µL. Two units of stat O negative blood are administered, and preparations are made to administer an exchange transfusion. A liver panel is obtained 3 hours later, which reveals an AST level of 1200 U/L and an ALT level of 1050 U/L. His bilirubin is 10 mg/dL, and his lactate dehydrogenase level is 1800 U/L. His urine is dark and is positive for bilirubin and ketones. He is transferred to the intensive care unit. A chest X-ray shows pulmonary congestion. Hematology/oncology is consulted.

He receives a 7-unit red blood cell exchange, which reduces his HbS to 11%. He continues to be hypotensive, and requires norepinephrine to support his blood pressure. Antibiotic therapy is started. His creatinine concentration rises to 2.3 mg/dL, potassium is 7.8 mEq/L, and bicarbonate is 12 mEq/L. He is placed on hemodialysis.

Computed tomography of the chest and abdomen reveals lower posterior lung infiltrates and a grossly enlarged spleen. He requires intubation. He is given a diagnosis of ACS in addition to kidney failure, liver failure, and “sickle crisis.” He continues to require daily to twice daily transfusions to maintain a hemoglobin of 7 to 9 g/dL, and his abdominal distension increases. As his condition worsens, surgery is consulted to discuss a liver transplant. He is deemed to not be a surgical candidate, and he passes away 6 days after entering the hospital. The immediate cause of death is listed as vaso-occlusive crisis, with ACS and sickle crisis listed as contributors.

• Are the causes of death accurate and complete?

If vaso-occlusive crisis is used to indicate a pain event, it is not an accurate cause of death. Pain is one of the most distressing complications of sickle cell disease, and frequent pain events are associated with early mortality,^4,80 but they are not in themselves fatal. ACS is the number one cause of death in sickle cell disease,⁴ and it likely contributed to this patient’s death. Sickle crisis is a vague term that should not be used in this context. Causes of death should include splenic sequestration and multisystem organ failure. Multisystem organ failure in sickle cell disease often responds to aggressive transfusion therapy, which this patient received.^116–118

CONCLUSION

Sickle cell disease is a complex chronic disease that impacts almost every organ system in the body. Clinicians may be inclined to attribute most pain in a patient with sickle cell disease to a simple vaso-occlusive crisis, treat them for this, and not investigate further. As the case presented here demonstrates, failure to identify the actual life-threatening process occurring in a patient with sickle cell disease presenting with pain can result in preventable early mortality. Clinicians must approach a sickle cell patient reporting pain in a thoughtful manner, and consider a complete differential diagnosis, including both sickle cell disease complications and those unrelated to sickle cell disease. Knowledge of the disease courses of the different sickle cell genotypes is essential, and must go beyond a superficial hierarchy of severity, but rather include an understanding of the complications each genotype is most prone to, and at what ages. Complete laboratory assessment, including a comprehensive metabolic panel, should be performed on all admitted patients, not just a complete blood count. Treating pain with high-dose opioids, while appropriate in an uncomplicated pain crisis, can lead to ACS or even respiratory failure in a patient with uninvestigated liver and kidney dysfunction. The most important lesson to remember is that even the sickle cell disease patient who has been given the unfortunate and pejorative label of “frequent flyer” by some providers has the potential for rapid deterioration into multisystem organ failure and death.

Empagliflozin (Jardiance) and liraglutide (Victoza) should be considered in type 2 diabetes patients with documented cardiovascular disease to reduce the risk of CV death, according to the American Diabetes Association 2017 Standards of Medical Care.

ADA updates it standards annually based on new information and research; like its predecessors, the 2017 guidance is comprehensive, addressing mental, social, and other challenges faced by patients with diabetes, along with clinical care (Diabetes Care. 2017 Jan;40(Suppl 1):S4-S5).

The 2017 guidance contains a great deal of new information. At 135 pages, there are 22 more pages than in 2016. “They did a really nice job. This guide is useful for anyone helping patients with diabetes,” including diabetologists, dietitians, educators, psychologists, and social workers, Richard Hellman, MD, a clinical endocrinologist in North Kansas City, Mo., said in an interview.

[email protected]

Empagliflozin (Jardiance) and liraglutide (Victoza) should be considered in type 2 diabetes patients with documented cardiovascular disease to reduce the risk of CV death, according to the American Diabetes Association 2017 Standards of Medical Care.

ADA updates it standards annually based on new information and research; like its predecessors, the 2017 guidance is comprehensive, addressing mental, social, and other challenges faced by patients with diabetes, along with clinical care (Diabetes Care. 2017 Jan;40(Suppl 1):S4-S5).

The 2017 guidance contains a great deal of new information. At 135 pages, there are 22 more pages than in 2016. “They did a really nice job. This guide is useful for anyone helping patients with diabetes,” including diabetologists, dietitians, educators, psychologists, and social workers, Richard Hellman, MD, a clinical endocrinologist in North Kansas City, Mo., said in an interview.

[email protected]

Empagliflozin (Jardiance) and liraglutide (Victoza) should be considered in type 2 diabetes patients with documented cardiovascular disease to reduce the risk of CV death, according to the American Diabetes Association 2017 Standards of Medical Care.

ADA updates it standards annually based on new information and research; like its predecessors, the 2017 guidance is comprehensive, addressing mental, social, and other challenges faced by patients with diabetes, along with clinical care (Diabetes Care. 2017 Jan;40(Suppl 1):S4-S5).

The 2017 guidance contains a great deal of new information. At 135 pages, there are 22 more pages than in 2016. “They did a really nice job. This guide is useful for anyone helping patients with diabetes,” including diabetologists, dietitians, educators, psychologists, and social workers, Richard Hellman, MD, a clinical endocrinologist in North Kansas City, Mo., said in an interview.

[email protected]

The American Diabetes Association has updated its guidance for those treating and researching hypoglycemia , joining with the European Association for the Study of Diabetes to specify that a level of less than 3 mmol/L (54 mg/dL) should be considered “clinically important hypoglycemia.”

“A single glucose level should be agreed to that has serious clinical and health-economic consequences,” the ADA and EASD stated in the new guidelines. “This would enable the diabetes and regulatory communities to compare the effectiveness of interventions in reducing hypoglycemia, be they pharmacological, technological, or educational. It would also permit the use of meta-analysis as a statistical tool to increase power when comparing interventions.”

An international, multidisciplinary group – the International Hypoglycemia Study Group – was formed to create distinct definitions of the various levels of severity that hypoglycemia can have. The new guidelines contain three levels, which should be used by clinicians to determine what amounts of blood glucose are significant enough to be clinically reported.

[email protected]
 

The American Diabetes Association has updated its guidance for those treating and researching hypoglycemia , joining with the European Association for the Study of Diabetes to specify that a level of less than 3 mmol/L (54 mg/dL) should be considered “clinically important hypoglycemia.”

“A single glucose level should be agreed to that has serious clinical and health-economic consequences,” the ADA and EASD stated in the new guidelines. “This would enable the diabetes and regulatory communities to compare the effectiveness of interventions in reducing hypoglycemia, be they pharmacological, technological, or educational. It would also permit the use of meta-analysis as a statistical tool to increase power when comparing interventions.”

An international, multidisciplinary group – the International Hypoglycemia Study Group – was formed to create distinct definitions of the various levels of severity that hypoglycemia can have. The new guidelines contain three levels, which should be used by clinicians to determine what amounts of blood glucose are significant enough to be clinically reported.

[email protected]
 

The American Diabetes Association has updated its guidance for those treating and researching hypoglycemia , joining with the European Association for the Study of Diabetes to specify that a level of less than 3 mmol/L (54 mg/dL) should be considered “clinically important hypoglycemia.”

“A single glucose level should be agreed to that has serious clinical and health-economic consequences,” the ADA and EASD stated in the new guidelines. “This would enable the diabetes and regulatory communities to compare the effectiveness of interventions in reducing hypoglycemia, be they pharmacological, technological, or educational. It would also permit the use of meta-analysis as a statistical tool to increase power when comparing interventions.”

An international, multidisciplinary group – the International Hypoglycemia Study Group – was formed to create distinct definitions of the various levels of severity that hypoglycemia can have. The new guidelines contain three levels, which should be used by clinicians to determine what amounts of blood glucose are significant enough to be clinically reported.

[email protected]
 

For patients with submassive acute pulmonary embolism—the intermediate category between massive and low-risk—whether to give thrombolytic therapy is controversial. In general, patients with massive pulmonary embolism need this therapy, whereas those with low-risk pulmonary embolism do not—and neither do most of those with submassive embolism. But where should we draw the line?

See related editorial

More than 600,000 patients suffer pulmonary embolisms every year in the United States, and 50,000 to 200,000 people die of them.^1–3 In various studies,^4–6 within 1 year, 12.9% of patients had another pulmonary embolism, 7.3% developed chronic venous insufficiency, and 3.8% developed chronic thromboembolic pulmonary hypertension.

THREE CATEGORIES OF RISK

Episodes of acute pulmonary embolism are classified as low-risk (about 70% of cases), hemodynamically unstable or massive (5%), or submassive (25%).^7,8

Low-risk acute pulmonary embolism is defined by the absence of right ventricular dysfunction and the absence of myocardial necrosis. The death rate in such cases is less than 1%.⁹ Its pharmacologic management includes parenteral anticoagulation and early initiation of long-term anticoagulation therapy, which the American College of Chest Physicians (ACCP) gives a grade IB recommendation (strong, based on moderate-quality evidence).¹⁰

Massive or hemodynamically unstable pulmonary embolism is characterized by any of the following, in the absence of other causes⁸:

• Sustained hypotension (systolic blood pressure < 90 mm Hg for ≥ 15 minutes)
• An absolute decrease in systolic blood pressure of 40 mm Hg or more
• Need for inotropic support
• Cardiac arrest
• Bradycardia (heart rate < 40 beats per minute).

The death rate is more than 30% in patients presenting with shock and approaches 70% in those presenting with cardiac arrest.^11,12 Therefore, the consensus is that this category of pulmonary embolism merits aggressive treatment. Systemic thrombolytic therapy is recommended in those who are not at high risk of major bleeding, though the ACCP gives it only a grade 2C recommendation (weak, based on low-quality evidence).¹⁰

Submassive pulmonary embolism is defined by evidence of right ventricular dysfunction with normal blood pressure. According to the ACCP guidelines, thrombolytic therapy should be considered (grade 2C recommendation) for patients with acute pulmonary embolism without hypotension and with a low bleeding risk (with no renal failure and not on dual antiplatelet therapy), but at high risk of developing hypotension.¹⁰

DIAGNOSING SUBMASSIVE PULMONARY EMBOLISM, DELINEATING ITS SEVERITY

In managing acute pulmonary embolism, it is critical to recognize whether a patient is at high risk of clinical deterioration.

Blood pressure

The systolic blood pressure not only indicates whether the patient has hypotension (systolic blood pressure < 90 mm Hg) and therefore massive rather than submassive or low-risk pulmonary embolism; it is also important as a baseline value. A decrease in systolic blood pressure of 40 mm Hg or more is associated with worse outcomes.¹²

Right ventricular dysfunction

The physiologic response to occlusion of the pulmonary arteries can result in early myocardial injury and right ventricular dysfunction, which can be assessed by various methods (Table 1).

Electrocardiographic signs. Right heart strain may be recognized on electrocardiography as:

• Evidence of new complete or incomplete right bundle branch block
• T-wave inversion in the anterolateral leads V₁ to V₄
• S1Q3T3 (a large S wave in lead I, a Q wave in lead III, and an inverted T wave in lead III, the classic pattern of acute cor pulmonale).¹³ 

These findings add incremental prognostic value to echocardiographic findings in patients with submassive pulmonary embolism.¹⁴

Cardiac biomarkers such as B-type natriuretic peptide (BNP), N-terminal-pro-BNP (NT-pro-BNP), cardiac troponins, and heart-type fatty acid-binding protein (H-FABP) are also markers of right-sided myocardial damage and strain and can be used to identify patients with submassive pulmonary embolism.¹⁵ Abnormal levels of these substances are as follows:

• Troponin T greater than 0.1 ng/mL
• Troponin I greater than 0.4 ng/mL
• BNP greater than 90 pg/mL
• NT-pro-BNP greater than 500 pg/mL
• H-FABP less than 6 ng/mL.

These levels have prognostic value, identifying patients with submassive pulmonary embolism at risk of deterioration or death,^14,16,17

Echocardiographic signs. Right ventricular dysfunction can be assessed quickly at the bedside with portable transthoracic echocardiography. A meta-analysis showed that close to 37% of hemodynamically stable patients with pulmonary embolism had echocardiographic evidence of right ventricular dysfunction on presentation and a higher short-term mortality rate.¹⁸ Evidence of right ventricular dysfunction includes the following:

• New-onset hypokinesis or akinesis
• Right ventricular dilation
• Right ventricular free-wall hypokinesis with apical sparing (the McConnell sign)
• Paradoxical movement of the interventricular septum
• Newly increased right ventricular systolic pressure
• Pulmonary hypertension, defined as tricuspid regurgitation jet velocity greater than 2.8 m/s.^15,19

Computed tomographic (CT) angiography is widely available. Findings that have prognostic value in determining those at higher risk of death include the following^20,21:

• A dilated right ventricle—ratio of right ventricle to left ventricle diameter (RV:LV ratio) greater than 0.9
• Interventricular septal bowing.

PESI and sPESI scores. The European Society of Cardiology 2014 guidelines stratify the risk in normotensive patients with pulmonary embolism according to their score on the Pulmonary Embolism Severity Index (PESI) or the simplified PESI (sPESI). There are five PESI classes. Those in PESI class III or IV or with an sPESI score of 1 or more (on a scale of 0 to 6) are considered at intermediate risk of clinical deterioration and are then further risk-stratified according to whether they have right ventricular dysfunction (based on echocardiography or computed tomography) and elevated cardiac biomarkers. These scoring systems are based on easily obtainable clinical information such as age, male sex, history of cancer, history of heart failure, history of chronic lung disease, heart rate, systolic blood pressure, respiratory rate, temperature, and altered mental status, and calculators are readily available.

Anticoagulation for all, plus thrombolysis for some

Patients with neither right ventricular dysfunction nor elevated cardiac biomarkers are at intermediate to low risk of clinical deterioration, and it is recommended that they be given anticoagulation therapy in an inpatient setting.

On the other hand, patients with both right ventricular dysfunction and elevated cardiac biomarkers are considered at intermediate to high risk of clinical deterioration; they should also be managed with anticoagulation and monitored closely for the need for rescue reperfusion therapy with thrombolytics.²²

THROMBOLYTIC AGENTS

Thrombolytic agents are the cornerstone of management for patients presenting with pulmonary embolism who are at high risk. As noted above, these agents are recommended in massive pulmonary embolism, but their role in submassive pulmonary embolism remains controversial.

Thrombolytics work by activating endogenous plasminogen. The resulting plasmin promotes clot lysis, reducing the size of the thrombus, decreasing pulmonary vasculature resistance, and improving right ventricular function.²³

To date, three thrombolytic agents have received US Food and Drug Administration approval for use in massive pulmonary embolism: alteplase, urokinase, and streptokinase. But only alteplase is still available in the United States. Alteplase is also the best tolerated, whereas streptokinase is highly antigenic and may cause further deterioration in an already unstable patient. Alteplase is also the most fibrin-specific and is considered the most potent of the three agents.²⁴

Additional thrombolytic agents under investigation for use in acute pulmonary embolism include reteplase, tenecteplase, and desmoteplase. These agents are more fibrin-specific than alteplase. Reteplase is a second-generation recombinant tissue-type plasminogen activator with a quicker onset of action and longer half-life than alteplase, allowing for bolus dosing. Tenecteplase, a variant of alteplase, is cleared more slowly and is 14 times more fibrin-specific than alteplase, also allowing for bolus dosing. Desmoteplase, a fibrin-specific agent currently in phase 2 trials, also has a longer half-life and appears to be more potent than alteplase. Table 2 lists the dosing and the degree of fibrin specificity of these agents.

Complications of thrombolytic therapy

Submassive pulmonary embolism has a low death rate, and the benefit of systemic thrombolytic therapy for this condition is controversial. Therefore, risk stratification is very important before pursuing this therapy.

A meta-analysis²⁵ of 16 randomized controlled trials included 2,125 patients with pulmonary embolism:

• 210 (9.88%) in the low-risk category
• 1,499 (70.54%) in the submassive category
• 31 (1.46%) in the massive category
• 385 (18.11%) whose disease severity could not be determined.

Major bleeding occurred in:

• 98 (9.24%) of 1,061 patients receiving anticoagulation plus thrombolytics
• 36 (3.42%) of 1,054 patients receiving anticoagulation without thrombolytics (odds ratio [OR] 2.73, 95% confidence interval [CI] 1.91–3.91; number needed to harm [NNH] 18, 95% CI 13–27).

Intracranial hemorrhage occurred in:

• 15 (1.46%) of 2,014 patients on thrombolytic therapy
• 2 (0.19%) of 1,019 patients not on thrombolytic therapy (OR 4.63, 95% CI 1.78–12.04; NNH 78, 95% CI 48–206).

Of note, the incidence of major bleeding was not significantly increased in those age 65 or younger receiving thrombolytics (OR 1.25, 95% CI 0.5–3.14).

Comments. Definitions of major bleeding varied in the individual trials. Additionally, intracranial hemorrhage was included as a major bleeding end point in any trial in which it was not prespecified.

These findings emphasize the importance of risk stratification before pursuing thrombolytic therapy in submassive pulmonary embolism.

Table 3 lists absolute and relative contraindications to thrombolytic therapy.

MAJOR STUDIES IN SUBMASSIVE PULMONARY EMBOLISM

The MAPPET-3 trial

The Management Strategies and Prognosis of Pulmonary Embolism-3 (MAPPET-3) trial,²⁶ in 2002, was the first major trial to study thrombolytic therapy in submassive pulmonary embolism.

In this prospective, randomized, double-blinded trial conducted in Germany, 118 patients received heparin with alteplase (100 mg over 2 hours) and 138 received heparin with placebo. The primary end point was in-hospital death or clinical deterioration requiring escalation of treatment. Secondary outcomes included recurrent pulmonary embolism, major bleeding, and stroke. Major bleeding was defined as fatal bleeding, hemorrhagic stroke, or drop in the hemoglobin concentration by more than 4 g/dL, with or without the need for red blood cell transfusion.

Right ventricular dysfunction was diagnosed by echocardiography in 30% of the participants, and the rest of the patients were classified as having submassive pulmonary embolism on the basis of electrocardiographic criteria alone. It is likely that the latter group had a less severe form of the disease and did not benefit from thrombolytic therapy as much as patients with echocardiographic findings of right ventricular dysfunction and elevated serum cardiac biomarkers.

Results. At 30 days, 11% of the alteplase-plus-heparin group had met the primary end point, compared with 24.6% of the placebo-plus-heparin group (P = .006). The difference was mostly driven by the need for secondary thrombolysis (7.6% vs 23.2%, P = .001), since 32 (23.2%) of the 138 patients in the control group required secondary thrombolysis, accounting for 94% of the 34 patients in this group who required escalation of treatment. Most cases of clinical deterioration in this group occurred within the first 5 days.

Mortality rates were 3.4% in the heparin-plus-alteplase group and 2.2% in the heparin-plus-placebo group, but the difference was not statistically significant (P = .71).

Major bleeding occurred in 1 patient in the heparin-plus-alteplase group and 5 patients in the heparin-plus-placebo group, but the trial’s definition of major bleeding may have resulted in underestimation of this event. The definition put forth by the International Society on Thrombosis and Haemostasis is less strict, defining bleeding in nonsurgical patients as major if it is fatal, symptomatic in a critical area or organ, or causing a fall in hemoglobin level of 2.0 g/dL or more, leading to transfusion of two or more units of whole blood or red cells.²⁷

MOPETT trial

The Moderate Pulmonary Embolism Treated with Thrombolysis (MOPETT) trial²⁸ was a single-center, randomized trial in 121 normotensive patients with “moderate” pulmonary embolism and right ventricular dysfunction. Moderate pulmonary embolism was defined as signs and symptoms of pulmonary embolism with evidence on computed tomographic angiography of greater than 70% involvement with thrombus in two or more lobes or left or right main pulmonary arteries, or by a high-probability ventilation-perfusion scan showing a mismatch in two or more lobes.

The authors defined right ventricular dysfunction by elevated cardiac markers or by findings on echocardiography. Only 20% of the participants were enrolled on the basis of right ventricular dysfunction on echocardiography, whereas almost 60% had elevated cardiac biomarkers.

The primary outcome was the development of pulmonary hypertension, based on echocardiography.

Patients were randomized to either anticoagulation alone (unfractionated heparin or low-molecular-weight heparin) or anticoagulation plus half-dose alteplase (0.5 mg/kg, to a maximum of 50 mg). Echocardiography was performed within 2 hours of study entry, at 48 hours, and every 6 months thereafter. The mean duration of follow-up was 28 months.

Results. Pulmonary hypertension developed in 16% of the anticoagulation-plus-alteplase group vs 57% of the anticoagulation-only group (P < .001). However, the clinical relevance of elevated right-sided pressures observed by echocardiography in asymptomatic patients is uncertain. Alteplase had no impact on the rates of death or recurrent pulmonary embolism.

PEITHO trial