Reviews

In recent years, hospitals and health systems have engaged in considerable efforts to reduce readmissions, in part due to financial incentives from the Medicare Hospital Readmission Reduction Program.^1,2 Though efforts to improve transitions of care for all patients are laudable, risk for readmission is not distributed equally; a small subset of patients accounts for a disproportionate number of hospital readmissions.³ This phenomenon of frequently hospitalized patients is similar to that seen in other populations in which a small proportion of patients account for a majority of healthcare utilization.^3,4

Recognizing that the current system of healthcare delivery does not meet the needs of this population, healthcare organizations have begun to implement interventions that supplement or redesign the system of care for frequently hospitalized patients.^5-7 Descriptive reviews of ambulatory "high-need, high-cost" patients emphasize complex case management and interdisciplinary, team-based care.^8,9 Prior systematic reviews of studies aimed at patients with high use of emergency care demonstrate improvements in social outcomes such as homelessness but mixed results in reducing emergency department (ED) use.¹⁰ However, we were unable to identify any prior reviews that evaluated interventions intended specifically for patients with frequent hospital admissions. Our objective in this systematic review was to characterize interventions for frequently admitted patients and determine whether these interventions decrease use of healthcare resources, improve health outcomes, and/or reduce costs.

Literature Search

We registered our study protocol in the PROSPERO database. A librarian (L.O.) collaboratively developed the search strategies with other review authors (A.G., B.H., N.N.) and in January 2018 ran searches on "super users," "high utilizers," and similar terms in the following databases: PubMed MEDLINE, Embase (embase.com), and Cochrane Central Register of Controlled Trials (CENTRAL) on the Wiley platform. The complete search strategies used are available in Appendix A.  

We attempted to discover additional studies by searching the reference lists of key publications and contacted authors of relevant abstracts to determine whether studies had been published or were planned for peer-reviewed publication. We also contacted authors of included studies to locate additional studies meeting inclusion criteria.

Data Collection Process

Studies were eligible for inclusion in our review if they were (1) published in a peer-reviewed source, (2) defined a study population of patients frequently admitted to inpatient medical services, (3) evaluated an intervention targeting frequently hospitalized patients, and (4) included patients who were >18 years old and (5) admitted as inpatients on medical services. Of note, studies with patients admitted to psychiatric, obstetric, or surgical wards were not included, as the authors did not define these as "general medicine" units. Studies focused solely on an ambulatory population were similarly excluded. Given the heterogeneity of how studies defined frequently hospitalized patients, we did not establish a prespecified number of admissions for inclusion to ensure that we did not exclude interventions not meeting a strict set of criteria. The goal was not to examine interventions to reduce all readmissions, but rather, to look at patients who were recurrently hospitalized. Thus, patients had to be repeatedly admitted, but we let the studies define that usage explicitly.  

Two members of a four-physician team (A.G., B.H., K.O., and N.N.) screened all initial results for eligibility through title and abstract review; potentially relevant articles were retained for full-text review to assess each study's eligibility. If a study's abstract did not clearly indicate whether inclusion criteria were met, we retained the article for full-text review. Two team members (A.G. and B.H.) independently reviewed the full text of each selected article to determine final inclusion in the study. The previously described inclusion criteria were again applied, and a final set of articles was identified for data extraction. Disagreements regarding inclusion in the final review (such as whether a study measured medical or psychiatric hospitalizations) were resolved through discussion among the entire four-physician review team to achieve consensus or, when required, by contacting authors of individual studies.  

Data Abstraction and Risk of Bias Assessment

After selecting the final set of articles, we abstracted data using a tool developed by the Cochrane Effective Practice and Organization of Care Group.¹¹ We then compiled study-level data into a single database for reporting. Extracted elements included study design, setting, patient characteristics, inclusion and exclusion criteria, control group identification, outcome measures, results, and length of follow-up. We also extracted individual characteristics of each intervention, including common intervention elements such as intervention setting, use of health information technology resources, and whether programs developed interdisciplinary care plans. We assessed the risk of bias of each study and the quality of studies using the Downs and Black Scale.^12,13 Two team members (A.G. and B.H.) independently assessed the risk of bias for all nine studies, and differences were resolved by consensus. Due to the variation in the outcomes used, we were unable to conduct a meta-analysis.

Search Results

We found a total of 4,762 references in the three databases. After de-duplication using the EndNote software, there were 3,314 references to screen. We identified 116 studies for full-text review. Of those, we selected nine studies that met the criteria for this study (Figure). The most common reason for exclusion of an article for full-text review was that the patients studied were not defined as high utilizers of inpatient resources and were instead high-utilizers of ambulatory or emergency care (32 studies). We identified five of the included studies through the primary search and four through review of the references of the included papers.

Study Designs and  Included Studies

Of the nine included studies, three were randomized controlled trials, three were controlled retrospective cohort studies, and three were uncontrolled pre-post studies. The key characteristics of each study are described in Table 1.^14-22 The included studies had different definitions for patients who were high utilizers of hospital care. Eight used a "threshold" model that predicted future admissions using past patterns; these studies included patients with at least two admissions over a period of 6 to 12 months, although many had higher thresholds. Zulman et al. used a prediction algorithm to identify patients at risk of future admission. Four studies also included some measure of medical complexity, such as a certain number of chronic medical conditions;^14,17,18,22  in contrast, Sledge et al. excluded the most complex and high-cost patients.²⁰

All studies measured hospital admissions as a primary or a secondary outcome (Table 1). Although all studies demonstrated a reduction in hospital admissions following implementation, those with the greatest reductions did not have a control group.^14,15,17 All three randomized controlled trials showed equal reductions in admission rates between the intervention and control groups.^18,20,22 Among those specifically examining readmissions to the hospital, similar trends emerged, although one study (Plant et al.) found a nonsignificant decrease in hospital readmissions (17% reduction in 24 months, P = .07).¹⁸

In the secondary outcome analysis, six of the nine studies found nonsignificant reductions in ED admissions (Table 1). Four studies measured costs to the hospital or the local hospital system, though none examined costs to patients or payors. Studies estimated cost differently, including the use of estimated hospital costs,^17,20 "facility patient costs" at the VA,²² and a combination of inpatient and ED costs.¹⁹ The latter study (Shah et al., which implemented complex case management services) was the only one to find a statistically significant decrease in mean cost per year pre- and postintervention ($20,298 versus $7,053, P < .001).¹⁹

Only one study measured the quality of life, finding no significant change in summary scores after the intervention compared with controls (93.4 versus 92, P = .32).21 Another study conducted at a VA clinic network found no difference in a patient activation scale following the intervention but found significantly increased satisfaction with overall VA care (3.16 versus 2.90, P = .04).²²

Intervention Characteristics

Intervention characteristics are summarized in Table 2. Although there was heterogeneity in study interventions, we identified common themes. Five of the nine interventions^14-17,22 consisted of interdisciplinary teams that included community health workers, nurses, social workers, and physicians. Physicians were not included on every team; three interventions used them in direct care roles while two others contained physicians as advisors or in indirect roles. Intervention teams also had a variable level of involvement in a patient's care. Mercer et al. developed care plans for patients without physical interaction,¹⁷ whereas Zulman et al. recruited patients to a separate, intensive outpatient clinic outside the usual VA care team structure.²²

The majority of interventions added direct services or support - most commonly, a social worker - to usual care processes. Patient panel sizes were relatively small, with most of the teams recruiting fewer than 150 patients per interdisciplinary team (range, 25-251). There was variation in the length of intervention, from 35 days of case management following hospital discharge to one year of intensive social work support to others of an indefinite length.^15,17,22
Additional common themes included caring for patients across settings and incorporating information technology (IT) into workflows. Four interventions reported either interacting with patients in multiple settings, such as the hospital, clinic, and day hospital, ED, at home, or in the community.^14,19,21,22 Two others^16,20 interacted with patients only in the clinic but expanded the scope of a "traditional" primary care practice to include open scheduling, flexible appointment times, interdisciplinary visits, or outreach. In addition, IT resources assisted seven of the nine interventions, most commonly by identifying eligible patients via an electronic data tracking system or by automated alerts when their patients arrived at affiliated care locations. 

Risk of Study Bias

We systematically assessed the risk of bias of the nine included studies (Appendix B). Using the scale published by Downs and Black, a point-based scale in which a score of 18 denotes a high-quality study, the studies in this review scored 15.55 on average (range 6-22, standard deviation [SD] 5.0). Four of the nine studies met the benchmark for high quality.^12,13,18-22 The risk of bias was highest for measures of internal validity and confounding (range 0-5, mean 2.83, SD 1.94). The risk of bias was lowest for reporting measures (range 0-13, mean 7.40, SD 3.43).

Overall, studies reported mixed results related to readmissions and hospital utilization. While low-quality studies found reductions in hospital use over time, higher quality studies found similar reductions in utilization between the intervention and control groups. Johnson et al. showed that frequent hospitalization rates in a cohort of high-utilizer patients declined naturally over the course of 1-2 years; only 10% of individuals in the initial cohort remained "chronically hospitalized."⁶ Thus, expanding on these findings, the decline in hospitalizations over time as observed in some of the studies included in this review may be due to study patients being identified during a "spike" in utilization, which naturally decreases as the underlying medical or social factors driving rehospitalization resolve. Alternatively, reduction in hospitalizations may represent patients choosing to pursue care at other neighboring hospitals.²³ No study included in our review evaluated healthcare use at institutions other than their study hospital or health system.

A striking theme of this review was the heterogeneity in each study's patient population. Thresholds for "high utilizers" varied from two hospital admissions in six months to two to three admissions in 30 days, to a combination of ED and hospital admissions, and to the use of predictive algorithms. A standard "case definition" for this population could guide future research, enabling comparison of outcomes across settings. Thus, we propose that future studies use three or more hospital admissions within six months when evaluating interventions targeting "high utilizer" patients. Although patients with one prior hospitalization in the past year are at elevated risk of rehospitalization,² we feel that a higher "threshold" for this population will identify those at the highest strata of risk. Although predictive models may be better than "threshold" models, more work in validating these tools needs to be done before these can be put to use across settings.

In contrast to interventions designed to reduce readmissions for heart failure, pneumonia, or other diagnoses, frequently admitted patients do not encompass one disease or pathology pattern. Rinehart et al., in a study characterizing frequently admitted patients across a health system, identified five "subgroups" of patients, including those with (1) unstable housing, (2) comorbid medical and psychiatric illness, (3) severe complex medical illness, (4) dual-diagnosis psychiatric illness and substance abuse, and (5) a combination of medical and psychosocial barriers.25 In light of this population's heterogeneity, interventions may need to be flexible and tailored to the needs of individual patients, while simultaneously accounting for the capabilities and priorities of the health system. More specific and standardized interventions, targeting more homogenous groups, may be appropriate for populations defined according to pathology (such as heart failure or sickle cell disease).²⁷

The components of interventions used for frequently hospitalized patients were diverse. Although most of the studies used interdisciplinary teams, they focused their efforts in a variety of settings, often crossing modern "boundaries of care" by providing direct or indirect input on care across healthcare settings. Care fragmentation probably plays an important role in the risk for readmissions in this population;9 as such, interventions that address factors across the continuum of care may be more likely to succeed.²¹ Notably, six of nine studies were conducted at academic medical centers and an additional one at a VA facility affiliated with an academic center. Only two were located at community-based clinical networks, indicating a theoretical potential for publication bias as academic centers may be more likely to study and publish their work. There may be successful interventions that have not been formally studied or published in the peer-reviewed literature.

The breadth of the outcome measures in the included studies raises questions about what metrics should define success. Although all the studies looked at hospital utilization and readmission, measure definitions varied. Importantly, a minority of studies investigated quality of life and patient satisfaction, outcomes that may ultimately provide a more fertile ground for inquiry and intervention. Two studies looked at quality of life as an outcome,^19,22 but only one found that patients reported increased satisfaction despite showing nonsignificant reductions in hospital use.²² As shown in multiple prior studies, patient engagement is associated with increased satisfaction and can be associated with lower healthcare costs.^26,27 Hibbard et al. have demonstrated that patient activation is a specific component of patient engagement and inversely impacts healthcare cost, with lower levels of patient activation showing increased costs in comparison to those patients more engaged in their own care.²⁷ By focusing on changing patients' perceptions about their own health and involvement in their own care team as a partner, programs may be able to make a greater impact.  

Our systematic review has several limitations. Although we used a search strategy designed to identify all relevant studies, reviewed the references of included studies, and contacted the authors, we identified only nine studies meeting our inclusion criteria. Four of the nine studies were identified from a manual review of references of the included studies, suggesting the possibility of a suboptimal search strategy. Although the inclusion of articles that appear in a check of reference lists is a valid step in the systematic review article acquisition process, we conducted a post hoc investigation of alternate search strategies. We checked the titles, abstracts, and subject headings of the four articles found by reference review to determine whether the original search could have been improved. An analysis of the articles revealed that the terminology used was not consistent with the super user/utilizer terminology we were operating under, and that the four articles used terms such as "high risk" and "complex patients," which are more generic than our targeted terms. Only on a careful read of the abstracts and full-text did we find that these articles were useful to the study. Adjusting the original search to include these general terms would have resulted in an unwieldy set of results; hence, we felt it best to adhere to our original search strategy.

Additional limitations include that only four of the nine included studies were at low risk of bias. In addition to limitations based on study design and small sample sizes, the interventions were often limited to a short period. In light of the multiple factors that contribute to frequent hospitalizations, some of which cannot be addressed quickly, studies to evaluate interventions for longer durations are warranted.  

We found mixed results for the effect of interventions on outcomes for frequently hospitalized patients. While low-quality studies found reductions in hospital use over time, higher quality studies generally found similar reductions in utilization between the intervention and control groups. The range of definitions, interventions, and outcomes used for frequently hospitalized patients is partly explained by the heterogeneity of the population. More rigorous studies using multifaceted interventions that adapt to patients' unique needs should be conducted to assess the effect on outcomes relevant to both providers and patients.

Acknowledgments

The authors would like the thank Dr. Luke Hansen, Dr. Margaret Chapman, and McKay Barra for their support and contributions to this paper and to Northwestern Memorial Hospital's CHAMP (Complex High Admission Management Program).

Disclosures

The authors have nothing to disclose.

Funding

The authors received no funding from external or internal sources for the completion of this project.

In recent years, hospitals and health systems have engaged in considerable efforts to reduce readmissions, in part due to financial incentives from the Medicare Hospital Readmission Reduction Program.^1,2 Though efforts to improve transitions of care for all patients are laudable, risk for readmission is not distributed equally; a small subset of patients accounts for a disproportionate number of hospital readmissions.³ This phenomenon of frequently hospitalized patients is similar to that seen in other populations in which a small proportion of patients account for a majority of healthcare utilization.^3,4

Recognizing that the current system of healthcare delivery does not meet the needs of this population, healthcare organizations have begun to implement interventions that supplement or redesign the system of care for frequently hospitalized patients.^5-7 Descriptive reviews of ambulatory "high-need, high-cost" patients emphasize complex case management and interdisciplinary, team-based care.^8,9 Prior systematic reviews of studies aimed at patients with high use of emergency care demonstrate improvements in social outcomes such as homelessness but mixed results in reducing emergency department (ED) use.¹⁰ However, we were unable to identify any prior reviews that evaluated interventions intended specifically for patients with frequent hospital admissions. Our objective in this systematic review was to characterize interventions for frequently admitted patients and determine whether these interventions decrease use of healthcare resources, improve health outcomes, and/or reduce costs.

Literature Search

We registered our study protocol in the PROSPERO database. A librarian (L.O.) collaboratively developed the search strategies with other review authors (A.G., B.H., N.N.) and in January 2018 ran searches on "super users," "high utilizers," and similar terms in the following databases: PubMed MEDLINE, Embase (embase.com), and Cochrane Central Register of Controlled Trials (CENTRAL) on the Wiley platform. The complete search strategies used are available in Appendix A.  

We attempted to discover additional studies by searching the reference lists of key publications and contacted authors of relevant abstracts to determine whether studies had been published or were planned for peer-reviewed publication. We also contacted authors of included studies to locate additional studies meeting inclusion criteria.

Data Collection Process

Studies were eligible for inclusion in our review if they were (1) published in a peer-reviewed source, (2) defined a study population of patients frequently admitted to inpatient medical services, (3) evaluated an intervention targeting frequently hospitalized patients, and (4) included patients who were >18 years old and (5) admitted as inpatients on medical services. Of note, studies with patients admitted to psychiatric, obstetric, or surgical wards were not included, as the authors did not define these as "general medicine" units. Studies focused solely on an ambulatory population were similarly excluded. Given the heterogeneity of how studies defined frequently hospitalized patients, we did not establish a prespecified number of admissions for inclusion to ensure that we did not exclude interventions not meeting a strict set of criteria. The goal was not to examine interventions to reduce all readmissions, but rather, to look at patients who were recurrently hospitalized. Thus, patients had to be repeatedly admitted, but we let the studies define that usage explicitly.  

Two members of a four-physician team (A.G., B.H., K.O., and N.N.) screened all initial results for eligibility through title and abstract review; potentially relevant articles were retained for full-text review to assess each study's eligibility. If a study's abstract did not clearly indicate whether inclusion criteria were met, we retained the article for full-text review. Two team members (A.G. and B.H.) independently reviewed the full text of each selected article to determine final inclusion in the study. The previously described inclusion criteria were again applied, and a final set of articles was identified for data extraction. Disagreements regarding inclusion in the final review (such as whether a study measured medical or psychiatric hospitalizations) were resolved through discussion among the entire four-physician review team to achieve consensus or, when required, by contacting authors of individual studies.  

Data Abstraction and Risk of Bias Assessment

After selecting the final set of articles, we abstracted data using a tool developed by the Cochrane Effective Practice and Organization of Care Group.¹¹ We then compiled study-level data into a single database for reporting. Extracted elements included study design, setting, patient characteristics, inclusion and exclusion criteria, control group identification, outcome measures, results, and length of follow-up. We also extracted individual characteristics of each intervention, including common intervention elements such as intervention setting, use of health information technology resources, and whether programs developed interdisciplinary care plans. We assessed the risk of bias of each study and the quality of studies using the Downs and Black Scale.^12,13 Two team members (A.G. and B.H.) independently assessed the risk of bias for all nine studies, and differences were resolved by consensus. Due to the variation in the outcomes used, we were unable to conduct a meta-analysis.

Search Results

We found a total of 4,762 references in the three databases. After de-duplication using the EndNote software, there were 3,314 references to screen. We identified 116 studies for full-text review. Of those, we selected nine studies that met the criteria for this study (Figure). The most common reason for exclusion of an article for full-text review was that the patients studied were not defined as high utilizers of inpatient resources and were instead high-utilizers of ambulatory or emergency care (32 studies). We identified five of the included studies through the primary search and four through review of the references of the included papers.

Study Designs and  Included Studies

Of the nine included studies, three were randomized controlled trials, three were controlled retrospective cohort studies, and three were uncontrolled pre-post studies. The key characteristics of each study are described in Table 1.^14-22 The included studies had different definitions for patients who were high utilizers of hospital care. Eight used a "threshold" model that predicted future admissions using past patterns; these studies included patients with at least two admissions over a period of 6 to 12 months, although many had higher thresholds. Zulman et al. used a prediction algorithm to identify patients at risk of future admission. Four studies also included some measure of medical complexity, such as a certain number of chronic medical conditions;^14,17,18,22  in contrast, Sledge et al. excluded the most complex and high-cost patients.²⁰

All studies measured hospital admissions as a primary or a secondary outcome (Table 1). Although all studies demonstrated a reduction in hospital admissions following implementation, those with the greatest reductions did not have a control group.^14,15,17 All three randomized controlled trials showed equal reductions in admission rates between the intervention and control groups.^18,20,22 Among those specifically examining readmissions to the hospital, similar trends emerged, although one study (Plant et al.) found a nonsignificant decrease in hospital readmissions (17% reduction in 24 months, P = .07).¹⁸

In the secondary outcome analysis, six of the nine studies found nonsignificant reductions in ED admissions (Table 1). Four studies measured costs to the hospital or the local hospital system, though none examined costs to patients or payors. Studies estimated cost differently, including the use of estimated hospital costs,^17,20 "facility patient costs" at the VA,²² and a combination of inpatient and ED costs.¹⁹ The latter study (Shah et al., which implemented complex case management services) was the only one to find a statistically significant decrease in mean cost per year pre- and postintervention ($20,298 versus $7,053, P < .001).¹⁹

Only one study measured the quality of life, finding no significant change in summary scores after the intervention compared with controls (93.4 versus 92, P = .32).21 Another study conducted at a VA clinic network found no difference in a patient activation scale following the intervention but found significantly increased satisfaction with overall VA care (3.16 versus 2.90, P = .04).²²

Intervention Characteristics

Intervention characteristics are summarized in Table 2. Although there was heterogeneity in study interventions, we identified common themes. Five of the nine interventions^14-17,22 consisted of interdisciplinary teams that included community health workers, nurses, social workers, and physicians. Physicians were not included on every team; three interventions used them in direct care roles while two others contained physicians as advisors or in indirect roles. Intervention teams also had a variable level of involvement in a patient's care. Mercer et al. developed care plans for patients without physical interaction,¹⁷ whereas Zulman et al. recruited patients to a separate, intensive outpatient clinic outside the usual VA care team structure.²²

The majority of interventions added direct services or support - most commonly, a social worker - to usual care processes. Patient panel sizes were relatively small, with most of the teams recruiting fewer than 150 patients per interdisciplinary team (range, 25-251). There was variation in the length of intervention, from 35 days of case management following hospital discharge to one year of intensive social work support to others of an indefinite length.^15,17,22
Additional common themes included caring for patients across settings and incorporating information technology (IT) into workflows. Four interventions reported either interacting with patients in multiple settings, such as the hospital, clinic, and day hospital, ED, at home, or in the community.^14,19,21,22 Two others^16,20 interacted with patients only in the clinic but expanded the scope of a "traditional" primary care practice to include open scheduling, flexible appointment times, interdisciplinary visits, or outreach. In addition, IT resources assisted seven of the nine interventions, most commonly by identifying eligible patients via an electronic data tracking system or by automated alerts when their patients arrived at affiliated care locations. 

Risk of Study Bias

We systematically assessed the risk of bias of the nine included studies (Appendix B). Using the scale published by Downs and Black, a point-based scale in which a score of 18 denotes a high-quality study, the studies in this review scored 15.55 on average (range 6-22, standard deviation [SD] 5.0). Four of the nine studies met the benchmark for high quality.^12,13,18-22 The risk of bias was highest for measures of internal validity and confounding (range 0-5, mean 2.83, SD 1.94). The risk of bias was lowest for reporting measures (range 0-13, mean 7.40, SD 3.43).

Overall, studies reported mixed results related to readmissions and hospital utilization. While low-quality studies found reductions in hospital use over time, higher quality studies found similar reductions in utilization between the intervention and control groups. Johnson et al. showed that frequent hospitalization rates in a cohort of high-utilizer patients declined naturally over the course of 1-2 years; only 10% of individuals in the initial cohort remained "chronically hospitalized."⁶ Thus, expanding on these findings, the decline in hospitalizations over time as observed in some of the studies included in this review may be due to study patients being identified during a "spike" in utilization, which naturally decreases as the underlying medical or social factors driving rehospitalization resolve. Alternatively, reduction in hospitalizations may represent patients choosing to pursue care at other neighboring hospitals.²³ No study included in our review evaluated healthcare use at institutions other than their study hospital or health system.

A striking theme of this review was the heterogeneity in each study's patient population. Thresholds for "high utilizers" varied from two hospital admissions in six months to two to three admissions in 30 days, to a combination of ED and hospital admissions, and to the use of predictive algorithms. A standard "case definition" for this population could guide future research, enabling comparison of outcomes across settings. Thus, we propose that future studies use three or more hospital admissions within six months when evaluating interventions targeting "high utilizer" patients. Although patients with one prior hospitalization in the past year are at elevated risk of rehospitalization,² we feel that a higher "threshold" for this population will identify those at the highest strata of risk. Although predictive models may be better than "threshold" models, more work in validating these tools needs to be done before these can be put to use across settings.

In contrast to interventions designed to reduce readmissions for heart failure, pneumonia, or other diagnoses, frequently admitted patients do not encompass one disease or pathology pattern. Rinehart et al., in a study characterizing frequently admitted patients across a health system, identified five "subgroups" of patients, including those with (1) unstable housing, (2) comorbid medical and psychiatric illness, (3) severe complex medical illness, (4) dual-diagnosis psychiatric illness and substance abuse, and (5) a combination of medical and psychosocial barriers.25 In light of this population's heterogeneity, interventions may need to be flexible and tailored to the needs of individual patients, while simultaneously accounting for the capabilities and priorities of the health system. More specific and standardized interventions, targeting more homogenous groups, may be appropriate for populations defined according to pathology (such as heart failure or sickle cell disease).²⁷

The components of interventions used for frequently hospitalized patients were diverse. Although most of the studies used interdisciplinary teams, they focused their efforts in a variety of settings, often crossing modern "boundaries of care" by providing direct or indirect input on care across healthcare settings. Care fragmentation probably plays an important role in the risk for readmissions in this population;9 as such, interventions that address factors across the continuum of care may be more likely to succeed.²¹ Notably, six of nine studies were conducted at academic medical centers and an additional one at a VA facility affiliated with an academic center. Only two were located at community-based clinical networks, indicating a theoretical potential for publication bias as academic centers may be more likely to study and publish their work. There may be successful interventions that have not been formally studied or published in the peer-reviewed literature.

The breadth of the outcome measures in the included studies raises questions about what metrics should define success. Although all the studies looked at hospital utilization and readmission, measure definitions varied. Importantly, a minority of studies investigated quality of life and patient satisfaction, outcomes that may ultimately provide a more fertile ground for inquiry and intervention. Two studies looked at quality of life as an outcome,^19,22 but only one found that patients reported increased satisfaction despite showing nonsignificant reductions in hospital use.²² As shown in multiple prior studies, patient engagement is associated with increased satisfaction and can be associated with lower healthcare costs.^26,27 Hibbard et al. have demonstrated that patient activation is a specific component of patient engagement and inversely impacts healthcare cost, with lower levels of patient activation showing increased costs in comparison to those patients more engaged in their own care.²⁷ By focusing on changing patients' perceptions about their own health and involvement in their own care team as a partner, programs may be able to make a greater impact.  

Our systematic review has several limitations. Although we used a search strategy designed to identify all relevant studies, reviewed the references of included studies, and contacted the authors, we identified only nine studies meeting our inclusion criteria. Four of the nine studies were identified from a manual review of references of the included studies, suggesting the possibility of a suboptimal search strategy. Although the inclusion of articles that appear in a check of reference lists is a valid step in the systematic review article acquisition process, we conducted a post hoc investigation of alternate search strategies. We checked the titles, abstracts, and subject headings of the four articles found by reference review to determine whether the original search could have been improved. An analysis of the articles revealed that the terminology used was not consistent with the super user/utilizer terminology we were operating under, and that the four articles used terms such as "high risk" and "complex patients," which are more generic than our targeted terms. Only on a careful read of the abstracts and full-text did we find that these articles were useful to the study. Adjusting the original search to include these general terms would have resulted in an unwieldy set of results; hence, we felt it best to adhere to our original search strategy.

Additional limitations include that only four of the nine included studies were at low risk of bias. In addition to limitations based on study design and small sample sizes, the interventions were often limited to a short period. In light of the multiple factors that contribute to frequent hospitalizations, some of which cannot be addressed quickly, studies to evaluate interventions for longer durations are warranted.  

We found mixed results for the effect of interventions on outcomes for frequently hospitalized patients. While low-quality studies found reductions in hospital use over time, higher quality studies generally found similar reductions in utilization between the intervention and control groups. The range of definitions, interventions, and outcomes used for frequently hospitalized patients is partly explained by the heterogeneity of the population. More rigorous studies using multifaceted interventions that adapt to patients' unique needs should be conducted to assess the effect on outcomes relevant to both providers and patients.

Acknowledgments

The authors would like the thank Dr. Luke Hansen, Dr. Margaret Chapman, and McKay Barra for their support and contributions to this paper and to Northwestern Memorial Hospital's CHAMP (Complex High Admission Management Program).

Disclosures

The authors have nothing to disclose.

Funding

The authors received no funding from external or internal sources for the completion of this project.

In recent years, hospitals and health systems have engaged in considerable efforts to reduce readmissions, in part due to financial incentives from the Medicare Hospital Readmission Reduction Program.^1,2 Though efforts to improve transitions of care for all patients are laudable, risk for readmission is not distributed equally; a small subset of patients accounts for a disproportionate number of hospital readmissions.³ This phenomenon of frequently hospitalized patients is similar to that seen in other populations in which a small proportion of patients account for a majority of healthcare utilization.^3,4

Recognizing that the current system of healthcare delivery does not meet the needs of this population, healthcare organizations have begun to implement interventions that supplement or redesign the system of care for frequently hospitalized patients.^5-7 Descriptive reviews of ambulatory "high-need, high-cost" patients emphasize complex case management and interdisciplinary, team-based care.^8,9 Prior systematic reviews of studies aimed at patients with high use of emergency care demonstrate improvements in social outcomes such as homelessness but mixed results in reducing emergency department (ED) use.¹⁰ However, we were unable to identify any prior reviews that evaluated interventions intended specifically for patients with frequent hospital admissions. Our objective in this systematic review was to characterize interventions for frequently admitted patients and determine whether these interventions decrease use of healthcare resources, improve health outcomes, and/or reduce costs.

Literature Search

We registered our study protocol in the PROSPERO database. A librarian (L.O.) collaboratively developed the search strategies with other review authors (A.G., B.H., N.N.) and in January 2018 ran searches on "super users," "high utilizers," and similar terms in the following databases: PubMed MEDLINE, Embase (embase.com), and Cochrane Central Register of Controlled Trials (CENTRAL) on the Wiley platform. The complete search strategies used are available in Appendix A.  

We attempted to discover additional studies by searching the reference lists of key publications and contacted authors of relevant abstracts to determine whether studies had been published or were planned for peer-reviewed publication. We also contacted authors of included studies to locate additional studies meeting inclusion criteria.

Data Collection Process

Studies were eligible for inclusion in our review if they were (1) published in a peer-reviewed source, (2) defined a study population of patients frequently admitted to inpatient medical services, (3) evaluated an intervention targeting frequently hospitalized patients, and (4) included patients who were >18 years old and (5) admitted as inpatients on medical services. Of note, studies with patients admitted to psychiatric, obstetric, or surgical wards were not included, as the authors did not define these as "general medicine" units. Studies focused solely on an ambulatory population were similarly excluded. Given the heterogeneity of how studies defined frequently hospitalized patients, we did not establish a prespecified number of admissions for inclusion to ensure that we did not exclude interventions not meeting a strict set of criteria. The goal was not to examine interventions to reduce all readmissions, but rather, to look at patients who were recurrently hospitalized. Thus, patients had to be repeatedly admitted, but we let the studies define that usage explicitly.  

Two members of a four-physician team (A.G., B.H., K.O., and N.N.) screened all initial results for eligibility through title and abstract review; potentially relevant articles were retained for full-text review to assess each study's eligibility. If a study's abstract did not clearly indicate whether inclusion criteria were met, we retained the article for full-text review. Two team members (A.G. and B.H.) independently reviewed the full text of each selected article to determine final inclusion in the study. The previously described inclusion criteria were again applied, and a final set of articles was identified for data extraction. Disagreements regarding inclusion in the final review (such as whether a study measured medical or psychiatric hospitalizations) were resolved through discussion among the entire four-physician review team to achieve consensus or, when required, by contacting authors of individual studies.  

Data Abstraction and Risk of Bias Assessment

After selecting the final set of articles, we abstracted data using a tool developed by the Cochrane Effective Practice and Organization of Care Group.¹¹ We then compiled study-level data into a single database for reporting. Extracted elements included study design, setting, patient characteristics, inclusion and exclusion criteria, control group identification, outcome measures, results, and length of follow-up. We also extracted individual characteristics of each intervention, including common intervention elements such as intervention setting, use of health information technology resources, and whether programs developed interdisciplinary care plans. We assessed the risk of bias of each study and the quality of studies using the Downs and Black Scale.^12,13 Two team members (A.G. and B.H.) independently assessed the risk of bias for all nine studies, and differences were resolved by consensus. Due to the variation in the outcomes used, we were unable to conduct a meta-analysis.

Search Results

We found a total of 4,762 references in the three databases. After de-duplication using the EndNote software, there were 3,314 references to screen. We identified 116 studies for full-text review. Of those, we selected nine studies that met the criteria for this study (Figure). The most common reason for exclusion of an article for full-text review was that the patients studied were not defined as high utilizers of inpatient resources and were instead high-utilizers of ambulatory or emergency care (32 studies). We identified five of the included studies through the primary search and four through review of the references of the included papers.

Study Designs and  Included Studies

Of the nine included studies, three were randomized controlled trials, three were controlled retrospective cohort studies, and three were uncontrolled pre-post studies. The key characteristics of each study are described in Table 1.^14-22 The included studies had different definitions for patients who were high utilizers of hospital care. Eight used a "threshold" model that predicted future admissions using past patterns; these studies included patients with at least two admissions over a period of 6 to 12 months, although many had higher thresholds. Zulman et al. used a prediction algorithm to identify patients at risk of future admission. Four studies also included some measure of medical complexity, such as a certain number of chronic medical conditions;^14,17,18,22  in contrast, Sledge et al. excluded the most complex and high-cost patients.²⁰

All studies measured hospital admissions as a primary or a secondary outcome (Table 1). Although all studies demonstrated a reduction in hospital admissions following implementation, those with the greatest reductions did not have a control group.^14,15,17 All three randomized controlled trials showed equal reductions in admission rates between the intervention and control groups.^18,20,22 Among those specifically examining readmissions to the hospital, similar trends emerged, although one study (Plant et al.) found a nonsignificant decrease in hospital readmissions (17% reduction in 24 months, P = .07).¹⁸

In the secondary outcome analysis, six of the nine studies found nonsignificant reductions in ED admissions (Table 1). Four studies measured costs to the hospital or the local hospital system, though none examined costs to patients or payors. Studies estimated cost differently, including the use of estimated hospital costs,^17,20 "facility patient costs" at the VA,²² and a combination of inpatient and ED costs.¹⁹ The latter study (Shah et al., which implemented complex case management services) was the only one to find a statistically significant decrease in mean cost per year pre- and postintervention ($20,298 versus $7,053, P < .001).¹⁹

Only one study measured the quality of life, finding no significant change in summary scores after the intervention compared with controls (93.4 versus 92, P = .32).21 Another study conducted at a VA clinic network found no difference in a patient activation scale following the intervention but found significantly increased satisfaction with overall VA care (3.16 versus 2.90, P = .04).²²

Intervention Characteristics

Intervention characteristics are summarized in Table 2. Although there was heterogeneity in study interventions, we identified common themes. Five of the nine interventions^14-17,22 consisted of interdisciplinary teams that included community health workers, nurses, social workers, and physicians. Physicians were not included on every team; three interventions used them in direct care roles while two others contained physicians as advisors or in indirect roles. Intervention teams also had a variable level of involvement in a patient's care. Mercer et al. developed care plans for patients without physical interaction,¹⁷ whereas Zulman et al. recruited patients to a separate, intensive outpatient clinic outside the usual VA care team structure.²²

The majority of interventions added direct services or support - most commonly, a social worker - to usual care processes. Patient panel sizes were relatively small, with most of the teams recruiting fewer than 150 patients per interdisciplinary team (range, 25-251). There was variation in the length of intervention, from 35 days of case management following hospital discharge to one year of intensive social work support to others of an indefinite length.^15,17,22
Additional common themes included caring for patients across settings and incorporating information technology (IT) into workflows. Four interventions reported either interacting with patients in multiple settings, such as the hospital, clinic, and day hospital, ED, at home, or in the community.^14,19,21,22 Two others^16,20 interacted with patients only in the clinic but expanded the scope of a "traditional" primary care practice to include open scheduling, flexible appointment times, interdisciplinary visits, or outreach. In addition, IT resources assisted seven of the nine interventions, most commonly by identifying eligible patients via an electronic data tracking system or by automated alerts when their patients arrived at affiliated care locations. 

Risk of Study Bias

We systematically assessed the risk of bias of the nine included studies (Appendix B). Using the scale published by Downs and Black, a point-based scale in which a score of 18 denotes a high-quality study, the studies in this review scored 15.55 on average (range 6-22, standard deviation [SD] 5.0). Four of the nine studies met the benchmark for high quality.^12,13,18-22 The risk of bias was highest for measures of internal validity and confounding (range 0-5, mean 2.83, SD 1.94). The risk of bias was lowest for reporting measures (range 0-13, mean 7.40, SD 3.43).

Overall, studies reported mixed results related to readmissions and hospital utilization. While low-quality studies found reductions in hospital use over time, higher quality studies found similar reductions in utilization between the intervention and control groups. Johnson et al. showed that frequent hospitalization rates in a cohort of high-utilizer patients declined naturally over the course of 1-2 years; only 10% of individuals in the initial cohort remained "chronically hospitalized."⁶ Thus, expanding on these findings, the decline in hospitalizations over time as observed in some of the studies included in this review may be due to study patients being identified during a "spike" in utilization, which naturally decreases as the underlying medical or social factors driving rehospitalization resolve. Alternatively, reduction in hospitalizations may represent patients choosing to pursue care at other neighboring hospitals.²³ No study included in our review evaluated healthcare use at institutions other than their study hospital or health system.

A striking theme of this review was the heterogeneity in each study's patient population. Thresholds for "high utilizers" varied from two hospital admissions in six months to two to three admissions in 30 days, to a combination of ED and hospital admissions, and to the use of predictive algorithms. A standard "case definition" for this population could guide future research, enabling comparison of outcomes across settings. Thus, we propose that future studies use three or more hospital admissions within six months when evaluating interventions targeting "high utilizer" patients. Although patients with one prior hospitalization in the past year are at elevated risk of rehospitalization,² we feel that a higher "threshold" for this population will identify those at the highest strata of risk. Although predictive models may be better than "threshold" models, more work in validating these tools needs to be done before these can be put to use across settings.

In contrast to interventions designed to reduce readmissions for heart failure, pneumonia, or other diagnoses, frequently admitted patients do not encompass one disease or pathology pattern. Rinehart et al., in a study characterizing frequently admitted patients across a health system, identified five "subgroups" of patients, including those with (1) unstable housing, (2) comorbid medical and psychiatric illness, (3) severe complex medical illness, (4) dual-diagnosis psychiatric illness and substance abuse, and (5) a combination of medical and psychosocial barriers.25 In light of this population's heterogeneity, interventions may need to be flexible and tailored to the needs of individual patients, while simultaneously accounting for the capabilities and priorities of the health system. More specific and standardized interventions, targeting more homogenous groups, may be appropriate for populations defined according to pathology (such as heart failure or sickle cell disease).²⁷

The components of interventions used for frequently hospitalized patients were diverse. Although most of the studies used interdisciplinary teams, they focused their efforts in a variety of settings, often crossing modern "boundaries of care" by providing direct or indirect input on care across healthcare settings. Care fragmentation probably plays an important role in the risk for readmissions in this population;9 as such, interventions that address factors across the continuum of care may be more likely to succeed.²¹ Notably, six of nine studies were conducted at academic medical centers and an additional one at a VA facility affiliated with an academic center. Only two were located at community-based clinical networks, indicating a theoretical potential for publication bias as academic centers may be more likely to study and publish their work. There may be successful interventions that have not been formally studied or published in the peer-reviewed literature.

The breadth of the outcome measures in the included studies raises questions about what metrics should define success. Although all the studies looked at hospital utilization and readmission, measure definitions varied. Importantly, a minority of studies investigated quality of life and patient satisfaction, outcomes that may ultimately provide a more fertile ground for inquiry and intervention. Two studies looked at quality of life as an outcome,^19,22 but only one found that patients reported increased satisfaction despite showing nonsignificant reductions in hospital use.²² As shown in multiple prior studies, patient engagement is associated with increased satisfaction and can be associated with lower healthcare costs.^26,27 Hibbard et al. have demonstrated that patient activation is a specific component of patient engagement and inversely impacts healthcare cost, with lower levels of patient activation showing increased costs in comparison to those patients more engaged in their own care.²⁷ By focusing on changing patients' perceptions about their own health and involvement in their own care team as a partner, programs may be able to make a greater impact.  

Our systematic review has several limitations. Although we used a search strategy designed to identify all relevant studies, reviewed the references of included studies, and contacted the authors, we identified only nine studies meeting our inclusion criteria. Four of the nine studies were identified from a manual review of references of the included studies, suggesting the possibility of a suboptimal search strategy. Although the inclusion of articles that appear in a check of reference lists is a valid step in the systematic review article acquisition process, we conducted a post hoc investigation of alternate search strategies. We checked the titles, abstracts, and subject headings of the four articles found by reference review to determine whether the original search could have been improved. An analysis of the articles revealed that the terminology used was not consistent with the super user/utilizer terminology we were operating under, and that the four articles used terms such as "high risk" and "complex patients," which are more generic than our targeted terms. Only on a careful read of the abstracts and full-text did we find that these articles were useful to the study. Adjusting the original search to include these general terms would have resulted in an unwieldy set of results; hence, we felt it best to adhere to our original search strategy.

Additional limitations include that only four of the nine included studies were at low risk of bias. In addition to limitations based on study design and small sample sizes, the interventions were often limited to a short period. In light of the multiple factors that contribute to frequent hospitalizations, some of which cannot be addressed quickly, studies to evaluate interventions for longer durations are warranted.  

We found mixed results for the effect of interventions on outcomes for frequently hospitalized patients. While low-quality studies found reductions in hospital use over time, higher quality studies generally found similar reductions in utilization between the intervention and control groups. The range of definitions, interventions, and outcomes used for frequently hospitalized patients is partly explained by the heterogeneity of the population. More rigorous studies using multifaceted interventions that adapt to patients' unique needs should be conducted to assess the effect on outcomes relevant to both providers and patients.

Acknowledgments

The authors would like the thank Dr. Luke Hansen, Dr. Margaret Chapman, and McKay Barra for their support and contributions to this paper and to Northwestern Memorial Hospital's CHAMP (Complex High Admission Management Program).

Disclosures

The authors have nothing to disclose.

Funding

The authors received no funding from external or internal sources for the completion of this project.

Botulinum toxin, a potent neurotoxic protein produced by the bacterium Clostridium botulinum, has been used as treatment for a variety of medical indications for more than 25 years (Box^1-12). Recently, researchers have been exploring the role of botulinum toxin in psychiatry, primarily as an adjunctive treatment for depression, but also for several other possible indications. Several studies, including randomized controlled trials (RCTs), have provided evidence that glabellar botulinum toxin injections may be a safe and effective treatment for depression. In this article, we provide an update on the latest clinical trials that evaluated botulinum toxin for depression, and also summarize the evidence regarding other potential clinical psychiatric applications of botulinum toxin.

Several RCTs suggest efficacy for depression

The use of botulinum toxin to treat depression is based on the facial feedback hypothesis, which was first proposed by Charles Darwin in 1872¹³ and further elaborated by William James,^14,15 who emphasized the importance of the sensation of bodily changes in emotion. Contrary to the popular belief that emotions trigger physiological changes in the body, James postulated that peripheral bodily changes secondary to stimuli perception would exert a sensory feedback, generating emotions. The manipulation of human facial expression with an expression that is associated with a particular emotion (eg, holding a pen with teeth, leading to risorius/zygomaticus muscles contraction and a smile simulation) was found to influence participants’ affective responses in the presence of emotional stimuli (eg, rating cartoons as funnier), reinforcing the facial-feedback hypothesis.^16,17

From a neurobiologic standpoint, facial botulinum toxin A (BTA) injections in rats were associated with increased serotonin and norepinephrine concentrations in the hypothalamus and striatum, respectively.¹⁸ Moreover, amygdala activity in response to angry vs happy faces, measured via functional magnetic resonance imaging (fMRI), was found to be attenuated after BTA applications to muscles involved in angry facial expressions.^19,20 Both the neurotransmitters as well as the aforementioned brain regions have been implicated in the pathophysiology of depression.^21,22

Box

Continue to: In a second RCT involving 74 patients with depression...

Continue to: Benefits for other psychiatric indications

Borderline personality disorder. In a case series of 6 women, BTA injections in the glabellar region were reported to be particularly effective for the treatment of borderline personality disorder symptoms that were resistant to psychotherapy and pharmacotherapy.³¹ Two to 6 weeks after a 29-unit injection, borderline personality disorder symptoms as measured by the Zanarini Rating Scale for Borderline Personality Disorder and/or the Borderline Symptom List were shown to significantly improve by 49% to 94% from baseline (P ≤ .05). These findings emphasize the promising therapeutic role of BTA on depressive symptoms concomitant with the emotional lability, impulsivity, and negative emotions that usually characterize this personality disorder.^31,32 A small sample size and lack of a placebo comparator are limitations of this research.

Neuroleptic-induced sialorrhea. Botulinum toxin injections in the salivary glands have been investigated for treating clozapine-induced sialorrhea because they are thought to directly inhibit the release of acetylcholine from salivary glands. One small RCT that used botulinum toxin B (BTB)³³ and 1 case report that used BTA³⁴ reported successful reduction in hypersalivation, with doses ranging from 150 to 500 units injected in each of the parotid and/or submandibular glands bilaterally. Although the treatment was well tolerated and lasted up to 16 weeks, larger studies are needed to replicate these findings.^33-35

Orofacial tardive dyskinesia. Several case reports of orofacial tardive dyskinesia, including lingual dyskinesia and lingual protrusion dystonia, have found improvements in hyperkinetic movements following muscular BTA injections, such as in the genioglossus muscle in the case of tongue involvement.^36-39 These cases were, however, described in the literature before the recent FDA approval of the vesicular monoamine transporter 2 inhibitors valbenazine and deutetrabenazine for the treatment of tardive dyskinesia.^40,41

Studies examining botulinum toxin’s application in areas of psychiatry other than depression are summarized in Table 2.^31,33,36-38

Continue to: Promising initial findings but multiple limitations

Although BTA injections have been explored as a potential treatment for several psychiatric conditions, the bulk of recent evidence is derived from studies in patients with depressive disorders. BTA injections in the glabellar regions have been shown in small RCTs to be well-tolerated with overall promising improvement of depressive symptoms, optimally 6 weeks after a single injection. Moreover, BTA has been shown to be safe and long-lasting, which would be convenient for patients and might improve adherence to therapy.^42-44 BTA’s antidepressant effects were shown to be independent of frown line severity or patient satisfaction with cosmetic effects.⁴⁵ The trials by Wollmer et al,²³ Finzi and Rosenthal,²⁵ and Magid et al²⁶ mainly studied BTA as an adjunctive treatment to antidepressants in patients with ongoing unipolar depression. However, Finzi and Rosenthal²⁵ included patients who were not medicated at the time of the study.

Pooled analysis of these 3 RCTs found that patients who received BTA monotherapy improved equally to those who received it as an adjunctive treatment to antidepressants. Overall, on primary endpoint measures, a response rate of 54.2% was obtained in the BTA group compared with 10.7% among patients who received placebo saline injections (odds ratio [OR] 11.1, 95% confidence interval [CI], 4.3 to 28.8, number needed to treat [NNT] = 2.3) and a remission rate of 30.5% with BTA compared with 6.7% with placebo (OR 7.3, 95 % CI, 2.4 to 22.5, NNT = 4.2).⁴⁶ However, remission rates tend to be higher in the augmentation groups, and so further studies are needed to compare both treatment strategies.

Nevertheless, these positive findings have been recently challenged by the results of the largest U.S. multicenter phase II RCT,²⁸ which failed to find a significant antidepressant effect at 6 weeks with the 30-unit BTA injection, and also failed to prove a dose-effect relationship, as the 50-unit injection wasn’t superior to the lower dose and didn’t significantly differ from placebo. One hypothesis to explain this discrepancy may be the difference in injection sites between the treatment and placebo groups.⁴⁷ Future studies need to address the various limitations of earlier clinical trials that mainly yielded promising results with BTA.

A major concern is the high rate of unblinding of participants and researchers in BTA trials, as the cosmetic effects of botulinum toxin injections make them easy to distinguish from saline injections. Ninety percent of participants in the Wollmer et al study²³ were able to correctly guess their group allocation, while 60% of evaluators guessed correctly. Finzi and Rozenthal²⁵ reported 52% of participants in the BTA group, 46% in the placebo group, and 73% of evaluators correctly guessed their allocation. Magid et al²⁶ reported 75% of participants were able to guess the order of intervention they received. The high unblinding rates in these trials remains a significant limitation. There is a concern that this may lead to an underestimation of the placebo effect relative to clinical improvement, thus causing inflation of outcome differences between groups. Although various methods have been tried to minimize evaluator unblinding, such as placing surgical caps on participants’ faces during visits to hide the glabellar region, better methods need to be implemented to prevent unblinding of both raters and participants.

Furthermore, except for the multicenter phase II trial, most studies have been conducted in small samples, which limits their statistical power. Larger controlled trials will be needed to replicate the positive findings obtained in smaller RCTs.

Another limitation is that the majority of the well-designed RCTs were conducted in populations that were predominantly female, which makes it difficult to reliably assess treatment efficacy in men. This may be because cosmetic treatment with botulinum toxin injection is more favorably received by women than by men. A recent comparison⁴⁸ of the studies by Wollmer et al²³ and Finzi and Rosenthal²⁵ discussed an interesting observation. Wollmer et al did not explicitly mention botulinum toxin when recruiting for the study, while Finzi and Rosenthal did. While approximately a quarter of the participants in the Wollmer et al study were male, Finzi and Rosenthal attracted an almost entirely female population. Perhaps there is a potential bias for females to be more attracted to these studies due to the secondary gain of receiving a cosmetic procedure.

In an attempt to understand predictors of positive response to botulinum toxin in patients with depression, Wollmer et al⁴⁹ conducted a follow-up study in which they reassessed the data obtained from their initial RCT using the HAM-D agitation item scores to separate the 15 participants who received BTA into low-agitation (≤1 score on agitation item of the HAM-D scale) and high-agitation (≥2 score on agitation item of the HAM-D scale) groups. They found that the 9 participants who responded to BTA treatment had significantly higher baseline agitation scores than participants who did not respond (1.56 ± 0.88 vs 0.33 ± 0.52, P = .01). All of the participants who presented with higher agitation levels experienced response, compared with 40% of those with lower agitation levels (P = .04), although there was no significant difference in magnitude of improvement (14.2 ± 1.92 vs 8.0 ± 9.37, P = .07). The study added additional support to the facial feedback hypothesis, as it links the improvement of depression to facial muscle activation targeted by the injections. It also introduced a potential predictor of response to botulinum toxin treatment, highlighting potential factors to consider when enrolling patients in future investigations.

The case series of patients with borderline personality disorder³¹ also shed light on the potential positive effect of BTA treatment for a particular subtype of patients with depression—those with comorbid emotional instability—to consider as a therapeutic target for the future. Hence, inclusion criteria for future trials might potentially include patient age, gender, existence/quantification of prominent frown lines at baseline, severity of MDD, duration of depression, and personality characteristics of enrolled participants.

In conclusion, BTA injections appear promising as a treatment for depression as well as for other psychiatric disorders. Future studies should focus on identifying optimal candidates for this innovative treatment modality. Furthermore, BTA dosing and administration strategies (monotherapy vs adjunctive treatment to antidepressants) need to be further explored. As retrograde axonal transport of botulinum toxin has been demonstrated in animal studies, it would be interesting to further examine its effects in the human CNS to enhance our knowledge of the pathophysiology of botulinum and its potential applications in psychiatry.⁵⁰

Bottom Line

Botulinum toxin shows promising antidepressant effects and may have a role in the treatment of several other psychiatric disorders. More research is needed to address limitations of previous studies and to establish an adequate treatment regimen.

Related Resources

• Wollmer MA, Magid M, Kruger TH. Botulinum toxin treatment in depression. In: Bewley A, Taylor RE, Reichenberg JS, et al (eds). Practical psychodermatology. Oxford, UK: Wiley; 2014.
• Wollmer MA, Neumann I, Magid M. et al. Shrink that frown! Botulinum toxin therapy is lifting the face of psychiatry. G Ital Dermatol Venereol. 2018;153(4):540-548.

Drug Brand Names

Alprazolam • Xanax
Aripiprazole • Abilify
Biperiden • Akineton
Botulinum toxin A • Botox
Botulinum toxin B • Myobloc
Clozapine • Clozaril
Deutetrabenazine • Austedo
Flupentixol • Prolixin
Imipramine • Tofranil
Olanzapine • Zyprexa
Reserpine • Serpalan, Serpasil
Tetrabenazine • Xenazine
Valbenazine • Ingrezza
Ziprasidone • Geodon

Botulinum toxin, a potent neurotoxic protein produced by the bacterium Clostridium botulinum, has been used as treatment for a variety of medical indications for more than 25 years (Box^1-12). Recently, researchers have been exploring the role of botulinum toxin in psychiatry, primarily as an adjunctive treatment for depression, but also for several other possible indications. Several studies, including randomized controlled trials (RCTs), have provided evidence that glabellar botulinum toxin injections may be a safe and effective treatment for depression. In this article, we provide an update on the latest clinical trials that evaluated botulinum toxin for depression, and also summarize the evidence regarding other potential clinical psychiatric applications of botulinum toxin.

Several RCTs suggest efficacy for depression

The use of botulinum toxin to treat depression is based on the facial feedback hypothesis, which was first proposed by Charles Darwin in 1872¹³ and further elaborated by William James,^14,15 who emphasized the importance of the sensation of bodily changes in emotion. Contrary to the popular belief that emotions trigger physiological changes in the body, James postulated that peripheral bodily changes secondary to stimuli perception would exert a sensory feedback, generating emotions. The manipulation of human facial expression with an expression that is associated with a particular emotion (eg, holding a pen with teeth, leading to risorius/zygomaticus muscles contraction and a smile simulation) was found to influence participants’ affective responses in the presence of emotional stimuli (eg, rating cartoons as funnier), reinforcing the facial-feedback hypothesis.^16,17

From a neurobiologic standpoint, facial botulinum toxin A (BTA) injections in rats were associated with increased serotonin and norepinephrine concentrations in the hypothalamus and striatum, respectively.¹⁸ Moreover, amygdala activity in response to angry vs happy faces, measured via functional magnetic resonance imaging (fMRI), was found to be attenuated after BTA applications to muscles involved in angry facial expressions.^19,20 Both the neurotransmitters as well as the aforementioned brain regions have been implicated in the pathophysiology of depression.^21,22

Box

Continue to: In a second RCT involving 74 patients with depression...

Continue to: Benefits for other psychiatric indications

Borderline personality disorder. In a case series of 6 women, BTA injections in the glabellar region were reported to be particularly effective for the treatment of borderline personality disorder symptoms that were resistant to psychotherapy and pharmacotherapy.³¹ Two to 6 weeks after a 29-unit injection, borderline personality disorder symptoms as measured by the Zanarini Rating Scale for Borderline Personality Disorder and/or the Borderline Symptom List were shown to significantly improve by 49% to 94% from baseline (P ≤ .05). These findings emphasize the promising therapeutic role of BTA on depressive symptoms concomitant with the emotional lability, impulsivity, and negative emotions that usually characterize this personality disorder.^31,32 A small sample size and lack of a placebo comparator are limitations of this research.

Neuroleptic-induced sialorrhea. Botulinum toxin injections in the salivary glands have been investigated for treating clozapine-induced sialorrhea because they are thought to directly inhibit the release of acetylcholine from salivary glands. One small RCT that used botulinum toxin B (BTB)³³ and 1 case report that used BTA³⁴ reported successful reduction in hypersalivation, with doses ranging from 150 to 500 units injected in each of the parotid and/or submandibular glands bilaterally. Although the treatment was well tolerated and lasted up to 16 weeks, larger studies are needed to replicate these findings.^33-35

Orofacial tardive dyskinesia. Several case reports of orofacial tardive dyskinesia, including lingual dyskinesia and lingual protrusion dystonia, have found improvements in hyperkinetic movements following muscular BTA injections, such as in the genioglossus muscle in the case of tongue involvement.^36-39 These cases were, however, described in the literature before the recent FDA approval of the vesicular monoamine transporter 2 inhibitors valbenazine and deutetrabenazine for the treatment of tardive dyskinesia.^40,41

Studies examining botulinum toxin’s application in areas of psychiatry other than depression are summarized in Table 2.^31,33,36-38

Continue to: Promising initial findings but multiple limitations

Although BTA injections have been explored as a potential treatment for several psychiatric conditions, the bulk of recent evidence is derived from studies in patients with depressive disorders. BTA injections in the glabellar regions have been shown in small RCTs to be well-tolerated with overall promising improvement of depressive symptoms, optimally 6 weeks after a single injection. Moreover, BTA has been shown to be safe and long-lasting, which would be convenient for patients and might improve adherence to therapy.^42-44 BTA’s antidepressant effects were shown to be independent of frown line severity or patient satisfaction with cosmetic effects.⁴⁵ The trials by Wollmer et al,²³ Finzi and Rosenthal,²⁵ and Magid et al²⁶ mainly studied BTA as an adjunctive treatment to antidepressants in patients with ongoing unipolar depression. However, Finzi and Rosenthal²⁵ included patients who were not medicated at the time of the study.

Pooled analysis of these 3 RCTs found that patients who received BTA monotherapy improved equally to those who received it as an adjunctive treatment to antidepressants. Overall, on primary endpoint measures, a response rate of 54.2% was obtained in the BTA group compared with 10.7% among patients who received placebo saline injections (odds ratio [OR] 11.1, 95% confidence interval [CI], 4.3 to 28.8, number needed to treat [NNT] = 2.3) and a remission rate of 30.5% with BTA compared with 6.7% with placebo (OR 7.3, 95 % CI, 2.4 to 22.5, NNT = 4.2).⁴⁶ However, remission rates tend to be higher in the augmentation groups, and so further studies are needed to compare both treatment strategies.

Nevertheless, these positive findings have been recently challenged by the results of the largest U.S. multicenter phase II RCT,²⁸ which failed to find a significant antidepressant effect at 6 weeks with the 30-unit BTA injection, and also failed to prove a dose-effect relationship, as the 50-unit injection wasn’t superior to the lower dose and didn’t significantly differ from placebo. One hypothesis to explain this discrepancy may be the difference in injection sites between the treatment and placebo groups.⁴⁷ Future studies need to address the various limitations of earlier clinical trials that mainly yielded promising results with BTA.

A major concern is the high rate of unblinding of participants and researchers in BTA trials, as the cosmetic effects of botulinum toxin injections make them easy to distinguish from saline injections. Ninety percent of participants in the Wollmer et al study²³ were able to correctly guess their group allocation, while 60% of evaluators guessed correctly. Finzi and Rozenthal²⁵ reported 52% of participants in the BTA group, 46% in the placebo group, and 73% of evaluators correctly guessed their allocation. Magid et al²⁶ reported 75% of participants were able to guess the order of intervention they received. The high unblinding rates in these trials remains a significant limitation. There is a concern that this may lead to an underestimation of the placebo effect relative to clinical improvement, thus causing inflation of outcome differences between groups. Although various methods have been tried to minimize evaluator unblinding, such as placing surgical caps on participants’ faces during visits to hide the glabellar region, better methods need to be implemented to prevent unblinding of both raters and participants.

Furthermore, except for the multicenter phase II trial, most studies have been conducted in small samples, which limits their statistical power. Larger controlled trials will be needed to replicate the positive findings obtained in smaller RCTs.

Another limitation is that the majority of the well-designed RCTs were conducted in populations that were predominantly female, which makes it difficult to reliably assess treatment efficacy in men. This may be because cosmetic treatment with botulinum toxin injection is more favorably received by women than by men. A recent comparison⁴⁸ of the studies by Wollmer et al²³ and Finzi and Rosenthal²⁵ discussed an interesting observation. Wollmer et al did not explicitly mention botulinum toxin when recruiting for the study, while Finzi and Rosenthal did. While approximately a quarter of the participants in the Wollmer et al study were male, Finzi and Rosenthal attracted an almost entirely female population. Perhaps there is a potential bias for females to be more attracted to these studies due to the secondary gain of receiving a cosmetic procedure.

In an attempt to understand predictors of positive response to botulinum toxin in patients with depression, Wollmer et al⁴⁹ conducted a follow-up study in which they reassessed the data obtained from their initial RCT using the HAM-D agitation item scores to separate the 15 participants who received BTA into low-agitation (≤1 score on agitation item of the HAM-D scale) and high-agitation (≥2 score on agitation item of the HAM-D scale) groups. They found that the 9 participants who responded to BTA treatment had significantly higher baseline agitation scores than participants who did not respond (1.56 ± 0.88 vs 0.33 ± 0.52, P = .01). All of the participants who presented with higher agitation levels experienced response, compared with 40% of those with lower agitation levels (P = .04), although there was no significant difference in magnitude of improvement (14.2 ± 1.92 vs 8.0 ± 9.37, P = .07). The study added additional support to the facial feedback hypothesis, as it links the improvement of depression to facial muscle activation targeted by the injections. It also introduced a potential predictor of response to botulinum toxin treatment, highlighting potential factors to consider when enrolling patients in future investigations.

The case series of patients with borderline personality disorder³¹ also shed light on the potential positive effect of BTA treatment for a particular subtype of patients with depression—those with comorbid emotional instability—to consider as a therapeutic target for the future. Hence, inclusion criteria for future trials might potentially include patient age, gender, existence/quantification of prominent frown lines at baseline, severity of MDD, duration of depression, and personality characteristics of enrolled participants.

In conclusion, BTA injections appear promising as a treatment for depression as well as for other psychiatric disorders. Future studies should focus on identifying optimal candidates for this innovative treatment modality. Furthermore, BTA dosing and administration strategies (monotherapy vs adjunctive treatment to antidepressants) need to be further explored. As retrograde axonal transport of botulinum toxin has been demonstrated in animal studies, it would be interesting to further examine its effects in the human CNS to enhance our knowledge of the pathophysiology of botulinum and its potential applications in psychiatry.⁵⁰

Bottom Line

Botulinum toxin shows promising antidepressant effects and may have a role in the treatment of several other psychiatric disorders. More research is needed to address limitations of previous studies and to establish an adequate treatment regimen.

Related Resources

• Wollmer MA, Magid M, Kruger TH. Botulinum toxin treatment in depression. In: Bewley A, Taylor RE, Reichenberg JS, et al (eds). Practical psychodermatology. Oxford, UK: Wiley; 2014.
• Wollmer MA, Neumann I, Magid M. et al. Shrink that frown! Botulinum toxin therapy is lifting the face of psychiatry. G Ital Dermatol Venereol. 2018;153(4):540-548.

Drug Brand Names

Alprazolam • Xanax
Aripiprazole • Abilify
Biperiden • Akineton
Botulinum toxin A • Botox
Botulinum toxin B • Myobloc
Clozapine • Clozaril
Deutetrabenazine • Austedo
Flupentixol • Prolixin
Imipramine • Tofranil
Olanzapine • Zyprexa
Reserpine • Serpalan, Serpasil
Tetrabenazine • Xenazine
Valbenazine • Ingrezza
Ziprasidone • Geodon

Botulinum toxin, a potent neurotoxic protein produced by the bacterium Clostridium botulinum, has been used as treatment for a variety of medical indications for more than 25 years (Box^1-12). Recently, researchers have been exploring the role of botulinum toxin in psychiatry, primarily as an adjunctive treatment for depression, but also for several other possible indications. Several studies, including randomized controlled trials (RCTs), have provided evidence that glabellar botulinum toxin injections may be a safe and effective treatment for depression. In this article, we provide an update on the latest clinical trials that evaluated botulinum toxin for depression, and also summarize the evidence regarding other potential clinical psychiatric applications of botulinum toxin.

Several RCTs suggest efficacy for depression

The use of botulinum toxin to treat depression is based on the facial feedback hypothesis, which was first proposed by Charles Darwin in 1872¹³ and further elaborated by William James,^14,15 who emphasized the importance of the sensation of bodily changes in emotion. Contrary to the popular belief that emotions trigger physiological changes in the body, James postulated that peripheral bodily changes secondary to stimuli perception would exert a sensory feedback, generating emotions. The manipulation of human facial expression with an expression that is associated with a particular emotion (eg, holding a pen with teeth, leading to risorius/zygomaticus muscles contraction and a smile simulation) was found to influence participants’ affective responses in the presence of emotional stimuli (eg, rating cartoons as funnier), reinforcing the facial-feedback hypothesis.^16,17

From a neurobiologic standpoint, facial botulinum toxin A (BTA) injections in rats were associated with increased serotonin and norepinephrine concentrations in the hypothalamus and striatum, respectively.¹⁸ Moreover, amygdala activity in response to angry vs happy faces, measured via functional magnetic resonance imaging (fMRI), was found to be attenuated after BTA applications to muscles involved in angry facial expressions.^19,20 Both the neurotransmitters as well as the aforementioned brain regions have been implicated in the pathophysiology of depression.^21,22

Box

Continue to: In a second RCT involving 74 patients with depression...

Continue to: Benefits for other psychiatric indications

Borderline personality disorder. In a case series of 6 women, BTA injections in the glabellar region were reported to be particularly effective for the treatment of borderline personality disorder symptoms that were resistant to psychotherapy and pharmacotherapy.³¹ Two to 6 weeks after a 29-unit injection, borderline personality disorder symptoms as measured by the Zanarini Rating Scale for Borderline Personality Disorder and/or the Borderline Symptom List were shown to significantly improve by 49% to 94% from baseline (P ≤ .05). These findings emphasize the promising therapeutic role of BTA on depressive symptoms concomitant with the emotional lability, impulsivity, and negative emotions that usually characterize this personality disorder.^31,32 A small sample size and lack of a placebo comparator are limitations of this research.

Neuroleptic-induced sialorrhea. Botulinum toxin injections in the salivary glands have been investigated for treating clozapine-induced sialorrhea because they are thought to directly inhibit the release of acetylcholine from salivary glands. One small RCT that used botulinum toxin B (BTB)³³ and 1 case report that used BTA³⁴ reported successful reduction in hypersalivation, with doses ranging from 150 to 500 units injected in each of the parotid and/or submandibular glands bilaterally. Although the treatment was well tolerated and lasted up to 16 weeks, larger studies are needed to replicate these findings.^33-35

Orofacial tardive dyskinesia. Several case reports of orofacial tardive dyskinesia, including lingual dyskinesia and lingual protrusion dystonia, have found improvements in hyperkinetic movements following muscular BTA injections, such as in the genioglossus muscle in the case of tongue involvement.^36-39 These cases were, however, described in the literature before the recent FDA approval of the vesicular monoamine transporter 2 inhibitors valbenazine and deutetrabenazine for the treatment of tardive dyskinesia.^40,41

Studies examining botulinum toxin’s application in areas of psychiatry other than depression are summarized in Table 2.^31,33,36-38

Continue to: Promising initial findings but multiple limitations

Although BTA injections have been explored as a potential treatment for several psychiatric conditions, the bulk of recent evidence is derived from studies in patients with depressive disorders. BTA injections in the glabellar regions have been shown in small RCTs to be well-tolerated with overall promising improvement of depressive symptoms, optimally 6 weeks after a single injection. Moreover, BTA has been shown to be safe and long-lasting, which would be convenient for patients and might improve adherence to therapy.^42-44 BTA’s antidepressant effects were shown to be independent of frown line severity or patient satisfaction with cosmetic effects.⁴⁵ The trials by Wollmer et al,²³ Finzi and Rosenthal,²⁵ and Magid et al²⁶ mainly studied BTA as an adjunctive treatment to antidepressants in patients with ongoing unipolar depression. However, Finzi and Rosenthal²⁵ included patients who were not medicated at the time of the study.

Pooled analysis of these 3 RCTs found that patients who received BTA monotherapy improved equally to those who received it as an adjunctive treatment to antidepressants. Overall, on primary endpoint measures, a response rate of 54.2% was obtained in the BTA group compared with 10.7% among patients who received placebo saline injections (odds ratio [OR] 11.1, 95% confidence interval [CI], 4.3 to 28.8, number needed to treat [NNT] = 2.3) and a remission rate of 30.5% with BTA compared with 6.7% with placebo (OR 7.3, 95 % CI, 2.4 to 22.5, NNT = 4.2).⁴⁶ However, remission rates tend to be higher in the augmentation groups, and so further studies are needed to compare both treatment strategies.

Nevertheless, these positive findings have been recently challenged by the results of the largest U.S. multicenter phase II RCT,²⁸ which failed to find a significant antidepressant effect at 6 weeks with the 30-unit BTA injection, and also failed to prove a dose-effect relationship, as the 50-unit injection wasn’t superior to the lower dose and didn’t significantly differ from placebo. One hypothesis to explain this discrepancy may be the difference in injection sites between the treatment and placebo groups.⁴⁷ Future studies need to address the various limitations of earlier clinical trials that mainly yielded promising results with BTA.

A major concern is the high rate of unblinding of participants and researchers in BTA trials, as the cosmetic effects of botulinum toxin injections make them easy to distinguish from saline injections. Ninety percent of participants in the Wollmer et al study²³ were able to correctly guess their group allocation, while 60% of evaluators guessed correctly. Finzi and Rozenthal²⁵ reported 52% of participants in the BTA group, 46% in the placebo group, and 73% of evaluators correctly guessed their allocation. Magid et al²⁶ reported 75% of participants were able to guess the order of intervention they received. The high unblinding rates in these trials remains a significant limitation. There is a concern that this may lead to an underestimation of the placebo effect relative to clinical improvement, thus causing inflation of outcome differences between groups. Although various methods have been tried to minimize evaluator unblinding, such as placing surgical caps on participants’ faces during visits to hide the glabellar region, better methods need to be implemented to prevent unblinding of both raters and participants.

Furthermore, except for the multicenter phase II trial, most studies have been conducted in small samples, which limits their statistical power. Larger controlled trials will be needed to replicate the positive findings obtained in smaller RCTs.

Another limitation is that the majority of the well-designed RCTs were conducted in populations that were predominantly female, which makes it difficult to reliably assess treatment efficacy in men. This may be because cosmetic treatment with botulinum toxin injection is more favorably received by women than by men. A recent comparison⁴⁸ of the studies by Wollmer et al²³ and Finzi and Rosenthal²⁵ discussed an interesting observation. Wollmer et al did not explicitly mention botulinum toxin when recruiting for the study, while Finzi and Rosenthal did. While approximately a quarter of the participants in the Wollmer et al study were male, Finzi and Rosenthal attracted an almost entirely female population. Perhaps there is a potential bias for females to be more attracted to these studies due to the secondary gain of receiving a cosmetic procedure.

In an attempt to understand predictors of positive response to botulinum toxin in patients with depression, Wollmer et al⁴⁹ conducted a follow-up study in which they reassessed the data obtained from their initial RCT using the HAM-D agitation item scores to separate the 15 participants who received BTA into low-agitation (≤1 score on agitation item of the HAM-D scale) and high-agitation (≥2 score on agitation item of the HAM-D scale) groups. They found that the 9 participants who responded to BTA treatment had significantly higher baseline agitation scores than participants who did not respond (1.56 ± 0.88 vs 0.33 ± 0.52, P = .01). All of the participants who presented with higher agitation levels experienced response, compared with 40% of those with lower agitation levels (P = .04), although there was no significant difference in magnitude of improvement (14.2 ± 1.92 vs 8.0 ± 9.37, P = .07). The study added additional support to the facial feedback hypothesis, as it links the improvement of depression to facial muscle activation targeted by the injections. It also introduced a potential predictor of response to botulinum toxin treatment, highlighting potential factors to consider when enrolling patients in future investigations.

The case series of patients with borderline personality disorder³¹ also shed light on the potential positive effect of BTA treatment for a particular subtype of patients with depression—those with comorbid emotional instability—to consider as a therapeutic target for the future. Hence, inclusion criteria for future trials might potentially include patient age, gender, existence/quantification of prominent frown lines at baseline, severity of MDD, duration of depression, and personality characteristics of enrolled participants.

In conclusion, BTA injections appear promising as a treatment for depression as well as for other psychiatric disorders. Future studies should focus on identifying optimal candidates for this innovative treatment modality. Furthermore, BTA dosing and administration strategies (monotherapy vs adjunctive treatment to antidepressants) need to be further explored. As retrograde axonal transport of botulinum toxin has been demonstrated in animal studies, it would be interesting to further examine its effects in the human CNS to enhance our knowledge of the pathophysiology of botulinum and its potential applications in psychiatry.⁵⁰

Bottom Line

Botulinum toxin shows promising antidepressant effects and may have a role in the treatment of several other psychiatric disorders. More research is needed to address limitations of previous studies and to establish an adequate treatment regimen.

Related Resources

• Wollmer MA, Magid M, Kruger TH. Botulinum toxin treatment in depression. In: Bewley A, Taylor RE, Reichenberg JS, et al (eds). Practical psychodermatology. Oxford, UK: Wiley; 2014.
• Wollmer MA, Neumann I, Magid M. et al. Shrink that frown! Botulinum toxin therapy is lifting the face of psychiatry. G Ital Dermatol Venereol. 2018;153(4):540-548.

Drug Brand Names

Alprazolam • Xanax
Aripiprazole • Abilify
Biperiden • Akineton
Botulinum toxin A • Botox
Botulinum toxin B • Myobloc
Clozapine • Clozaril
Deutetrabenazine • Austedo
Flupentixol • Prolixin
Imipramine • Tofranil
Olanzapine • Zyprexa
Reserpine • Serpalan, Serpasil
Tetrabenazine • Xenazine
Valbenazine • Ingrezza
Ziprasidone • Geodon

In schizophrenia, negative symptoms such as social withdrawal, avoidance, lack of spontaneity and flow of conversation, reduced initiative, anhedonia, and blunted affect are among the most challenging to treat. These symptoms commonly persist after positive symptoms such as hallucinations, delusions, and paranoia have subsided. In an analysis of 20 pivotal placebo-controlled trials of second-generation antipsychotics (SGAs), almost 45% of patients who completed 6 weeks of treatment still had at least 1 residual negative symptom of at least moderate severity, and approximately 25% had 2 or more.¹ Negative symptoms are viewed as being intrinsic to schizophrenia, and also as the result of extrapyramidal symptoms, depression, and psychosis.

Nearly a decade ago, the Schizophrenia Patient Outcomes Research Team (PORT) published its recommendations for psychopharmacologic and psycho­social treatments of schizophrenia. Unfortunately, due to insufficient evidence, there is still no proven treatment for negative symptoms.^2-4 This is particularly problematic because negative symptoms are a major determinant of the poor social and vocational abilities of patients with schizophrenia.

This review focuses on treatments for negative symptoms of schizophrenia that have been evaluated since the PORT treatment recommendations were published and highlights those approaches that show promise.

_

The limitations of antipsychotics

Antipsychotics can both worsen and alleviate negative symptoms by reducing psychotic symptoms. Double-blind, placebo-controlled trials have found that most, if not all, antipsychotics are superior to placebo for treating negative symptoms in patients with acute psychosis.⁴ However, because these improvements occur in the early stages of treatment, concomitantly with improvement of psychotic symptoms, antipsychotics generally are not viewed as being very effective in the treatment of primary negative symptoms.⁴ Indeed, an examination of patients with prominent negative symptoms without prominent positive symptoms in the NEWMEDS cohort, which was extracted from 20 pivotal placebo-controlled trials of SGAs, revealed no clinically meaningful treatment effect on negative symptoms.¹

There is evidence that antipsychotics can contribute to the development of apathy, flat affect, and other negative symptoms.⁵ Dopamine (D2)-blocking antipsychotics produce secondary negative symptoms that are not always easy to distinguish from primary negative symptoms.⁶ In a double-blind, placebo-controlled trial of single doses of risperidone, haloperidol, or placebo in healthy participants, the antipsychotics increased negative symptoms, particularly avolition/apathy.⁷ Another study found that chronic treatment with antipsychotics did not necessarily affect motivation in patients with schizophrenia.⁸

Adverse effects, such as anhedonia, often produce and enhance negative symptoms and therefore can limit the use of pharmacologic treatment options. Other adverse effects associated with specific antipsycho­tics include extrapyramidal symptoms, sedation, increased prolactin secretion, weight gain, and other metabolic abnormalities.

Continue to: Seeking new pharmacologic options

The years since the PORT review have been filled with initial promise, a series of bitter disappointments, and a renewed spark of hope in the quest to treat negative symptoms in schizophrenia.


Compounds that have been abandoned. Since PORT, researchers have evaluated 5 major compounds that mainly targeted cognition and negative symptoms in patients with schizophrenia (Box^9-17). Unfortunately, 4 of them failed to provide significant superiority over placebo, and 1 was withdrawn due to safety concerns.

Box

Initial favorable results. Registered, robust trials of other compounds have had some initial favorable results that need to be replicated. These agents include:

• MIN-101 is a novel cyclic amide derivative.¹⁸ In a phase IIb 12-week study of MIN-101 monotherapy (32 mg, n = 78; 64 mg, n = 83) vs placebo (n = 83), both dose groups had significantly more improvement on the Positive and Negative Syndrome Scale (PANSS) negative factor score, which was the primary outcome measure, than placebo (32 mg/d; effect size = .45, P < .02, 64 mg/d; effect size = .57, P < .004) as well as on PANSS negative symptom score and other measures of negative symptoms.¹⁸
• Cariprazine is a D2 and D3 receptor partial agonist with high selectivity towards the D3 receptor¹⁹
• Minocycline is a broad-spectrum tetracyclic antibiotic displaying neuroprotective properties^18,20,21
• Raloxifene is a selective estrogen receptor modulator for postmenopausal women^22,23
• Pimavanserin, which was FDA-approved in 2016 for the treatment of Parkinson’s disease psychosis, is being tested in a large trial for adjunctive treatment of patients with negative symptoms of schizophrenia. This medication is a nondopaminergic antipsychotic that acts as a selective serotonin inverse agonist that preferentially targets 5-HT2A receptors while avoiding activity at common targets such as dopamine.²⁴

All of these compounds except MIN-101 are currently available in the U.S. but have not been approved for the treatment of negative symptoms in patients with schizophrenia. MIN-101 is in phase III testing (NCT03397134).

Continue to: Nonpharmacologic treatments

Recent studies of nonpharmacologic treatments for negative symptoms, including psychosocial approaches and noninvasive electromagnetic neurostimulation, have also been performed. The major psychosocial approaches that have been studied include social skills training (SST), cognitive-behavioral therapy (CBT) for psychosis, cognitive remediation, and family intervention. Some positive findings have been reported. A recent review of psychosocial treatments for negative symptoms in schizophrenia concluded that CBT and SST have the most empirical support, with some evidence even suggesting that gains from CBT are maintained as long as 6 months after treatment.²⁵ Another review found that CBT was significantly more efficacious for reducing positive symptoms and SST in reducing negative symptoms.²⁶

It remains unclear if a combined treatment approach provides improvements above and beyond those associated with each individual treatment modality. Motivation and Enhancement therapy (MOVE) is a potentially promising approach that combines environmental support, CBT, skills training, and other components in an attempt to address all domains of negative symptoms.²⁷ Preliminary results from a randomized controlled trial examining 51 patients with clinically meaningful negative symptoms suggested that MOVE improves negative symptoms. However, the group differences were not significant until after 9 months of treatment and not for all negative symptom scales. A follow-up study has been completed, but the results are not yet available.²⁸

Some small studies have suggested improvement of negative symptoms after noninvasive electromagnetic neurostimulation,^29-31 but this has not been replicated in larger studies.³² In the last few years, there were several studies underway that could help clarify if there is a role for noninvasive electromagnetic neurostimulation in the treatment of negative symptoms in schizophrenia; however, results have not been reported at this time.^33-35

_

Social skills training and combined interventions

Taken together, the data suggest that treating negative symptoms in schizophrenia remains a major challenge. Patients with negative symptoms are difficult to engage and motivate for treatment and there are no well-supported treatment options. Given the lack of evidence, it is not possible to synthesize this data into clear treatment recommendations. Because many of the negative symptoms are social in nature, it is perhaps not surprising that some evidence has emerged supporting the role of psycho­social approaches. Studies have pointed to the potential role of SST. It is believed to be beneficial as it targets participants’ social functioning by training verbal and nonverbal communication alongside perception and responses to social cues.³⁶ Some evidence suggests that treatment packages that combine several psychosocial interventions (eg, family psychoeducation and skill training) achieve better outcomes than standalone interventions.³⁷ Thus, psychosocial approaches appear to be potentially effective for the treatment of negative symptoms in patients with schizophrenia. In addition, because some antipsychotics has been shown to be associated with fewer negative symptoms than others, another treatment strategy could be to attempt the use of a different antipsychotic, or to revisit whether continued antipsychotic treatment is needed in the absence of positive symptoms.

Bottom Line

Treating negative symptoms in schizophrenia remains a major challenge. There is a lack of evidence for pharmacologic treatments; psychosocial approaches may be beneficial due to the social nature of many negative symptoms. Further, some evidence suggests that treatment packages that combine several psychosocial interventions may achieve better outcomes than standalone interventions.

Related Resource

Tandon R, Jibson M. Negative symptoms of schizophrenia: How to treat them most effectively. Current Psychiatry. 2002;1(9):36-42.

Drug Brand Names

Cariprazine • Vraylar
Haloperidol • Haldol
Minocycline • Dynacin, Minocin
Pimavanserin • Nuplazid
Raloxifene • Evista
Risperidone • Risperdal

In schizophrenia, negative symptoms such as social withdrawal, avoidance, lack of spontaneity and flow of conversation, reduced initiative, anhedonia, and blunted affect are among the most challenging to treat. These symptoms commonly persist after positive symptoms such as hallucinations, delusions, and paranoia have subsided. In an analysis of 20 pivotal placebo-controlled trials of second-generation antipsychotics (SGAs), almost 45% of patients who completed 6 weeks of treatment still had at least 1 residual negative symptom of at least moderate severity, and approximately 25% had 2 or more.¹ Negative symptoms are viewed as being intrinsic to schizophrenia, and also as the result of extrapyramidal symptoms, depression, and psychosis.

Nearly a decade ago, the Schizophrenia Patient Outcomes Research Team (PORT) published its recommendations for psychopharmacologic and psycho­social treatments of schizophrenia. Unfortunately, due to insufficient evidence, there is still no proven treatment for negative symptoms.^2-4 This is particularly problematic because negative symptoms are a major determinant of the poor social and vocational abilities of patients with schizophrenia.

This review focuses on treatments for negative symptoms of schizophrenia that have been evaluated since the PORT treatment recommendations were published and highlights those approaches that show promise.

_

The limitations of antipsychotics

Antipsychotics can both worsen and alleviate negative symptoms by reducing psychotic symptoms. Double-blind, placebo-controlled trials have found that most, if not all, antipsychotics are superior to placebo for treating negative symptoms in patients with acute psychosis.⁴ However, because these improvements occur in the early stages of treatment, concomitantly with improvement of psychotic symptoms, antipsychotics generally are not viewed as being very effective in the treatment of primary negative symptoms.⁴ Indeed, an examination of patients with prominent negative symptoms without prominent positive symptoms in the NEWMEDS cohort, which was extracted from 20 pivotal placebo-controlled trials of SGAs, revealed no clinically meaningful treatment effect on negative symptoms.¹

There is evidence that antipsychotics can contribute to the development of apathy, flat affect, and other negative symptoms.⁵ Dopamine (D2)-blocking antipsychotics produce secondary negative symptoms that are not always easy to distinguish from primary negative symptoms.⁶ In a double-blind, placebo-controlled trial of single doses of risperidone, haloperidol, or placebo in healthy participants, the antipsychotics increased negative symptoms, particularly avolition/apathy.⁷ Another study found that chronic treatment with antipsychotics did not necessarily affect motivation in patients with schizophrenia.⁸

Adverse effects, such as anhedonia, often produce and enhance negative symptoms and therefore can limit the use of pharmacologic treatment options. Other adverse effects associated with specific antipsycho­tics include extrapyramidal symptoms, sedation, increased prolactin secretion, weight gain, and other metabolic abnormalities.

Continue to: Seeking new pharmacologic options

The years since the PORT review have been filled with initial promise, a series of bitter disappointments, and a renewed spark of hope in the quest to treat negative symptoms in schizophrenia.


Compounds that have been abandoned. Since PORT, researchers have evaluated 5 major compounds that mainly targeted cognition and negative symptoms in patients with schizophrenia (Box^9-17). Unfortunately, 4 of them failed to provide significant superiority over placebo, and 1 was withdrawn due to safety concerns.

Box

Initial favorable results. Registered, robust trials of other compounds have had some initial favorable results that need to be replicated. These agents include:

• MIN-101 is a novel cyclic amide derivative.¹⁸ In a phase IIb 12-week study of MIN-101 monotherapy (32 mg, n = 78; 64 mg, n = 83) vs placebo (n = 83), both dose groups had significantly more improvement on the Positive and Negative Syndrome Scale (PANSS) negative factor score, which was the primary outcome measure, than placebo (32 mg/d; effect size = .45, P < .02, 64 mg/d; effect size = .57, P < .004) as well as on PANSS negative symptom score and other measures of negative symptoms.¹⁸
• Cariprazine is a D2 and D3 receptor partial agonist with high selectivity towards the D3 receptor¹⁹
• Minocycline is a broad-spectrum tetracyclic antibiotic displaying neuroprotective properties^18,20,21
• Raloxifene is a selective estrogen receptor modulator for postmenopausal women^22,23
• Pimavanserin, which was FDA-approved in 2016 for the treatment of Parkinson’s disease psychosis, is being tested in a large trial for adjunctive treatment of patients with negative symptoms of schizophrenia. This medication is a nondopaminergic antipsychotic that acts as a selective serotonin inverse agonist that preferentially targets 5-HT2A receptors while avoiding activity at common targets such as dopamine.²⁴

All of these compounds except MIN-101 are currently available in the U.S. but have not been approved for the treatment of negative symptoms in patients with schizophrenia. MIN-101 is in phase III testing (NCT03397134).

Continue to: Nonpharmacologic treatments

Recent studies of nonpharmacologic treatments for negative symptoms, including psychosocial approaches and noninvasive electromagnetic neurostimulation, have also been performed. The major psychosocial approaches that have been studied include social skills training (SST), cognitive-behavioral therapy (CBT) for psychosis, cognitive remediation, and family intervention. Some positive findings have been reported. A recent review of psychosocial treatments for negative symptoms in schizophrenia concluded that CBT and SST have the most empirical support, with some evidence even suggesting that gains from CBT are maintained as long as 6 months after treatment.²⁵ Another review found that CBT was significantly more efficacious for reducing positive symptoms and SST in reducing negative symptoms.²⁶

It remains unclear if a combined treatment approach provides improvements above and beyond those associated with each individual treatment modality. Motivation and Enhancement therapy (MOVE) is a potentially promising approach that combines environmental support, CBT, skills training, and other components in an attempt to address all domains of negative symptoms.²⁷ Preliminary results from a randomized controlled trial examining 51 patients with clinically meaningful negative symptoms suggested that MOVE improves negative symptoms. However, the group differences were not significant until after 9 months of treatment and not for all negative symptom scales. A follow-up study has been completed, but the results are not yet available.²⁸

Some small studies have suggested improvement of negative symptoms after noninvasive electromagnetic neurostimulation,^29-31 but this has not been replicated in larger studies.³² In the last few years, there were several studies underway that could help clarify if there is a role for noninvasive electromagnetic neurostimulation in the treatment of negative symptoms in schizophrenia; however, results have not been reported at this time.^33-35

_

Social skills training and combined interventions

Taken together, the data suggest that treating negative symptoms in schizophrenia remains a major challenge. Patients with negative symptoms are difficult to engage and motivate for treatment and there are no well-supported treatment options. Given the lack of evidence, it is not possible to synthesize this data into clear treatment recommendations. Because many of the negative symptoms are social in nature, it is perhaps not surprising that some evidence has emerged supporting the role of psycho­social approaches. Studies have pointed to the potential role of SST. It is believed to be beneficial as it targets participants’ social functioning by training verbal and nonverbal communication alongside perception and responses to social cues.³⁶ Some evidence suggests that treatment packages that combine several psychosocial interventions (eg, family psychoeducation and skill training) achieve better outcomes than standalone interventions.³⁷ Thus, psychosocial approaches appear to be potentially effective for the treatment of negative symptoms in patients with schizophrenia. In addition, because some antipsychotics has been shown to be associated with fewer negative symptoms than others, another treatment strategy could be to attempt the use of a different antipsychotic, or to revisit whether continued antipsychotic treatment is needed in the absence of positive symptoms.

Bottom Line

Treating negative symptoms in schizophrenia remains a major challenge. There is a lack of evidence for pharmacologic treatments; psychosocial approaches may be beneficial due to the social nature of many negative symptoms. Further, some evidence suggests that treatment packages that combine several psychosocial interventions may achieve better outcomes than standalone interventions.

Related Resource

Tandon R, Jibson M. Negative symptoms of schizophrenia: How to treat them most effectively. Current Psychiatry. 2002;1(9):36-42.

Drug Brand Names

Cariprazine • Vraylar
Haloperidol • Haldol
Minocycline • Dynacin, Minocin
Pimavanserin • Nuplazid
Raloxifene • Evista
Risperidone • Risperdal

In schizophrenia, negative symptoms such as social withdrawal, avoidance, lack of spontaneity and flow of conversation, reduced initiative, anhedonia, and blunted affect are among the most challenging to treat. These symptoms commonly persist after positive symptoms such as hallucinations, delusions, and paranoia have subsided. In an analysis of 20 pivotal placebo-controlled trials of second-generation antipsychotics (SGAs), almost 45% of patients who completed 6 weeks of treatment still had at least 1 residual negative symptom of at least moderate severity, and approximately 25% had 2 or more.¹ Negative symptoms are viewed as being intrinsic to schizophrenia, and also as the result of extrapyramidal symptoms, depression, and psychosis.

Nearly a decade ago, the Schizophrenia Patient Outcomes Research Team (PORT) published its recommendations for psychopharmacologic and psycho­social treatments of schizophrenia. Unfortunately, due to insufficient evidence, there is still no proven treatment for negative symptoms.^2-4 This is particularly problematic because negative symptoms are a major determinant of the poor social and vocational abilities of patients with schizophrenia.

This review focuses on treatments for negative symptoms of schizophrenia that have been evaluated since the PORT treatment recommendations were published and highlights those approaches that show promise.

_

The limitations of antipsychotics

Antipsychotics can both worsen and alleviate negative symptoms by reducing psychotic symptoms. Double-blind, placebo-controlled trials have found that most, if not all, antipsychotics are superior to placebo for treating negative symptoms in patients with acute psychosis.⁴ However, because these improvements occur in the early stages of treatment, concomitantly with improvement of psychotic symptoms, antipsychotics generally are not viewed as being very effective in the treatment of primary negative symptoms.⁴ Indeed, an examination of patients with prominent negative symptoms without prominent positive symptoms in the NEWMEDS cohort, which was extracted from 20 pivotal placebo-controlled trials of SGAs, revealed no clinically meaningful treatment effect on negative symptoms.¹

There is evidence that antipsychotics can contribute to the development of apathy, flat affect, and other negative symptoms.⁵ Dopamine (D2)-blocking antipsychotics produce secondary negative symptoms that are not always easy to distinguish from primary negative symptoms.⁶ In a double-blind, placebo-controlled trial of single doses of risperidone, haloperidol, or placebo in healthy participants, the antipsychotics increased negative symptoms, particularly avolition/apathy.⁷ Another study found that chronic treatment with antipsychotics did not necessarily affect motivation in patients with schizophrenia.⁸

Adverse effects, such as anhedonia, often produce and enhance negative symptoms and therefore can limit the use of pharmacologic treatment options. Other adverse effects associated with specific antipsycho­tics include extrapyramidal symptoms, sedation, increased prolactin secretion, weight gain, and other metabolic abnormalities.

Continue to: Seeking new pharmacologic options

The years since the PORT review have been filled with initial promise, a series of bitter disappointments, and a renewed spark of hope in the quest to treat negative symptoms in schizophrenia.


Compounds that have been abandoned. Since PORT, researchers have evaluated 5 major compounds that mainly targeted cognition and negative symptoms in patients with schizophrenia (Box^9-17). Unfortunately, 4 of them failed to provide significant superiority over placebo, and 1 was withdrawn due to safety concerns.

Box

Initial favorable results. Registered, robust trials of other compounds have had some initial favorable results that need to be replicated. These agents include:

• MIN-101 is a novel cyclic amide derivative.¹⁸ In a phase IIb 12-week study of MIN-101 monotherapy (32 mg, n = 78; 64 mg, n = 83) vs placebo (n = 83), both dose groups had significantly more improvement on the Positive and Negative Syndrome Scale (PANSS) negative factor score, which was the primary outcome measure, than placebo (32 mg/d; effect size = .45, P < .02, 64 mg/d; effect size = .57, P < .004) as well as on PANSS negative symptom score and other measures of negative symptoms.¹⁸
• Cariprazine is a D2 and D3 receptor partial agonist with high selectivity towards the D3 receptor¹⁹
• Minocycline is a broad-spectrum tetracyclic antibiotic displaying neuroprotective properties^18,20,21
• Raloxifene is a selective estrogen receptor modulator for postmenopausal women^22,23
• Pimavanserin, which was FDA-approved in 2016 for the treatment of Parkinson’s disease psychosis, is being tested in a large trial for adjunctive treatment of patients with negative symptoms of schizophrenia. This medication is a nondopaminergic antipsychotic that acts as a selective serotonin inverse agonist that preferentially targets 5-HT2A receptors while avoiding activity at common targets such as dopamine.²⁴

All of these compounds except MIN-101 are currently available in the U.S. but have not been approved for the treatment of negative symptoms in patients with schizophrenia. MIN-101 is in phase III testing (NCT03397134).

Continue to: Nonpharmacologic treatments

Recent studies of nonpharmacologic treatments for negative symptoms, including psychosocial approaches and noninvasive electromagnetic neurostimulation, have also been performed. The major psychosocial approaches that have been studied include social skills training (SST), cognitive-behavioral therapy (CBT) for psychosis, cognitive remediation, and family intervention. Some positive findings have been reported. A recent review of psychosocial treatments for negative symptoms in schizophrenia concluded that CBT and SST have the most empirical support, with some evidence even suggesting that gains from CBT are maintained as long as 6 months after treatment.²⁵ Another review found that CBT was significantly more efficacious for reducing positive symptoms and SST in reducing negative symptoms.²⁶

It remains unclear if a combined treatment approach provides improvements above and beyond those associated with each individual treatment modality. Motivation and Enhancement therapy (MOVE) is a potentially promising approach that combines environmental support, CBT, skills training, and other components in an attempt to address all domains of negative symptoms.²⁷ Preliminary results from a randomized controlled trial examining 51 patients with clinically meaningful negative symptoms suggested that MOVE improves negative symptoms. However, the group differences were not significant until after 9 months of treatment and not for all negative symptom scales. A follow-up study has been completed, but the results are not yet available.²⁸

Some small studies have suggested improvement of negative symptoms after noninvasive electromagnetic neurostimulation,^29-31 but this has not been replicated in larger studies.³² In the last few years, there were several studies underway that could help clarify if there is a role for noninvasive electromagnetic neurostimulation in the treatment of negative symptoms in schizophrenia; however, results have not been reported at this time.^33-35

_

Social skills training and combined interventions

Taken together, the data suggest that treating negative symptoms in schizophrenia remains a major challenge. Patients with negative symptoms are difficult to engage and motivate for treatment and there are no well-supported treatment options. Given the lack of evidence, it is not possible to synthesize this data into clear treatment recommendations. Because many of the negative symptoms are social in nature, it is perhaps not surprising that some evidence has emerged supporting the role of psycho­social approaches. Studies have pointed to the potential role of SST. It is believed to be beneficial as it targets participants’ social functioning by training verbal and nonverbal communication alongside perception and responses to social cues.³⁶ Some evidence suggests that treatment packages that combine several psychosocial interventions (eg, family psychoeducation and skill training) achieve better outcomes than standalone interventions.³⁷ Thus, psychosocial approaches appear to be potentially effective for the treatment of negative symptoms in patients with schizophrenia. In addition, because some antipsychotics has been shown to be associated with fewer negative symptoms than others, another treatment strategy could be to attempt the use of a different antipsychotic, or to revisit whether continued antipsychotic treatment is needed in the absence of positive symptoms.

Bottom Line

Treating negative symptoms in schizophrenia remains a major challenge. There is a lack of evidence for pharmacologic treatments; psychosocial approaches may be beneficial due to the social nature of many negative symptoms. Further, some evidence suggests that treatment packages that combine several psychosocial interventions may achieve better outcomes than standalone interventions.

Related Resource

Tandon R, Jibson M. Negative symptoms of schizophrenia: How to treat them most effectively. Current Psychiatry. 2002;1(9):36-42.

Drug Brand Names

Cariprazine • Vraylar
Haloperidol • Haldol
Minocycline • Dynacin, Minocin
Pimavanserin • Nuplazid
Raloxifene • Evista
Risperidone • Risperdal

Oral antipsychotic nonadherence is a significant contributor to relapse in patients with schizophrenia spectrum disorders. Long-acting injectable (LAI) antipsychotics have been developed to provide sustained antipsychotic exposure, with evidence that use of LAIs significantly reduces hospitalization rates.¹ One limiting factor in transitioning patients to certain LAIs is the need for prolonged oral coverage at the onset of treatment for agents that cannot be loaded. Nonadherence with this bridging oral therapy places the patient at risk for symptom exacerbation until effective antipsychotic plasma levels are achieved from the LAI.² Although risperidone is one of the more widely used antipsychotics for treating schizophrenia, until recently the only available LAI preparation, risperidone microspheres (Risperdal Consta), required 3 weeks of oral coverage upon initiation.³

Clinical implications

Oral medication nonadherence remains a significant public health issue for patients with schizophrenia, with an estimated 50% of patients failing to achieve 80% adherence even when enrolled in clinical trials specifically designed to track adherence.⁵ Although LAI atypical antipsychotics have been available since the approval of Risperdal Consta, the LAI form of risperidone, and both LAI forms of aripiprazole, were not designed to be loaded. A 1-day initiation regimen for aripiprazole lauroxil has been developed to avoid the need for 3 weeks of oral medication coverage,^6,7 but aripiprazole monohydrate and risperidone microspheres mandate oral bridging of 2 and 3 weeks, respectively.² Because one of the primary indications for LAI antipsychotic therapy is oral medication nonadherence, this prolonged period of oral coverage creates a risk for symptom exacerbation when the bridging period occurs outside of a controlled setting, as is common when patients are discharged from inpatient hospitalization.

One solution to this problem has its antecedents in the development of the Atrigel biodegradable injectable polymer, which was designed to deliver prolonged medication exposure after subcutaneous injection.⁸ This biodegradable polymer drug delivery system suspends and dissolves the medication of interest (in this case, risperidone) in a poly DL-lactide-coglycolide gel and its biocompatible carrier.⁹ The viscous liquid undergoes a phase transition upon contact with tissue fluids after subcutaneous injection, resulting in an implant that releases risperidone in a controlled manner as it is resorbed. Importantly, the kinetic parameters of RBP-7000 are such that effective drug levels are seen within the first week without the need for oral coverage.¹⁰

Use in adults with schizophrenia. After establishing tolerability with oral risperidone, the recommended doses are 90 mg or 120 mg monthly, which correspond to oral daily risperidone doses of 3 mg or 4 mg. RBP-7000 must be administered as a subcutaneous abdominal injection by a health care professional. It is recommended that the patient be in the supine position for the injection and that the injection sites be rotated monthly among 4 quadrants in the abdominal region. The injection volumes for the 90 mg and 120 mg doses are 0.6 mL and 0.8 mL, respectively.¹⁰ As the gel implant becomes firmer, the patient will notice a lump for several weeks that will decrease in size over time. Patients should be advised not to rub or massage the injection site, and to be aware of the placement of any belts or clothing with waistbands.¹⁰

Pharmacologic profile, adverse reactions

Risperidone is an atypical antipsychotic that has been commercially available in the U.S. since December 29, 1993, and its adverse effect profile is well characterized. The most common adverse effects associated with risperidone include those related to dopamine D2 antagonism, metabolic adverse effects, and an increase in serum prolactin. In the 12-month long-term safety study of RBP-7000, 1-minute post-dose injection site pain scores (on a 100-point scale) were highest on Day 1 (mean of 25) and decreased over time with subsequent injections (14 to 16 following the last injection).¹⁰

Continue to: How the Atrigel system works

The kinetics for RBP-7000 are markedly different than those for oral risperidone (Figure 2¹⁵). After a single subcutaneous injection, RBP-7000 shows 2 absorption peaks for risperidone. The first lower peak occurs with a T_max of 4 to 6 hours due to initial release of risperidone during the implant formation process; a second risperidone peak occurs after 10 to 14 days and is associated with slow release from the subcutaneous depot.^9,16,17 For both 9-OH risperidone levels and the total active moiety (risperidone plus 9-OH risperidone levels), the median T_max of the first peak ranges from 4 to 48 hours and the second peak ranges from 7 to 11 days. Following a single subcutaneous injection of RBP-7000, the apparent terminal half-life of risperidone ranges from 9 to 11 days, on average. The mean apparent terminal half-life of the active moiety ranges from 8 to 9 days.^9,16,17 Based on population pharmacokinetic modeling, the 90 mg and 120 mg doses of RBP-7000 are estimated to provide drug exposure equivalent to 3 mg/d and 4 mg/d of oral risperidone, respectively.^9,16,17

Continue to: Efficacy of RBP-7000

Efficacy of RBP-7000 was established in an 8-week, double-blind, placebo-controlled trial of adult patients experiencing an acute exacerbation of schizo­phrenia (age 18 to 55).⁴ Eligible participants had:

• An acute exacerbation of schizophrenia that occurred ≤8 weeks before the screening visit and would have benefited from psychiatric hospitalization or continued hospitalization
• Positive and Negative Syndrome Scale (PANSS) total score between 80 and 120 at visit 1 and a score of >4 on at least 2 of the following 4 items: hallucinatory behavior, delusions, conceptual disorganization, or suspiciousness/persecution
• The diagnosis of acute exacerbation of schizophrenia and PANSS total score were confirmed through an independent video-conference interview conducted by an experienced rater.

Participants were excluded if they:

• Experienced a ≥20% improvement in PANSS total score between the initial screening visit and the first injection
• had been treated at any time with clozapine for treatment-resistant schizophrenia
• had met DSM-IV-TR criteria for substance dependence (with the exception of nicotine or caffeine) before screening.

During the initial screening visit, participants received a 0.25-mg tablet of oral risperidone on 2 consecutive days to assess the tolerability of risperidone.

Outcome. Participants were randomized in a 1:1:1 manner to placebo (n = 112) or 1 of 2 monthly doses of RBP-7000: 90 mg (n = 111) or 120 mg (n = 114). Using the least squares means of repeated-measures changes from baseline in PANSS total scores, there was a significant improvement in the difference in PANSS total scores from baseline to the end of the study compared with placebo: 90-mg RBP-7000, -6.148 points (95% confidence interval [CI], -9.982 to -2.314, P = .0004); 120-mg RBP-7000, -7.237 points (95% CI, -11.045 to -3.429, P < .0001). The absolute change from baseline in PANSS total score was -15.367 points for the 90-mg dose and -16.456 points for the 120-mg dose.⁴ Completion rates across all 3 arms were comparable: placebo 70.6%, RBP-7000 90 mg 77.6%, and RBP-7000 120 mg 71.4%.

Tolerability. In the 8-week phase III efficacy trial of RBP-7000, adverse effects occurring with an incidence ≥5% and at least twice the rate of placebo were weight gain (placebo 3.4%, 90 mg 13.0%, 120 mg 12.8%) and sedation (placebo 0%, 90 mg 7.0%, 120 mg 7.7%).¹⁰ Compared with baseline, participants had a mean weight gain at the end of the study of 2.83 kg in the placebo group, 5.15 kg in the 90-mg RBP-7000 group, and 4.69 kg in the 120-mg RBP-7000 group. There were no clinically significant differences at study endpoint in glucose and lipid parameters. Consistent with the known effects of risperidone, there were increases in mean prolactin levels during the 8-week study, the effects of which were greater for women. For men, mean prolactin levels from baseline to study end were: placebo: 9.8 ± 7.9 vs 9.9 ± 8.0 ng/mL; 90 mg: 8.9 ± 6.9 vs 22.4 ± 11.2 ng/mL; and 120 mg: 8.2 ± 5.2 vs 31.3 ± 14.8 ng/mL. For women, mean prolactin levels from baseline to study end were: placebo: 12.8 ± 11.7 vs 10.4 ± 8.0 ng/mL; 90 mg: 7.7 ± 5.3 vs 60.3 ± 46.9 ng/mL; and 120 mg: 10.9 ± 8.6 vs 85.5 ± 55.1 ng/mL. In the pivotal study, discontinuations due to adverse events were low across all treatment groups: 2.5% for placebo vs 0% for 90 mg and 1.7% for 120 mg.⁴ There was no single adverse reaction leading to discontinuation that occurred at a rate of ≥2% and greater than placebo in patients treated with RBP-7000.¹⁰ There were no clinically relevant differences in mean changes from baseline in corrected QT, QRS, and PR intervals, and in heart rate. Similarly, in the 12-month, long-term safety study, there were no clinically relevant changes in mean electrocardiography interval values from baseline to post-dose assessments.¹⁰

Using a 100-point visual analog scale (VAS), injection site pain scores 1 minute after the first dose decreased from a mean of 27 to the range of 3 to 7 for scores obtained 30 to 60 minutes post-dose. In the 12-month long-term safety study, 1-minute post-dose injection site pain VAS scores were highest on Day 1 (mean of 25) and decreased over time with subsequent injections (14 to 16 following last injection).¹⁰

Clinical considerations

Unique properties. RBP-7000 uses the established Atrigel system to provide effective antipsychotic levels in the first week of treatment, without the need for bridging oral coverage or a second loading injection. The abdominal subcutaneous injection volume is relatively small (0.6 mL or 0.8 mL).

Why Rx? The reasons to prescribe RBP-7000 for adult patients with schizophrenia include:

• no oral coverage required at the initiation of treatment
• effective plasma active moiety levels are seen within the first week without the need for a second loading injection
• monthly injection schedule.

Dosing. The recommended dosage of RBP-7000 is 90 mg or 120 mg once monthly, equivalent to 3 mg/d or 4 mg/d of oral risperidone, respectively. Oral risperidone tolerability should be established before the first injection. No oral risperidone coverage is required. RBP-7000 has not been studied in patients with renal or hepatic impairment and should be used with caution in these patients. Prior to initiating treatment in these patients, it is advised to carefully titrate up to at least 3 mg/d of oral risperidone. If a patient can tolerate 3 mg/d of oral risperidone and is psychiatrically stable, then the 90-mg dose of RBP-7000 can be considered.¹⁰ 

Contraindications. The only contraindications for RBP-7000 are known hypersensitivity to risperidone, paliperidone (9-OH risperidone), or other components of the injection.

Bottom Line

RBP-7000 (Perseris) is the second long-acting injectable (LAI) form of risperidone approved in the U.S. Unlike risperidone microspheres (Consta), RBP-7000 does not require any oral risperidone coverage at the beginning of therapy, provides effective drug levels within the first week of treatment with a single injection, and uses a monthly dosing interval. RBP-7000 does not require loading upon initiation. The monthly injection is <1 mL, is administered in abdominal subcutaneous tissue, and uses the Atrigel system.

Related Resource

• Carpenter J, Wong KK. Long-acting injectable antipsychotics: What to do about missed doses. Current Psychiatry. 2018;17(7):10-12,14-19,56.
   

Drug Brand Names
Aripiprazole • Abilify
Carbamazepine • Carbatrol, Tegretol
Doxycycline • Atridox
Leuprolide acetate injectable suspension • Eligard
Paliperidone palmitate • Invega Sustenna
Risperidone • Risperdal
Risperidone extended-release injectable suspension • Perseris
Risperidone long-acting injection • Risperdal Consta

Oral antipsychotic nonadherence is a significant contributor to relapse in patients with schizophrenia spectrum disorders. Long-acting injectable (LAI) antipsychotics have been developed to provide sustained antipsychotic exposure, with evidence that use of LAIs significantly reduces hospitalization rates.¹ One limiting factor in transitioning patients to certain LAIs is the need for prolonged oral coverage at the onset of treatment for agents that cannot be loaded. Nonadherence with this bridging oral therapy places the patient at risk for symptom exacerbation until effective antipsychotic plasma levels are achieved from the LAI.² Although risperidone is one of the more widely used antipsychotics for treating schizophrenia, until recently the only available LAI preparation, risperidone microspheres (Risperdal Consta), required 3 weeks of oral coverage upon initiation.³

Clinical implications

Oral medication nonadherence remains a significant public health issue for patients with schizophrenia, with an estimated 50% of patients failing to achieve 80% adherence even when enrolled in clinical trials specifically designed to track adherence.⁵ Although LAI atypical antipsychotics have been available since the approval of Risperdal Consta, the LAI form of risperidone, and both LAI forms of aripiprazole, were not designed to be loaded. A 1-day initiation regimen for aripiprazole lauroxil has been developed to avoid the need for 3 weeks of oral medication coverage,^6,7 but aripiprazole monohydrate and risperidone microspheres mandate oral bridging of 2 and 3 weeks, respectively.² Because one of the primary indications for LAI antipsychotic therapy is oral medication nonadherence, this prolonged period of oral coverage creates a risk for symptom exacerbation when the bridging period occurs outside of a controlled setting, as is common when patients are discharged from inpatient hospitalization.

One solution to this problem has its antecedents in the development of the Atrigel biodegradable injectable polymer, which was designed to deliver prolonged medication exposure after subcutaneous injection.⁸ This biodegradable polymer drug delivery system suspends and dissolves the medication of interest (in this case, risperidone) in a poly DL-lactide-coglycolide gel and its biocompatible carrier.⁹ The viscous liquid undergoes a phase transition upon contact with tissue fluids after subcutaneous injection, resulting in an implant that releases risperidone in a controlled manner as it is resorbed. Importantly, the kinetic parameters of RBP-7000 are such that effective drug levels are seen within the first week without the need for oral coverage.¹⁰

Use in adults with schizophrenia. After establishing tolerability with oral risperidone, the recommended doses are 90 mg or 120 mg monthly, which correspond to oral daily risperidone doses of 3 mg or 4 mg. RBP-7000 must be administered as a subcutaneous abdominal injection by a health care professional. It is recommended that the patient be in the supine position for the injection and that the injection sites be rotated monthly among 4 quadrants in the abdominal region. The injection volumes for the 90 mg and 120 mg doses are 0.6 mL and 0.8 mL, respectively.¹⁰ As the gel implant becomes firmer, the patient will notice a lump for several weeks that will decrease in size over time. Patients should be advised not to rub or massage the injection site, and to be aware of the placement of any belts or clothing with waistbands.¹⁰

Pharmacologic profile, adverse reactions

Risperidone is an atypical antipsychotic that has been commercially available in the U.S. since December 29, 1993, and its adverse effect profile is well characterized. The most common adverse effects associated with risperidone include those related to dopamine D2 antagonism, metabolic adverse effects, and an increase in serum prolactin. In the 12-month long-term safety study of RBP-7000, 1-minute post-dose injection site pain scores (on a 100-point scale) were highest on Day 1 (mean of 25) and decreased over time with subsequent injections (14 to 16 following the last injection).¹⁰

Continue to: How the Atrigel system works

The kinetics for RBP-7000 are markedly different than those for oral risperidone (Figure 2¹⁵). After a single subcutaneous injection, RBP-7000 shows 2 absorption peaks for risperidone. The first lower peak occurs with a T_max of 4 to 6 hours due to initial release of risperidone during the implant formation process; a second risperidone peak occurs after 10 to 14 days and is associated with slow release from the subcutaneous depot.^9,16,17 For both 9-OH risperidone levels and the total active moiety (risperidone plus 9-OH risperidone levels), the median T_max of the first peak ranges from 4 to 48 hours and the second peak ranges from 7 to 11 days. Following a single subcutaneous injection of RBP-7000, the apparent terminal half-life of risperidone ranges from 9 to 11 days, on average. The mean apparent terminal half-life of the active moiety ranges from 8 to 9 days.^9,16,17 Based on population pharmacokinetic modeling, the 90 mg and 120 mg doses of RBP-7000 are estimated to provide drug exposure equivalent to 3 mg/d and 4 mg/d of oral risperidone, respectively.^9,16,17

Continue to: Efficacy of RBP-7000

Efficacy of RBP-7000 was established in an 8-week, double-blind, placebo-controlled trial of adult patients experiencing an acute exacerbation of schizo­phrenia (age 18 to 55).⁴ Eligible participants had:

• An acute exacerbation of schizophrenia that occurred ≤8 weeks before the screening visit and would have benefited from psychiatric hospitalization or continued hospitalization
• Positive and Negative Syndrome Scale (PANSS) total score between 80 and 120 at visit 1 and a score of >4 on at least 2 of the following 4 items: hallucinatory behavior, delusions, conceptual disorganization, or suspiciousness/persecution
• The diagnosis of acute exacerbation of schizophrenia and PANSS total score were confirmed through an independent video-conference interview conducted by an experienced rater.

Participants were excluded if they:

• Experienced a ≥20% improvement in PANSS total score between the initial screening visit and the first injection
• had been treated at any time with clozapine for treatment-resistant schizophrenia
• had met DSM-IV-TR criteria for substance dependence (with the exception of nicotine or caffeine) before screening.

During the initial screening visit, participants received a 0.25-mg tablet of oral risperidone on 2 consecutive days to assess the tolerability of risperidone.

Outcome. Participants were randomized in a 1:1:1 manner to placebo (n = 112) or 1 of 2 monthly doses of RBP-7000: 90 mg (n = 111) or 120 mg (n = 114). Using the least squares means of repeated-measures changes from baseline in PANSS total scores, there was a significant improvement in the difference in PANSS total scores from baseline to the end of the study compared with placebo: 90-mg RBP-7000, -6.148 points (95% confidence interval [CI], -9.982 to -2.314, P = .0004); 120-mg RBP-7000, -7.237 points (95% CI, -11.045 to -3.429, P < .0001). The absolute change from baseline in PANSS total score was -15.367 points for the 90-mg dose and -16.456 points for the 120-mg dose.⁴ Completion rates across all 3 arms were comparable: placebo 70.6%, RBP-7000 90 mg 77.6%, and RBP-7000 120 mg 71.4%.

Tolerability. In the 8-week phase III efficacy trial of RBP-7000, adverse effects occurring with an incidence ≥5% and at least twice the rate of placebo were weight gain (placebo 3.4%, 90 mg 13.0%, 120 mg 12.8%) and sedation (placebo 0%, 90 mg 7.0%, 120 mg 7.7%).¹⁰ Compared with baseline, participants had a mean weight gain at the end of the study of 2.83 kg in the placebo group, 5.15 kg in the 90-mg RBP-7000 group, and 4.69 kg in the 120-mg RBP-7000 group. There were no clinically significant differences at study endpoint in glucose and lipid parameters. Consistent with the known effects of risperidone, there were increases in mean prolactin levels during the 8-week study, the effects of which were greater for women. For men, mean prolactin levels from baseline to study end were: placebo: 9.8 ± 7.9 vs 9.9 ± 8.0 ng/mL; 90 mg: 8.9 ± 6.9 vs 22.4 ± 11.2 ng/mL; and 120 mg: 8.2 ± 5.2 vs 31.3 ± 14.8 ng/mL. For women, mean prolactin levels from baseline to study end were: placebo: 12.8 ± 11.7 vs 10.4 ± 8.0 ng/mL; 90 mg: 7.7 ± 5.3 vs 60.3 ± 46.9 ng/mL; and 120 mg: 10.9 ± 8.6 vs 85.5 ± 55.1 ng/mL. In the pivotal study, discontinuations due to adverse events were low across all treatment groups: 2.5% for placebo vs 0% for 90 mg and 1.7% for 120 mg.⁴ There was no single adverse reaction leading to discontinuation that occurred at a rate of ≥2% and greater than placebo in patients treated with RBP-7000.¹⁰ There were no clinically relevant differences in mean changes from baseline in corrected QT, QRS, and PR intervals, and in heart rate. Similarly, in the 12-month, long-term safety study, there were no clinically relevant changes in mean electrocardiography interval values from baseline to post-dose assessments.¹⁰

Using a 100-point visual analog scale (VAS), injection site pain scores 1 minute after the first dose decreased from a mean of 27 to the range of 3 to 7 for scores obtained 30 to 60 minutes post-dose. In the 12-month long-term safety study, 1-minute post-dose injection site pain VAS scores were highest on Day 1 (mean of 25) and decreased over time with subsequent injections (14 to 16 following last injection).¹⁰

Clinical considerations

Unique properties. RBP-7000 uses the established Atrigel system to provide effective antipsychotic levels in the first week of treatment, without the need for bridging oral coverage or a second loading injection. The abdominal subcutaneous injection volume is relatively small (0.6 mL or 0.8 mL).

Why Rx? The reasons to prescribe RBP-7000 for adult patients with schizophrenia include:

• no oral coverage required at the initiation of treatment
• effective plasma active moiety levels are seen within the first week without the need for a second loading injection
• monthly injection schedule.

Dosing. The recommended dosage of RBP-7000 is 90 mg or 120 mg once monthly, equivalent to 3 mg/d or 4 mg/d of oral risperidone, respectively. Oral risperidone tolerability should be established before the first injection. No oral risperidone coverage is required. RBP-7000 has not been studied in patients with renal or hepatic impairment and should be used with caution in these patients. Prior to initiating treatment in these patients, it is advised to carefully titrate up to at least 3 mg/d of oral risperidone. If a patient can tolerate 3 mg/d of oral risperidone and is psychiatrically stable, then the 90-mg dose of RBP-7000 can be considered.¹⁰ 

Contraindications. The only contraindications for RBP-7000 are known hypersensitivity to risperidone, paliperidone (9-OH risperidone), or other components of the injection.

Bottom Line

RBP-7000 (Perseris) is the second long-acting injectable (LAI) form of risperidone approved in the U.S. Unlike risperidone microspheres (Consta), RBP-7000 does not require any oral risperidone coverage at the beginning of therapy, provides effective drug levels within the first week of treatment with a single injection, and uses a monthly dosing interval. RBP-7000 does not require loading upon initiation. The monthly injection is <1 mL, is administered in abdominal subcutaneous tissue, and uses the Atrigel system.

Related Resource

• Carpenter J, Wong KK. Long-acting injectable antipsychotics: What to do about missed doses. Current Psychiatry. 2018;17(7):10-12,14-19,56.
   

Drug Brand Names
Aripiprazole • Abilify
Carbamazepine • Carbatrol, Tegretol
Doxycycline • Atridox
Leuprolide acetate injectable suspension • Eligard
Paliperidone palmitate • Invega Sustenna
Risperidone • Risperdal
Risperidone extended-release injectable suspension • Perseris
Risperidone long-acting injection • Risperdal Consta

Oral antipsychotic nonadherence is a significant contributor to relapse in patients with schizophrenia spectrum disorders. Long-acting injectable (LAI) antipsychotics have been developed to provide sustained antipsychotic exposure, with evidence that use of LAIs significantly reduces hospitalization rates.¹ One limiting factor in transitioning patients to certain LAIs is the need for prolonged oral coverage at the onset of treatment for agents that cannot be loaded. Nonadherence with this bridging oral therapy places the patient at risk for symptom exacerbation until effective antipsychotic plasma levels are achieved from the LAI.² Although risperidone is one of the more widely used antipsychotics for treating schizophrenia, until recently the only available LAI preparation, risperidone microspheres (Risperdal Consta), required 3 weeks of oral coverage upon initiation.³

Clinical implications

Oral medication nonadherence remains a significant public health issue for patients with schizophrenia, with an estimated 50% of patients failing to achieve 80% adherence even when enrolled in clinical trials specifically designed to track adherence.⁵ Although LAI atypical antipsychotics have been available since the approval of Risperdal Consta, the LAI form of risperidone, and both LAI forms of aripiprazole, were not designed to be loaded. A 1-day initiation regimen for aripiprazole lauroxil has been developed to avoid the need for 3 weeks of oral medication coverage,^6,7 but aripiprazole monohydrate and risperidone microspheres mandate oral bridging of 2 and 3 weeks, respectively.² Because one of the primary indications for LAI antipsychotic therapy is oral medication nonadherence, this prolonged period of oral coverage creates a risk for symptom exacerbation when the bridging period occurs outside of a controlled setting, as is common when patients are discharged from inpatient hospitalization.

One solution to this problem has its antecedents in the development of the Atrigel biodegradable injectable polymer, which was designed to deliver prolonged medication exposure after subcutaneous injection.⁸ This biodegradable polymer drug delivery system suspends and dissolves the medication of interest (in this case, risperidone) in a poly DL-lactide-coglycolide gel and its biocompatible carrier.⁹ The viscous liquid undergoes a phase transition upon contact with tissue fluids after subcutaneous injection, resulting in an implant that releases risperidone in a controlled manner as it is resorbed. Importantly, the kinetic parameters of RBP-7000 are such that effective drug levels are seen within the first week without the need for oral coverage.¹⁰

Use in adults with schizophrenia. After establishing tolerability with oral risperidone, the recommended doses are 90 mg or 120 mg monthly, which correspond to oral daily risperidone doses of 3 mg or 4 mg. RBP-7000 must be administered as a subcutaneous abdominal injection by a health care professional. It is recommended that the patient be in the supine position for the injection and that the injection sites be rotated monthly among 4 quadrants in the abdominal region. The injection volumes for the 90 mg and 120 mg doses are 0.6 mL and 0.8 mL, respectively.¹⁰ As the gel implant becomes firmer, the patient will notice a lump for several weeks that will decrease in size over time. Patients should be advised not to rub or massage the injection site, and to be aware of the placement of any belts or clothing with waistbands.¹⁰

Pharmacologic profile, adverse reactions

Risperidone is an atypical antipsychotic that has been commercially available in the U.S. since December 29, 1993, and its adverse effect profile is well characterized. The most common adverse effects associated with risperidone include those related to dopamine D2 antagonism, metabolic adverse effects, and an increase in serum prolactin. In the 12-month long-term safety study of RBP-7000, 1-minute post-dose injection site pain scores (on a 100-point scale) were highest on Day 1 (mean of 25) and decreased over time with subsequent injections (14 to 16 following the last injection).¹⁰

Continue to: How the Atrigel system works

The kinetics for RBP-7000 are markedly different than those for oral risperidone (Figure 2¹⁵). After a single subcutaneous injection, RBP-7000 shows 2 absorption peaks for risperidone. The first lower peak occurs with a T_max of 4 to 6 hours due to initial release of risperidone during the implant formation process; a second risperidone peak occurs after 10 to 14 days and is associated with slow release from the subcutaneous depot.^9,16,17 For both 9-OH risperidone levels and the total active moiety (risperidone plus 9-OH risperidone levels), the median T_max of the first peak ranges from 4 to 48 hours and the second peak ranges from 7 to 11 days. Following a single subcutaneous injection of RBP-7000, the apparent terminal half-life of risperidone ranges from 9 to 11 days, on average. The mean apparent terminal half-life of the active moiety ranges from 8 to 9 days.^9,16,17 Based on population pharmacokinetic modeling, the 90 mg and 120 mg doses of RBP-7000 are estimated to provide drug exposure equivalent to 3 mg/d and 4 mg/d of oral risperidone, respectively.^9,16,17

Continue to: Efficacy of RBP-7000

Efficacy of RBP-7000 was established in an 8-week, double-blind, placebo-controlled trial of adult patients experiencing an acute exacerbation of schizo­phrenia (age 18 to 55).⁴ Eligible participants had:

• An acute exacerbation of schizophrenia that occurred ≤8 weeks before the screening visit and would have benefited from psychiatric hospitalization or continued hospitalization
• Positive and Negative Syndrome Scale (PANSS) total score between 80 and 120 at visit 1 and a score of >4 on at least 2 of the following 4 items: hallucinatory behavior, delusions, conceptual disorganization, or suspiciousness/persecution
• The diagnosis of acute exacerbation of schizophrenia and PANSS total score were confirmed through an independent video-conference interview conducted by an experienced rater.

Participants were excluded if they:

• Experienced a ≥20% improvement in PANSS total score between the initial screening visit and the first injection
• had been treated at any time with clozapine for treatment-resistant schizophrenia
• had met DSM-IV-TR criteria for substance dependence (with the exception of nicotine or caffeine) before screening.

During the initial screening visit, participants received a 0.25-mg tablet of oral risperidone on 2 consecutive days to assess the tolerability of risperidone.

Outcome. Participants were randomized in a 1:1:1 manner to placebo (n = 112) or 1 of 2 monthly doses of RBP-7000: 90 mg (n = 111) or 120 mg (n = 114). Using the least squares means of repeated-measures changes from baseline in PANSS total scores, there was a significant improvement in the difference in PANSS total scores from baseline to the end of the study compared with placebo: 90-mg RBP-7000, -6.148 points (95% confidence interval [CI], -9.982 to -2.314, P = .0004); 120-mg RBP-7000, -7.237 points (95% CI, -11.045 to -3.429, P < .0001). The absolute change from baseline in PANSS total score was -15.367 points for the 90-mg dose and -16.456 points for the 120-mg dose.⁴ Completion rates across all 3 arms were comparable: placebo 70.6%, RBP-7000 90 mg 77.6%, and RBP-7000 120 mg 71.4%.

Tolerability. In the 8-week phase III efficacy trial of RBP-7000, adverse effects occurring with an incidence ≥5% and at least twice the rate of placebo were weight gain (placebo 3.4%, 90 mg 13.0%, 120 mg 12.8%) and sedation (placebo 0%, 90 mg 7.0%, 120 mg 7.7%).¹⁰ Compared with baseline, participants had a mean weight gain at the end of the study of 2.83 kg in the placebo group, 5.15 kg in the 90-mg RBP-7000 group, and 4.69 kg in the 120-mg RBP-7000 group. There were no clinically significant differences at study endpoint in glucose and lipid parameters. Consistent with the known effects of risperidone, there were increases in mean prolactin levels during the 8-week study, the effects of which were greater for women. For men, mean prolactin levels from baseline to study end were: placebo: 9.8 ± 7.9 vs 9.9 ± 8.0 ng/mL; 90 mg: 8.9 ± 6.9 vs 22.4 ± 11.2 ng/mL; and 120 mg: 8.2 ± 5.2 vs 31.3 ± 14.8 ng/mL. For women, mean prolactin levels from baseline to study end were: placebo: 12.8 ± 11.7 vs 10.4 ± 8.0 ng/mL; 90 mg: 7.7 ± 5.3 vs 60.3 ± 46.9 ng/mL; and 120 mg: 10.9 ± 8.6 vs 85.5 ± 55.1 ng/mL. In the pivotal study, discontinuations due to adverse events were low across all treatment groups: 2.5% for placebo vs 0% for 90 mg and 1.7% for 120 mg.⁴ There was no single adverse reaction leading to discontinuation that occurred at a rate of ≥2% and greater than placebo in patients treated with RBP-7000.¹⁰ There were no clinically relevant differences in mean changes from baseline in corrected QT, QRS, and PR intervals, and in heart rate. Similarly, in the 12-month, long-term safety study, there were no clinically relevant changes in mean electrocardiography interval values from baseline to post-dose assessments.¹⁰

Using a 100-point visual analog scale (VAS), injection site pain scores 1 minute after the first dose decreased from a mean of 27 to the range of 3 to 7 for scores obtained 30 to 60 minutes post-dose. In the 12-month long-term safety study, 1-minute post-dose injection site pain VAS scores were highest on Day 1 (mean of 25) and decreased over time with subsequent injections (14 to 16 following last injection).¹⁰

Clinical considerations

Unique properties. RBP-7000 uses the established Atrigel system to provide effective antipsychotic levels in the first week of treatment, without the need for bridging oral coverage or a second loading injection. The abdominal subcutaneous injection volume is relatively small (0.6 mL or 0.8 mL).

Why Rx? The reasons to prescribe RBP-7000 for adult patients with schizophrenia include:

• no oral coverage required at the initiation of treatment
• effective plasma active moiety levels are seen within the first week without the need for a second loading injection
• monthly injection schedule.

Dosing. The recommended dosage of RBP-7000 is 90 mg or 120 mg once monthly, equivalent to 3 mg/d or 4 mg/d of oral risperidone, respectively. Oral risperidone tolerability should be established before the first injection. No oral risperidone coverage is required. RBP-7000 has not been studied in patients with renal or hepatic impairment and should be used with caution in these patients. Prior to initiating treatment in these patients, it is advised to carefully titrate up to at least 3 mg/d of oral risperidone. If a patient can tolerate 3 mg/d of oral risperidone and is psychiatrically stable, then the 90-mg dose of RBP-7000 can be considered.¹⁰ 

Contraindications. The only contraindications for RBP-7000 are known hypersensitivity to risperidone, paliperidone (9-OH risperidone), or other components of the injection.

Bottom Line

RBP-7000 (Perseris) is the second long-acting injectable (LAI) form of risperidone approved in the U.S. Unlike risperidone microspheres (Consta), RBP-7000 does not require any oral risperidone coverage at the beginning of therapy, provides effective drug levels within the first week of treatment with a single injection, and uses a monthly dosing interval. RBP-7000 does not require loading upon initiation. The monthly injection is <1 mL, is administered in abdominal subcutaneous tissue, and uses the Atrigel system.

Related Resource

• Carpenter J, Wong KK. Long-acting injectable antipsychotics: What to do about missed doses. Current Psychiatry. 2018;17(7):10-12,14-19,56.
   

Drug Brand Names
Aripiprazole • Abilify
Carbamazepine • Carbatrol, Tegretol
Doxycycline • Atridox
Leuprolide acetate injectable suspension • Eligard
Paliperidone palmitate • Invega Sustenna
Risperidone • Risperdal
Risperidone extended-release injectable suspension • Perseris
Risperidone long-acting injection • Risperdal Consta

Although urine drug screens (UDS) are most commonly used to screen job applicants, some clinicians have started to use them as a tool for improving their patients’ clinical outcomes.¹ Recently, some clinicians have begun using UDS to help patients who experience chronic pain and dependency (mainly on opioids) and for those who use diverted drugs to relieve these conditions. Many psychiatrists are concerned about the high cost of drug diversion, as well as the possibility of diversion-related patient mortality. Clinicians should therefore consider using UDS as a tool to help address these challenges.

Consider individualized UDS monitoring

The standard 5-substance UDS test panel consists of tetrahydrocannabinol, opiates, amphetamines, cocaine, and phencyclidine. Although this panel was sufficient for an employment screening-related UDS, the American Society of Addiction Medicine (ASAM) has rejected its use for patients with substance abuse. As part of its emphasis on the importance of incorporating preventative procedures, diagnostics, and surveillance protocols, the ASAM advocates using a rotating test panel in conjunction with a patient-specific UDS.² This type of patient-specific regimen would take into account the dynamic nature of a patient’s health profile factors, including comorbid and psychosocial status, subjective pain features, and diverted drug use. Furthermore, the ASAM recommends evaluating patients for the concurrent use of other substances and agents, such as benzodiazepines, sleep-inducing medications, stimulants, and alcohol, because these can interact with opioids.

Consider extending individualized monitoring by implementing standard “cutoff” values for each drug; patients whose levels of a specific substance are above the established cutoff value are categorized as testing positive for the use of that substance. The Substance Abuse Mental Health Services Administration favors adjusting UDS cutoffs, specifically the use of decreased cutoffs, to improve patient compliance.³ However, standardized drug concentration cutoff values may not be applicable for each patient; therefore, such values may need to be carefully tailored to each patient.

Additional drug monitoring techniques

Existing UDS practices, such as medication adherence and compliance, can be supplemented or alternately used with UDS panels that are modified to account for a patient’s fluctuating clinical conditions and concurrent medications. Point-of-care immunoassays, which provide accurate screening for medication compliance and adherence and possible drug diversion, should be used for routine monitoring. Using DNA-authenticated UDS also adds further control in monitoring a patient’s use of different drugs.^4,5

In addition to being helpful for monitoring opioid use, a DNA-verified UDS can be used to evaluate for the presence of synthetic urine substitutes.^6-8 Diversion remains a growing epidemiologic concern, and the number of cases is vastly underreported in the literature. The DNA-authenticated UDS can give clinicians greater precision in identifying synthetic and substituted urine among patient-provided samples.⁴

Using a combination of the methods described here can help expand a clinician’s ability to perform individualized drug monitoring, and verify whether a patient is adhering to his or her treatment regimen.

Although urine drug screens (UDS) are most commonly used to screen job applicants, some clinicians have started to use them as a tool for improving their patients’ clinical outcomes.¹ Recently, some clinicians have begun using UDS to help patients who experience chronic pain and dependency (mainly on opioids) and for those who use diverted drugs to relieve these conditions. Many psychiatrists are concerned about the high cost of drug diversion, as well as the possibility of diversion-related patient mortality. Clinicians should therefore consider using UDS as a tool to help address these challenges.

Consider individualized UDS monitoring

The standard 5-substance UDS test panel consists of tetrahydrocannabinol, opiates, amphetamines, cocaine, and phencyclidine. Although this panel was sufficient for an employment screening-related UDS, the American Society of Addiction Medicine (ASAM) has rejected its use for patients with substance abuse. As part of its emphasis on the importance of incorporating preventative procedures, diagnostics, and surveillance protocols, the ASAM advocates using a rotating test panel in conjunction with a patient-specific UDS.² This type of patient-specific regimen would take into account the dynamic nature of a patient’s health profile factors, including comorbid and psychosocial status, subjective pain features, and diverted drug use. Furthermore, the ASAM recommends evaluating patients for the concurrent use of other substances and agents, such as benzodiazepines, sleep-inducing medications, stimulants, and alcohol, because these can interact with opioids.

Consider extending individualized monitoring by implementing standard “cutoff” values for each drug; patients whose levels of a specific substance are above the established cutoff value are categorized as testing positive for the use of that substance. The Substance Abuse Mental Health Services Administration favors adjusting UDS cutoffs, specifically the use of decreased cutoffs, to improve patient compliance.³ However, standardized drug concentration cutoff values may not be applicable for each patient; therefore, such values may need to be carefully tailored to each patient.

Additional drug monitoring techniques

Existing UDS practices, such as medication adherence and compliance, can be supplemented or alternately used with UDS panels that are modified to account for a patient’s fluctuating clinical conditions and concurrent medications. Point-of-care immunoassays, which provide accurate screening for medication compliance and adherence and possible drug diversion, should be used for routine monitoring. Using DNA-authenticated UDS also adds further control in monitoring a patient’s use of different drugs.^4,5

In addition to being helpful for monitoring opioid use, a DNA-verified UDS can be used to evaluate for the presence of synthetic urine substitutes.^6-8 Diversion remains a growing epidemiologic concern, and the number of cases is vastly underreported in the literature. The DNA-authenticated UDS can give clinicians greater precision in identifying synthetic and substituted urine among patient-provided samples.⁴

Using a combination of the methods described here can help expand a clinician’s ability to perform individualized drug monitoring, and verify whether a patient is adhering to his or her treatment regimen.

Although urine drug screens (UDS) are most commonly used to screen job applicants, some clinicians have started to use them as a tool for improving their patients’ clinical outcomes.¹ Recently, some clinicians have begun using UDS to help patients who experience chronic pain and dependency (mainly on opioids) and for those who use diverted drugs to relieve these conditions. Many psychiatrists are concerned about the high cost of drug diversion, as well as the possibility of diversion-related patient mortality. Clinicians should therefore consider using UDS as a tool to help address these challenges.

Consider individualized UDS monitoring

The standard 5-substance UDS test panel consists of tetrahydrocannabinol, opiates, amphetamines, cocaine, and phencyclidine. Although this panel was sufficient for an employment screening-related UDS, the American Society of Addiction Medicine (ASAM) has rejected its use for patients with substance abuse. As part of its emphasis on the importance of incorporating preventative procedures, diagnostics, and surveillance protocols, the ASAM advocates using a rotating test panel in conjunction with a patient-specific UDS.² This type of patient-specific regimen would take into account the dynamic nature of a patient’s health profile factors, including comorbid and psychosocial status, subjective pain features, and diverted drug use. Furthermore, the ASAM recommends evaluating patients for the concurrent use of other substances and agents, such as benzodiazepines, sleep-inducing medications, stimulants, and alcohol, because these can interact with opioids.

Consider extending individualized monitoring by implementing standard “cutoff” values for each drug; patients whose levels of a specific substance are above the established cutoff value are categorized as testing positive for the use of that substance. The Substance Abuse Mental Health Services Administration favors adjusting UDS cutoffs, specifically the use of decreased cutoffs, to improve patient compliance.³ However, standardized drug concentration cutoff values may not be applicable for each patient; therefore, such values may need to be carefully tailored to each patient.

Additional drug monitoring techniques

Existing UDS practices, such as medication adherence and compliance, can be supplemented or alternately used with UDS panels that are modified to account for a patient’s fluctuating clinical conditions and concurrent medications. Point-of-care immunoassays, which provide accurate screening for medication compliance and adherence and possible drug diversion, should be used for routine monitoring. Using DNA-authenticated UDS also adds further control in monitoring a patient’s use of different drugs.^4,5

In addition to being helpful for monitoring opioid use, a DNA-verified UDS can be used to evaluate for the presence of synthetic urine substitutes.^6-8 Diversion remains a growing epidemiologic concern, and the number of cases is vastly underreported in the literature. The DNA-authenticated UDS can give clinicians greater precision in identifying synthetic and substituted urine among patient-provided samples.⁴

Using a combination of the methods described here can help expand a clinician’s ability to perform individualized drug monitoring, and verify whether a patient is adhering to his or her treatment regimen.

This centennial year update focuses primarily on immunization, but also reviews epidemiology, transmission, and treatment.

EPIDEMIOLOGY

2017–2018 was a bad season

The 2017–2018 influenza epidemic was memorable, dominated by influenza A(H3N2) viruses with morbidity and mortality rates approaching pandemic numbers. It lasted 19 weeks, killed more people than any other epidemic since 2010, particularly children, and was associated with 30,453 hospitalizations—almost twice the previous season high in some parts of the United States.²

Regrettably, 171 unvaccinated children died during 2017–2018, accounting for almost 80% of deaths.² The mean age of the children who died was 7.1 years; 51% had at least 1 underlying medical condition placing them at risk for influenza-related complications, and 57% died after hospitalization.²

Recent estimates of the incidence of symptomatic influenza among all ages ranged from 3% to 11%, which is slightly lower than historical estimates. The rates were higher for children under age 18 than for adults.³ Interestingly, influenza A(H3N2) accounted for 50% of cases of non-mumps viral parotitis during the 2014–2015 influenza season in the United States.⁴

Influenza C exists but is rare

Influenza A and B account for almost all influenza-related outpatient visits and hospitalizations. Surveillance data from May 2013 through December 2016 showed that influenza C accounts for 0.5% of influenza-related outpatient visits and hospitalizations, particularly affecting children ages 6 to 24 months. Medical comorbidities and copathogens were seen in all patients requiring intensive care and in most hospitalizations.⁵ Diagnostic tests for influenza C are not widely available.

Dogs and cats: Factories for new flu strains?

While pigs and birds are the major reservoirs of influenza viral genetic diversity from which infection is transmitted to humans, dogs and cats have recently emerged as possible sources of novel reassortant influenza A.⁶ With their frequent close contact with humans, our pets may prove to pose a significant threat.

Obesity a risk factor for influenza

Obesity emerged as a risk factor for severe influenza in the 2009 pandemic. Recent data also showed that obesity increases the duration of influenza A virus shedding, thus increasing duration of contagiousness.⁷

Influenza a cardiovascular risk factor

Previous data showed that influenza was a risk factor for cardiovascular events. Two recent epi­demiologic studies from the United Kingdom showed that laboratory-confirmed influenza was associated with higher rates of myocardial infarction and stroke for up to 4 weeks.^8,9

Which strain is the biggest threat?

Predicting which emerging influenza serotype may cause the next pandemic is difficult, but influenza A(H7N9), which had not infected humans until 2013 but has since infected about 1,600 people in China and killed 37% of them, appears to have the greatest potential.¹⁰

National influenza surveillance programs and influenza-related social media applications have been developed and may get a boost from technology. A smartphone equipped with a temperature sensor can instantly detect one’s temperature with great precision. A 2018 study suggested that a smartphone-driven thermometry application correlated well with national influenza-like illness activity and improved its forecast in real time and up to 3 weeks in advance.¹¹

TRANSMISSION

Humidity may not block transmission

Animal studies have suggested that humidity in the air interferes with transmission of airborne influenza virus, partially from biologic inactivation. But when a recent study used humidity-controlled chambers to investigate the stability of the 2009 influenza A(H1N1) virus in suspended aerosols and stationary droplets, the virus remained infectious in aerosols across a wide range of relative humidities, challenging the common belief that humidity destabilizes respiratory viruses in aerosols.¹²

One sick passenger may not infect the whole plane

Transmission of respiratory viruses on airplane flights has long been considered a potential avenue for spreading influenza. However, a recent study that monitored movements of individuals on 10 transcontinental US flights and simulated inflight transmission based on these data showed a low probability of direct transmission, except for passengers seated in close proximity to an infectious passenger.¹³

WHAT’S IN THE NEW FLU SHOT?

The 2018–2019 quadrivalent vaccine for the Northern Hemisphere¹⁴ contains the following strains:

• A/Michigan/45/2015 A(H1N1)pdm09-like virus
• A/Singapore/INFIMH-16-0019/2016 (H3N2)-like virus
• B/Colorado/06/2017-like virus (Victoria lineage)
• B/Phuket/3073/2013-like virus (Yamagata lineage).

The A(H3N2) (Singapore) and B/Victoria lineage components are new this year. The A(H3N2) strain was the main cause of the 2018 influenza epidemic in the Southern Hemisphere.

The quadrivalent live-attenuated vaccine, which was not recommended during the 2016–2017 and 2017–2018 influenza seasons, has made a comeback and is recommended for the 2018–2019 season in people for whom it is appropriate based on age and comorbidities.¹⁵ Although it was effective against influenza B and A(H3N2) viruses, it was less effective against the influenza A(H1N1)pdm09-like viruses during the 2013–2014 and 2015–2016 seasons.

A/Slovenia/2903/2015, the new A(H1N1)pdm09-like virus included in the 2018–2019 quadrivalent live-attenuated vaccine, is significantly more immunogenic than its predecessor, A/Bolivia/559/2013, but its clinical effectiveness remains to be seen.

How effective is it?

Influenza vaccine effectiveness in the 2017–2018 influenza season was 36% overall, 67% against A(H1N1), 42% against influenza B, and 25% against A(H3N2).¹⁶ It is estimated that influenza vaccine prevents 300 to 4,000 deaths annually in the United States alone.¹⁷

A 2018 Cochrane review¹⁷ concluded that vaccination reduced the incidence of influenza by about half, with 2.3% of the population contracting the flu without vaccination compared with 0.9% with vaccination (risk ratio 0.41, 95% confidence interval 0.36–0.47). The same review found that 71 healthy adults need to be vaccinated to prevent 1 from experiencing influenza, and 29 to prevent 1 influenza-like illness.

Several recent studies showed that influenza vaccine effectiveness varied based on age and influenza serotype, with higher effectiveness in people ages 5 to 17 and ages 18 to 64 than in those age 65 and older.^18–20 A mathematical model of influenza transmission and vaccination in the United States determined that even relatively low-efficacy influenza vaccines can be very useful if optimally distributed across age groups.²¹

Vaccination rates are low, and ‘antivaxxers’ are on the rise

Although the influenza vaccine is recommended in the United States for all people age 6 months and older regardless of the state of their health, vaccination rates remain low. In 2016, only 37% of employed adults were vaccinated. The highest rate was for government employees (45%), followed by private employees (36%), followed by the self-employed (30%).²²

A national goal is to immunize 80% of all Americans and 90% of at-risk populations (which include children and the elderly).²³ The number of US hospitals that require their employees to be vaccinated increased from 37.1% in 2013 to 61.4% in 2017.²⁴ Regrettably, as of March 2018, 14 lawsuits addressing religious objections to hospital influenza vaccination mandates have been filed.²⁵

Despite hundreds of studies demonstrating the efficacy, safety, and cost savings of influenza vaccination, the antivaccine movement has been growing in the United States and worldwide.²⁶ All US states except West Virginia, Mississippi, and California allow nonmedical exemptions from vaccination based on religious or personal belief.²⁷ Several US metropolitan areas represent “hot spots” for these exemptions.²⁸ This may render such areas vulnerable to vaccine-preventable diseases, including influenza.

Herd immunity: We’re all in this together

Some argue that the potential adverse effects and the cost of vaccination outweigh the benefits, but the protective benefits of herd immunity are significant for those with comorbidities or compromised immunity.

Educating the public about herd immunity and local influenza vaccination uptake increases people’s willingness to be vaccinated.²⁹ A key educational point is that at least 70% of a community needs to be vaccinated to prevent community outbreaks; this protects everyone, including those who do not mount a protective antibody response to influenza vaccination and those who are not vaccinated.

DOES ANNUAL VACCINATION BLUNT ITS EFFECTIVENESS?

Some studies from the 1970s and 1980s raised concern over a possible negative effect of annual influenza vaccination on vaccine effectiveness. The “antigenic distance hypothesis” holds that vaccine effectiveness is influenced by antigenic similarity between the previous season’s vaccine serotypes and the epidemic serotypes, as well as the antigenic similarity between the serotypes of the current and previous seasons.

A meta-analysis of studies from 2010 through 2015 showed significant inconsistencies in repeat vaccination effects within and between seasons and serotypes. It also showed that vaccine effectiveness may be influenced by more than 1 previous season, particularly for influenza A(H3N2), in which repeated vaccination can blunt the hemagglutinin antibody response.³⁰

A study from Japan showed that people who needed medical attention for influenza in the previous season were at lower risk of a similar event in the current season.³¹ Prior-season influenza vaccination reduced current-season vaccine effectiveness only in those who did not have medically attended influenza in the prior season. This suggests that infection is more immunogenic than vaccination, but only against the serotype causing the infection and not the other serotypes included in the vaccine.

An Australian study showed that annual influenza vaccination did not decrease vaccine effectiveness against influenza-associated hospitalization. Rather, effectiveness increased by about 15% in those vaccinated in both current and previous seasons compared with those vaccinated in either season alone.³²

European investigators showed that repeated seasonal influenza vaccination in the elderly prevented the need for hospitalization due to influenza A(H3N2) and B, but not A(H1N1)pdm09.³³

High-dose vaccine for older adults

The high-dose influenza vaccine has been licensed since 2009 for use in the United States for people ages 65 and older.

Recent studies confirmed that high-dose vaccine is more effective than standard-dose vaccine in veterans³⁴ and US Medicare beneficiaries.³⁵

The high-dose vaccine is rapidly becoming the primary vaccine given to people ages 65 and older in retail pharmacies, where vaccination begins earlier in the season than in providers’ offices.³⁶ Some studies have shown that the standard-dose vaccine wanes in effectiveness toward the end of the influenza season (particularly if the season is long) if it is given very early. It remains to be seen whether the same applies to the high-dose influenza vaccine.

Some advocate twice-annual influenza vaccination, particularly for older adults living in tropical and subtropical areas, where influenza seasons are more prolonged. However, a recently published study observed reductions in influenza-specific hemagglutination inhibition and cell-mediated immunity after twice-annual vaccination.³⁷

Vaccination is beneficial during pregnancy

Many studies have shown the value of influenza vaccination during pregnancy for both mothers and their infants.

One recently published study showed that 18% of infants who developed influenza required hospitalization.³⁸ In that study, prenatal and postpartum maternal influenza vaccination decreased the odds of influenza in infants by 61% and 53%, respectively.

Another study showed that vaccine effectiveness did not vary by gestational age at vaccination.³⁹

Some studies have shown that influenza virus infection can increase susceptibility to certain bacterial infections. A post hoc analysis of an influenza vaccination study in pregnant women suggested that the vaccine was also associated with decreased rates of pertussis in these women.⁴⁰

Factors that make vaccination less effective

Several factors including age-related frailty and iatrogenic and disease-related immunosuppression can affect vaccine effectiveness.

Frailty. A recent study showed that vaccine effectiveness was 77.6% in nonfrail older adults but only 58.7% in frail older adults.⁴¹

Immunosuppression. Temporary discontinuation of methotrexate for 2 weeks after influenza vaccination in patients with rheumatoid arthritis improves vaccine immunogenicity without precipitating disease flare.⁴² Solid-organ and hematopoietic stem cell transplant recipients who received influenza vaccine were less likely to develop pneumonia and require intensive care unit admission.⁴³

The high-dose influenza vaccine is more immunogenic than the standard-dose vaccine in solid-organ transplant recipients.⁴⁴

Statins are widely prescribed and have recently been associated with reduced influenza vaccine effectiveness against medically attended acute respiratory illness, but their benefits in preventing cardiovascular events outweigh this risk.⁴⁵

FUTURE VACCINE CONSIDERATIONS

Moving away from eggs

During the annual egg-based production process, which takes several months, the influenza vaccine acquires antigenic changes that allow replication in eggs, particularly in the hemagglutinin protein, which mediates receptor binding. This process of egg adaptation may cause antigenic changes that decrease vaccine effectiveness against circulating viruses.

The cell-based baculovirus influenza vaccine grown in dog kidney cells has higher antigenic content and is not subject to the limitations of egg-based vaccine, although it still requires annual updates. A recombinant influenza vaccine reduces the probability of influenza-like illness by 30% compared with the egg-based influenza vaccine, but also still requires annual updates.⁴⁶ The market share of these non-egg-based vaccines is small, and thus their effectiveness has yet to be demonstrated.

The US Department of Defense administered the cell-based influenza vaccine to about one-third of Armed Forces personnel, their families, and retirees in the 2017–2018 influenza seasons, and data on its effectiveness are expected in the near future.⁴⁷

A universal vaccine would be ideal

The quest continues for a universal influenza vaccine, one that remains protective for several years and does not require annual updates.⁴⁸ Such a vaccine would protect against seasonal epidemic influenza drift variants and pandemic strains. More people could likely be persuaded to be vaccinated once rather than every year.

David Schumick, medical illustrator; adapted from reference 49; illustration of the influenza hemagglutinin created in part with data from reference 50.

An ideal universal vaccine would be suitable for all age groups, at least 75% effective against symptomatic influenza virus infection, protective against all influenza A viruses (influenza A, not B, causes pandemics and seasonal epidemics), and durable through multiple influenza seasons.⁵¹

Research and production of such a vaccine are expected to require funding of about $1 billion over the next 5 years.

Boosting effectiveness

Estimates of influenza vaccine effectiveness range from 40% to 60% in years when the vaccine viruses closely match the circulating viruses, and variably lower when they do not match. The efficacy of most other vaccines given to prevent other infections is much higher.

New technologies to improve influenza vaccine effectiveness are needed, particularly for influenza A(H3N2) viruses, which are rapidly evolving and are highly susceptible to egg-adaptive mutations in the manufacturing process.

In one study, a nanoparticle vaccine formulated with a saponin-based adjuvant induced hemagglutination inhibition responses that were even greater than those induced by the high-dose vaccine.⁵²

Immunoglobulin A (IgA) may be a more effective vaccine target than traditional influenza vaccines that target IgG, since different parts of IgA may engage the influenza virus simultaneously.⁵³

Vaccines can be developed more quickly than in the past. The timeline from viral sequencing to human studies with deoxyribonucleic acid plasmid vaccines decreased from 20 months in 2003 for the severe acquired respiratory syndrome coronavirus to 11 months in 2006 for influenza A/Indonesia/2006 (H5), to 4 months in 2009 for influenza A/California/2009 (H1), to 3.5 months in 2016 for Zika virus.⁵⁴ This is because it is possible today to sequence a virus and insert the genetic material into a vaccine platform without ever having to grow the virus.

Numerous studies have found anti-influenza medications to be effective. Nevertheless, in an analysis of the 2011–2016 influenza seasons, only 15% of high-risk patients were prescribed anti-influenza medications within 2 days of symptom onset, including 37% in those with laboratory-confirmed influenza.⁵⁵ Fever was associated with an increased rate of antiviral treatment, but 25% of high-risk outpatients were afebrile. Empiric treatment of 4 high-risk outpatients with acute respiratory illness was needed to treat 1 patient with influenza.⁵⁵

Treatment with a neuraminidase inhibitor within 2 days of illness has recently been shown to improve survival and shorten duration of viral shedding in patients with avian influenza A(H7N9) infection.⁵⁶ Antiviral treatment within 2 days of illness is associated with improved outcomes in transplant recipients⁵⁷ and with a lower risk of otitis media in children.⁵⁸

Appropriate anti-influenza treatment is as important as avoiding unnecessary antibiotics. Regrettably, as many as one-third of patients with laboratory-confirmed influenza are prescribed antibiotics.⁵⁹

The US Food and Drug Administration warns against fraudulent unapproved over-the-counter influenza products.⁶⁰

Baloxavir marboxil

Baloxavir marboxil is a new anti-influenza medication approved in Japan in February 2018 and anticipated to be available in the United States sometime in 2019.

This prodrug is hydrolyzed in vivo to the active metabolite, which selectively inhibits cap-dependent endonuclease enzyme, a key enzyme in initiation of messenger ribonucleic acid synthesis required for influenza viral replication.⁶¹

In a double-blind phase 3 trial, the median time to alleviation of influenza symptoms is 26.5 hours shorter with baloxavir marboxil than with placebo. One tablet was as effective as 5 days of the neuraminidase inhibitor oseltamivir and was associated with greater reduction in viral load 1 day after initiation, and similar side effects.⁶² Of concern is the emergence of nucleic acid substitutions conferring resistance to baloxavir; this occurred in 2.2% and 9.7% of baloxavir recipients in the phase 2 and 3 trials, respectively.

CLOSING THE GAPS

Several gaps in the management of influenza persist since the 1918 pandemic.¹ These include gaps in epidemiology, prevention, diagnosis, treatment, and prognosis.

• Global networks wider than current ones are needed to address this global disease and to prioritize coordination efforts.
• Establishing and strengthening clinical capacity is needed in limited resource settings. New technologies are needed to expedite vaccine development and to achieve progress toward a universal vaccine.
• Current diagnostic tests do not distinguish between seasonal and novel influenza A viruses of zoonotic origin, which are expected to cause the next pandemic.
• Current antivirals have been shown to shorten duration of illness in outpatients with uncomplicated influenza, but the benefit in hospitalized patients has been less well established.
• In 2007, resistance of seasonal influenza A(H1N1) to oseltamivir became widespread. In 2009, pandemic influenza A(H1N1), which is highly susceptible to oseltamivir, replaced the seasonal virus and remains the predominantly circulating A(H1N1) strain.
• A small-molecule fragment, N-cyclohexyaltaurine, binds to the conserved hemagglutinin receptor-binding site in a manner that mimics the binding mode of the natural receptor sialic acid. This can serve as a template to guide the development of novel broad-spectrum small-molecule anti-influenza drugs.⁶³
• Biomarkers that can accurately predict development of severe disease in patients with influenza are needed.

This centennial year update focuses primarily on immunization, but also reviews epidemiology, transmission, and treatment.

EPIDEMIOLOGY

2017–2018 was a bad season

The 2017–2018 influenza epidemic was memorable, dominated by influenza A(H3N2) viruses with morbidity and mortality rates approaching pandemic numbers. It lasted 19 weeks, killed more people than any other epidemic since 2010, particularly children, and was associated with 30,453 hospitalizations—almost twice the previous season high in some parts of the United States.²

Regrettably, 171 unvaccinated children died during 2017–2018, accounting for almost 80% of deaths.² The mean age of the children who died was 7.1 years; 51% had at least 1 underlying medical condition placing them at risk for influenza-related complications, and 57% died after hospitalization.²

Recent estimates of the incidence of symptomatic influenza among all ages ranged from 3% to 11%, which is slightly lower than historical estimates. The rates were higher for children under age 18 than for adults.³ Interestingly, influenza A(H3N2) accounted for 50% of cases of non-mumps viral parotitis during the 2014–2015 influenza season in the United States.⁴

Influenza C exists but is rare

Influenza A and B account for almost all influenza-related outpatient visits and hospitalizations. Surveillance data from May 2013 through December 2016 showed that influenza C accounts for 0.5% of influenza-related outpatient visits and hospitalizations, particularly affecting children ages 6 to 24 months. Medical comorbidities and copathogens were seen in all patients requiring intensive care and in most hospitalizations.⁵ Diagnostic tests for influenza C are not widely available.

Dogs and cats: Factories for new flu strains?

While pigs and birds are the major reservoirs of influenza viral genetic diversity from which infection is transmitted to humans, dogs and cats have recently emerged as possible sources of novel reassortant influenza A.⁶ With their frequent close contact with humans, our pets may prove to pose a significant threat.

Obesity a risk factor for influenza

Obesity emerged as a risk factor for severe influenza in the 2009 pandemic. Recent data also showed that obesity increases the duration of influenza A virus shedding, thus increasing duration of contagiousness.⁷

Influenza a cardiovascular risk factor

Previous data showed that influenza was a risk factor for cardiovascular events. Two recent epi­demiologic studies from the United Kingdom showed that laboratory-confirmed influenza was associated with higher rates of myocardial infarction and stroke for up to 4 weeks.^8,9

Which strain is the biggest threat?

Predicting which emerging influenza serotype may cause the next pandemic is difficult, but influenza A(H7N9), which had not infected humans until 2013 but has since infected about 1,600 people in China and killed 37% of them, appears to have the greatest potential.¹⁰

National influenza surveillance programs and influenza-related social media applications have been developed and may get a boost from technology. A smartphone equipped with a temperature sensor can instantly detect one’s temperature with great precision. A 2018 study suggested that a smartphone-driven thermometry application correlated well with national influenza-like illness activity and improved its forecast in real time and up to 3 weeks in advance.¹¹

TRANSMISSION

Humidity may not block transmission

Animal studies have suggested that humidity in the air interferes with transmission of airborne influenza virus, partially from biologic inactivation. But when a recent study used humidity-controlled chambers to investigate the stability of the 2009 influenza A(H1N1) virus in suspended aerosols and stationary droplets, the virus remained infectious in aerosols across a wide range of relative humidities, challenging the common belief that humidity destabilizes respiratory viruses in aerosols.¹²

One sick passenger may not infect the whole plane

Transmission of respiratory viruses on airplane flights has long been considered a potential avenue for spreading influenza. However, a recent study that monitored movements of individuals on 10 transcontinental US flights and simulated inflight transmission based on these data showed a low probability of direct transmission, except for passengers seated in close proximity to an infectious passenger.¹³

WHAT’S IN THE NEW FLU SHOT?

The 2018–2019 quadrivalent vaccine for the Northern Hemisphere¹⁴ contains the following strains:

• A/Michigan/45/2015 A(H1N1)pdm09-like virus
• A/Singapore/INFIMH-16-0019/2016 (H3N2)-like virus
• B/Colorado/06/2017-like virus (Victoria lineage)
• B/Phuket/3073/2013-like virus (Yamagata lineage).

The A(H3N2) (Singapore) and B/Victoria lineage components are new this year. The A(H3N2) strain was the main cause of the 2018 influenza epidemic in the Southern Hemisphere.

The quadrivalent live-attenuated vaccine, which was not recommended during the 2016–2017 and 2017–2018 influenza seasons, has made a comeback and is recommended for the 2018–2019 season in people for whom it is appropriate based on age and comorbidities.¹⁵ Although it was effective against influenza B and A(H3N2) viruses, it was less effective against the influenza A(H1N1)pdm09-like viruses during the 2013–2014 and 2015–2016 seasons.

A/Slovenia/2903/2015, the new A(H1N1)pdm09-like virus included in the 2018–2019 quadrivalent live-attenuated vaccine, is significantly more immunogenic than its predecessor, A/Bolivia/559/2013, but its clinical effectiveness remains to be seen.

How effective is it?

Influenza vaccine effectiveness in the 2017–2018 influenza season was 36% overall, 67% against A(H1N1), 42% against influenza B, and 25% against A(H3N2).¹⁶ It is estimated that influenza vaccine prevents 300 to 4,000 deaths annually in the United States alone.¹⁷

A 2018 Cochrane review¹⁷ concluded that vaccination reduced the incidence of influenza by about half, with 2.3% of the population contracting the flu without vaccination compared with 0.9% with vaccination (risk ratio 0.41, 95% confidence interval 0.36–0.47). The same review found that 71 healthy adults need to be vaccinated to prevent 1 from experiencing influenza, and 29 to prevent 1 influenza-like illness.

Several recent studies showed that influenza vaccine effectiveness varied based on age and influenza serotype, with higher effectiveness in people ages 5 to 17 and ages 18 to 64 than in those age 65 and older.^18–20 A mathematical model of influenza transmission and vaccination in the United States determined that even relatively low-efficacy influenza vaccines can be very useful if optimally distributed across age groups.²¹

Vaccination rates are low, and ‘antivaxxers’ are on the rise

Although the influenza vaccine is recommended in the United States for all people age 6 months and older regardless of the state of their health, vaccination rates remain low. In 2016, only 37% of employed adults were vaccinated. The highest rate was for government employees (45%), followed by private employees (36%), followed by the self-employed (30%).²²

A national goal is to immunize 80% of all Americans and 90% of at-risk populations (which include children and the elderly).²³ The number of US hospitals that require their employees to be vaccinated increased from 37.1% in 2013 to 61.4% in 2017.²⁴ Regrettably, as of March 2018, 14 lawsuits addressing religious objections to hospital influenza vaccination mandates have been filed.²⁵

Despite hundreds of studies demonstrating the efficacy, safety, and cost savings of influenza vaccination, the antivaccine movement has been growing in the United States and worldwide.²⁶ All US states except West Virginia, Mississippi, and California allow nonmedical exemptions from vaccination based on religious or personal belief.²⁷ Several US metropolitan areas represent “hot spots” for these exemptions.²⁸ This may render such areas vulnerable to vaccine-preventable diseases, including influenza.

Herd immunity: We’re all in this together

Some argue that the potential adverse effects and the cost of vaccination outweigh the benefits, but the protective benefits of herd immunity are significant for those with comorbidities or compromised immunity.

Educating the public about herd immunity and local influenza vaccination uptake increases people’s willingness to be vaccinated.²⁹ A key educational point is that at least 70% of a community needs to be vaccinated to prevent community outbreaks; this protects everyone, including those who do not mount a protective antibody response to influenza vaccination and those who are not vaccinated.

DOES ANNUAL VACCINATION BLUNT ITS EFFECTIVENESS?

Some studies from the 1970s and 1980s raised concern over a possible negative effect of annual influenza vaccination on vaccine effectiveness. The “antigenic distance hypothesis” holds that vaccine effectiveness is influenced by antigenic similarity between the previous season’s vaccine serotypes and the epidemic serotypes, as well as the antigenic similarity between the serotypes of the current and previous seasons.

A meta-analysis of studies from 2010 through 2015 showed significant inconsistencies in repeat vaccination effects within and between seasons and serotypes. It also showed that vaccine effectiveness may be influenced by more than 1 previous season, particularly for influenza A(H3N2), in which repeated vaccination can blunt the hemagglutinin antibody response.³⁰

A study from Japan showed that people who needed medical attention for influenza in the previous season were at lower risk of a similar event in the current season.³¹ Prior-season influenza vaccination reduced current-season vaccine effectiveness only in those who did not have medically attended influenza in the prior season. This suggests that infection is more immunogenic than vaccination, but only against the serotype causing the infection and not the other serotypes included in the vaccine.

An Australian study showed that annual influenza vaccination did not decrease vaccine effectiveness against influenza-associated hospitalization. Rather, effectiveness increased by about 15% in those vaccinated in both current and previous seasons compared with those vaccinated in either season alone.³²

European investigators showed that repeated seasonal influenza vaccination in the elderly prevented the need for hospitalization due to influenza A(H3N2) and B, but not A(H1N1)pdm09.³³

High-dose vaccine for older adults

The high-dose influenza vaccine has been licensed since 2009 for use in the United States for people ages 65 and older.

Recent studies confirmed that high-dose vaccine is more effective than standard-dose vaccine in veterans³⁴ and US Medicare beneficiaries.³⁵

The high-dose vaccine is rapidly becoming the primary vaccine given to people ages 65 and older in retail pharmacies, where vaccination begins earlier in the season than in providers’ offices.³⁶ Some studies have shown that the standard-dose vaccine wanes in effectiveness toward the end of the influenza season (particularly if the season is long) if it is given very early. It remains to be seen whether the same applies to the high-dose influenza vaccine.

Some advocate twice-annual influenza vaccination, particularly for older adults living in tropical and subtropical areas, where influenza seasons are more prolonged. However, a recently published study observed reductions in influenza-specific hemagglutination inhibition and cell-mediated immunity after twice-annual vaccination.³⁷

Vaccination is beneficial during pregnancy

Many studies have shown the value of influenza vaccination during pregnancy for both mothers and their infants.

One recently published study showed that 18% of infants who developed influenza required hospitalization.³⁸ In that study, prenatal and postpartum maternal influenza vaccination decreased the odds of influenza in infants by 61% and 53%, respectively.

Another study showed that vaccine effectiveness did not vary by gestational age at vaccination.³⁹

Some studies have shown that influenza virus infection can increase susceptibility to certain bacterial infections. A post hoc analysis of an influenza vaccination study in pregnant women suggested that the vaccine was also associated with decreased rates of pertussis in these women.⁴⁰

Factors that make vaccination less effective

Several factors including age-related frailty and iatrogenic and disease-related immunosuppression can affect vaccine effectiveness.

Frailty. A recent study showed that vaccine effectiveness was 77.6% in nonfrail older adults but only 58.7% in frail older adults.⁴¹

Immunosuppression. Temporary discontinuation of methotrexate for 2 weeks after influenza vaccination in patients with rheumatoid arthritis improves vaccine immunogenicity without precipitating disease flare.⁴² Solid-organ and hematopoietic stem cell transplant recipients who received influenza vaccine were less likely to develop pneumonia and require intensive care unit admission.⁴³

The high-dose influenza vaccine is more immunogenic than the standard-dose vaccine in solid-organ transplant recipients.⁴⁴

Statins are widely prescribed and have recently been associated with reduced influenza vaccine effectiveness against medically attended acute respiratory illness, but their benefits in preventing cardiovascular events outweigh this risk.⁴⁵

FUTURE VACCINE CONSIDERATIONS

Moving away from eggs

During the annual egg-based production process, which takes several months, the influenza vaccine acquires antigenic changes that allow replication in eggs, particularly in the hemagglutinin protein, which mediates receptor binding. This process of egg adaptation may cause antigenic changes that decrease vaccine effectiveness against circulating viruses.

The cell-based baculovirus influenza vaccine grown in dog kidney cells has higher antigenic content and is not subject to the limitations of egg-based vaccine, although it still requires annual updates. A recombinant influenza vaccine reduces the probability of influenza-like illness by 30% compared with the egg-based influenza vaccine, but also still requires annual updates.⁴⁶ The market share of these non-egg-based vaccines is small, and thus their effectiveness has yet to be demonstrated.

The US Department of Defense administered the cell-based influenza vaccine to about one-third of Armed Forces personnel, their families, and retirees in the 2017–2018 influenza seasons, and data on its effectiveness are expected in the near future.⁴⁷

A universal vaccine would be ideal

The quest continues for a universal influenza vaccine, one that remains protective for several years and does not require annual updates.⁴⁸ Such a vaccine would protect against seasonal epidemic influenza drift variants and pandemic strains. More people could likely be persuaded to be vaccinated once rather than every year.

David Schumick, medical illustrator; adapted from reference 49; illustration of the influenza hemagglutinin created in part with data from reference 50.

An ideal universal vaccine would be suitable for all age groups, at least 75% effective against symptomatic influenza virus infection, protective against all influenza A viruses (influenza A, not B, causes pandemics and seasonal epidemics), and durable through multiple influenza seasons.⁵¹

Research and production of such a vaccine are expected to require funding of about $1 billion over the next 5 years.

Boosting effectiveness

Estimates of influenza vaccine effectiveness range from 40% to 60% in years when the vaccine viruses closely match the circulating viruses, and variably lower when they do not match. The efficacy of most other vaccines given to prevent other infections is much higher.

New technologies to improve influenza vaccine effectiveness are needed, particularly for influenza A(H3N2) viruses, which are rapidly evolving and are highly susceptible to egg-adaptive mutations in the manufacturing process.

In one study, a nanoparticle vaccine formulated with a saponin-based adjuvant induced hemagglutination inhibition responses that were even greater than those induced by the high-dose vaccine.⁵²

Immunoglobulin A (IgA) may be a more effective vaccine target than traditional influenza vaccines that target IgG, since different parts of IgA may engage the influenza virus simultaneously.⁵³

Vaccines can be developed more quickly than in the past. The timeline from viral sequencing to human studies with deoxyribonucleic acid plasmid vaccines decreased from 20 months in 2003 for the severe acquired respiratory syndrome coronavirus to 11 months in 2006 for influenza A/Indonesia/2006 (H5), to 4 months in 2009 for influenza A/California/2009 (H1), to 3.5 months in 2016 for Zika virus.⁵⁴ This is because it is possible today to sequence a virus and insert the genetic material into a vaccine platform without ever having to grow the virus.

Numerous studies have found anti-influenza medications to be effective. Nevertheless, in an analysis of the 2011–2016 influenza seasons, only 15% of high-risk patients were prescribed anti-influenza medications within 2 days of symptom onset, including 37% in those with laboratory-confirmed influenza.⁵⁵ Fever was associated with an increased rate of antiviral treatment, but 25% of high-risk outpatients were afebrile. Empiric treatment of 4 high-risk outpatients with acute respiratory illness was needed to treat 1 patient with influenza.⁵⁵

Treatment with a neuraminidase inhibitor within 2 days of illness has recently been shown to improve survival and shorten duration of viral shedding in patients with avian influenza A(H7N9) infection.⁵⁶ Antiviral treatment within 2 days of illness is associated with improved outcomes in transplant recipients⁵⁷ and with a lower risk of otitis media in children.⁵⁸

Appropriate anti-influenza treatment is as important as avoiding unnecessary antibiotics. Regrettably, as many as one-third of patients with laboratory-confirmed influenza are prescribed antibiotics.⁵⁹

The US Food and Drug Administration warns against fraudulent unapproved over-the-counter influenza products.⁶⁰

Baloxavir marboxil

Baloxavir marboxil is a new anti-influenza medication approved in Japan in February 2018 and anticipated to be available in the United States sometime in 2019.

This prodrug is hydrolyzed in vivo to the active metabolite, which selectively inhibits cap-dependent endonuclease enzyme, a key enzyme in initiation of messenger ribonucleic acid synthesis required for influenza viral replication.⁶¹

In a double-blind phase 3 trial, the median time to alleviation of influenza symptoms is 26.5 hours shorter with baloxavir marboxil than with placebo. One tablet was as effective as 5 days of the neuraminidase inhibitor oseltamivir and was associated with greater reduction in viral load 1 day after initiation, and similar side effects.⁶² Of concern is the emergence of nucleic acid substitutions conferring resistance to baloxavir; this occurred in 2.2% and 9.7% of baloxavir recipients in the phase 2 and 3 trials, respectively.

CLOSING THE GAPS

Several gaps in the management of influenza persist since the 1918 pandemic.¹ These include gaps in epidemiology, prevention, diagnosis, treatment, and prognosis.

• Global networks wider than current ones are needed to address this global disease and to prioritize coordination efforts.
• Establishing and strengthening clinical capacity is needed in limited resource settings. New technologies are needed to expedite vaccine development and to achieve progress toward a universal vaccine.
• Current diagnostic tests do not distinguish between seasonal and novel influenza A viruses of zoonotic origin, which are expected to cause the next pandemic.
• Current antivirals have been shown to shorten duration of illness in outpatients with uncomplicated influenza, but the benefit in hospitalized patients has been less well established.
• In 2007, resistance of seasonal influenza A(H1N1) to oseltamivir became widespread. In 2009, pandemic influenza A(H1N1), which is highly susceptible to oseltamivir, replaced the seasonal virus and remains the predominantly circulating A(H1N1) strain.
• A small-molecule fragment, N-cyclohexyaltaurine, binds to the conserved hemagglutinin receptor-binding site in a manner that mimics the binding mode of the natural receptor sialic acid. This can serve as a template to guide the development of novel broad-spectrum small-molecule anti-influenza drugs.⁶³
• Biomarkers that can accurately predict development of severe disease in patients with influenza are needed.

This centennial year update focuses primarily on immunization, but also reviews epidemiology, transmission, and treatment.

EPIDEMIOLOGY

2017–2018 was a bad season

The 2017–2018 influenza epidemic was memorable, dominated by influenza A(H3N2) viruses with morbidity and mortality rates approaching pandemic numbers. It lasted 19 weeks, killed more people than any other epidemic since 2010, particularly children, and was associated with 30,453 hospitalizations—almost twice the previous season high in some parts of the United States.²

Regrettably, 171 unvaccinated children died during 2017–2018, accounting for almost 80% of deaths.² The mean age of the children who died was 7.1 years; 51% had at least 1 underlying medical condition placing them at risk for influenza-related complications, and 57% died after hospitalization.²

Recent estimates of the incidence of symptomatic influenza among all ages ranged from 3% to 11%, which is slightly lower than historical estimates. The rates were higher for children under age 18 than for adults.³ Interestingly, influenza A(H3N2) accounted for 50% of cases of non-mumps viral parotitis during the 2014–2015 influenza season in the United States.⁴

Influenza C exists but is rare

Influenza A and B account for almost all influenza-related outpatient visits and hospitalizations. Surveillance data from May 2013 through December 2016 showed that influenza C accounts for 0.5% of influenza-related outpatient visits and hospitalizations, particularly affecting children ages 6 to 24 months. Medical comorbidities and copathogens were seen in all patients requiring intensive care and in most hospitalizations.⁵ Diagnostic tests for influenza C are not widely available.

Dogs and cats: Factories for new flu strains?

While pigs and birds are the major reservoirs of influenza viral genetic diversity from which infection is transmitted to humans, dogs and cats have recently emerged as possible sources of novel reassortant influenza A.⁶ With their frequent close contact with humans, our pets may prove to pose a significant threat.

Obesity a risk factor for influenza

Obesity emerged as a risk factor for severe influenza in the 2009 pandemic. Recent data also showed that obesity increases the duration of influenza A virus shedding, thus increasing duration of contagiousness.⁷

Influenza a cardiovascular risk factor

Previous data showed that influenza was a risk factor for cardiovascular events. Two recent epi­demiologic studies from the United Kingdom showed that laboratory-confirmed influenza was associated with higher rates of myocardial infarction and stroke for up to 4 weeks.^8,9

Which strain is the biggest threat?

Predicting which emerging influenza serotype may cause the next pandemic is difficult, but influenza A(H7N9), which had not infected humans until 2013 but has since infected about 1,600 people in China and killed 37% of them, appears to have the greatest potential.¹⁰

National influenza surveillance programs and influenza-related social media applications have been developed and may get a boost from technology. A smartphone equipped with a temperature sensor can instantly detect one’s temperature with great precision. A 2018 study suggested that a smartphone-driven thermometry application correlated well with national influenza-like illness activity and improved its forecast in real time and up to 3 weeks in advance.¹¹

TRANSMISSION

Humidity may not block transmission

Animal studies have suggested that humidity in the air interferes with transmission of airborne influenza virus, partially from biologic inactivation. But when a recent study used humidity-controlled chambers to investigate the stability of the 2009 influenza A(H1N1) virus in suspended aerosols and stationary droplets, the virus remained infectious in aerosols across a wide range of relative humidities, challenging the common belief that humidity destabilizes respiratory viruses in aerosols.¹²

One sick passenger may not infect the whole plane

Transmission of respiratory viruses on airplane flights has long been considered a potential avenue for spreading influenza. However, a recent study that monitored movements of individuals on 10 transcontinental US flights and simulated inflight transmission based on these data showed a low probability of direct transmission, except for passengers seated in close proximity to an infectious passenger.¹³

WHAT’S IN THE NEW FLU SHOT?

The 2018–2019 quadrivalent vaccine for the Northern Hemisphere¹⁴ contains the following strains:

• A/Michigan/45/2015 A(H1N1)pdm09-like virus
• A/Singapore/INFIMH-16-0019/2016 (H3N2)-like virus
• B/Colorado/06/2017-like virus (Victoria lineage)
• B/Phuket/3073/2013-like virus (Yamagata lineage).

The A(H3N2) (Singapore) and B/Victoria lineage components are new this year. The A(H3N2) strain was the main cause of the 2018 influenza epidemic in the Southern Hemisphere.

The quadrivalent live-attenuated vaccine, which was not recommended during the 2016–2017 and 2017–2018 influenza seasons, has made a comeback and is recommended for the 2018–2019 season in people for whom it is appropriate based on age and comorbidities.¹⁵ Although it was effective against influenza B and A(H3N2) viruses, it was less effective against the influenza A(H1N1)pdm09-like viruses during the 2013–2014 and 2015–2016 seasons.

A/Slovenia/2903/2015, the new A(H1N1)pdm09-like virus included in the 2018–2019 quadrivalent live-attenuated vaccine, is significantly more immunogenic than its predecessor, A/Bolivia/559/2013, but its clinical effectiveness remains to be seen.

How effective is it?

Influenza vaccine effectiveness in the 2017–2018 influenza season was 36% overall, 67% against A(H1N1), 42% against influenza B, and 25% against A(H3N2).¹⁶ It is estimated that influenza vaccine prevents 300 to 4,000 deaths annually in the United States alone.¹⁷

A 2018 Cochrane review¹⁷ concluded that vaccination reduced the incidence of influenza by about half, with 2.3% of the population contracting the flu without vaccination compared with 0.9% with vaccination (risk ratio 0.41, 95% confidence interval 0.36–0.47). The same review found that 71 healthy adults need to be vaccinated to prevent 1 from experiencing influenza, and 29 to prevent 1 influenza-like illness.

Several recent studies showed that influenza vaccine effectiveness varied based on age and influenza serotype, with higher effectiveness in people ages 5 to 17 and ages 18 to 64 than in those age 65 and older.^18–20 A mathematical model of influenza transmission and vaccination in the United States determined that even relatively low-efficacy influenza vaccines can be very useful if optimally distributed across age groups.²¹

Vaccination rates are low, and ‘antivaxxers’ are on the rise

Although the influenza vaccine is recommended in the United States for all people age 6 months and older regardless of the state of their health, vaccination rates remain low. In 2016, only 37% of employed adults were vaccinated. The highest rate was for government employees (45%), followed by private employees (36%), followed by the self-employed (30%).²²

A national goal is to immunize 80% of all Americans and 90% of at-risk populations (which include children and the elderly).²³ The number of US hospitals that require their employees to be vaccinated increased from 37.1% in 2013 to 61.4% in 2017.²⁴ Regrettably, as of March 2018, 14 lawsuits addressing religious objections to hospital influenza vaccination mandates have been filed.²⁵

Despite hundreds of studies demonstrating the efficacy, safety, and cost savings of influenza vaccination, the antivaccine movement has been growing in the United States and worldwide.²⁶ All US states except West Virginia, Mississippi, and California allow nonmedical exemptions from vaccination based on religious or personal belief.²⁷ Several US metropolitan areas represent “hot spots” for these exemptions.²⁸ This may render such areas vulnerable to vaccine-preventable diseases, including influenza.

Herd immunity: We’re all in this together

Some argue that the potential adverse effects and the cost of vaccination outweigh the benefits, but the protective benefits of herd immunity are significant for those with comorbidities or compromised immunity.

Educating the public about herd immunity and local influenza vaccination uptake increases people’s willingness to be vaccinated.²⁹ A key educational point is that at least 70% of a community needs to be vaccinated to prevent community outbreaks; this protects everyone, including those who do not mount a protective antibody response to influenza vaccination and those who are not vaccinated.

DOES ANNUAL VACCINATION BLUNT ITS EFFECTIVENESS?

Some studies from the 1970s and 1980s raised concern over a possible negative effect of annual influenza vaccination on vaccine effectiveness. The “antigenic distance hypothesis” holds that vaccine effectiveness is influenced by antigenic similarity between the previous season’s vaccine serotypes and the epidemic serotypes, as well as the antigenic similarity between the serotypes of the current and previous seasons.

A meta-analysis of studies from 2010 through 2015 showed significant inconsistencies in repeat vaccination effects within and between seasons and serotypes. It also showed that vaccine effectiveness may be influenced by more than 1 previous season, particularly for influenza A(H3N2), in which repeated vaccination can blunt the hemagglutinin antibody response.³⁰

A study from Japan showed that people who needed medical attention for influenza in the previous season were at lower risk of a similar event in the current season.³¹ Prior-season influenza vaccination reduced current-season vaccine effectiveness only in those who did not have medically attended influenza in the prior season. This suggests that infection is more immunogenic than vaccination, but only against the serotype causing the infection and not the other serotypes included in the vaccine.

An Australian study showed that annual influenza vaccination did not decrease vaccine effectiveness against influenza-associated hospitalization. Rather, effectiveness increased by about 15% in those vaccinated in both current and previous seasons compared with those vaccinated in either season alone.³²

European investigators showed that repeated seasonal influenza vaccination in the elderly prevented the need for hospitalization due to influenza A(H3N2) and B, but not A(H1N1)pdm09.³³

High-dose vaccine for older adults

The high-dose influenza vaccine has been licensed since 2009 for use in the United States for people ages 65 and older.

Recent studies confirmed that high-dose vaccine is more effective than standard-dose vaccine in veterans³⁴ and US Medicare beneficiaries.³⁵

The high-dose vaccine is rapidly becoming the primary vaccine given to people ages 65 and older in retail pharmacies, where vaccination begins earlier in the season than in providers’ offices.³⁶ Some studies have shown that the standard-dose vaccine wanes in effectiveness toward the end of the influenza season (particularly if the season is long) if it is given very early. It remains to be seen whether the same applies to the high-dose influenza vaccine.

Some advocate twice-annual influenza vaccination, particularly for older adults living in tropical and subtropical areas, where influenza seasons are more prolonged. However, a recently published study observed reductions in influenza-specific hemagglutination inhibition and cell-mediated immunity after twice-annual vaccination.³⁷

Vaccination is beneficial during pregnancy

Many studies have shown the value of influenza vaccination during pregnancy for both mothers and their infants.

One recently published study showed that 18% of infants who developed influenza required hospitalization.³⁸ In that study, prenatal and postpartum maternal influenza vaccination decreased the odds of influenza in infants by 61% and 53%, respectively.

Another study showed that vaccine effectiveness did not vary by gestational age at vaccination.³⁹

Some studies have shown that influenza virus infection can increase susceptibility to certain bacterial infections. A post hoc analysis of an influenza vaccination study in pregnant women suggested that the vaccine was also associated with decreased rates of pertussis in these women.⁴⁰

Factors that make vaccination less effective

Several factors including age-related frailty and iatrogenic and disease-related immunosuppression can affect vaccine effectiveness.

Frailty. A recent study showed that vaccine effectiveness was 77.6% in nonfrail older adults but only 58.7% in frail older adults.⁴¹

Immunosuppression. Temporary discontinuation of methotrexate for 2 weeks after influenza vaccination in patients with rheumatoid arthritis improves vaccine immunogenicity without precipitating disease flare.⁴² Solid-organ and hematopoietic stem cell transplant recipients who received influenza vaccine were less likely to develop pneumonia and require intensive care unit admission.⁴³

The high-dose influenza vaccine is more immunogenic than the standard-dose vaccine in solid-organ transplant recipients.⁴⁴

Statins are widely prescribed and have recently been associated with reduced influenza vaccine effectiveness against medically attended acute respiratory illness, but their benefits in preventing cardiovascular events outweigh this risk.⁴⁵

FUTURE VACCINE CONSIDERATIONS

Moving away from eggs

During the annual egg-based production process, which takes several months, the influenza vaccine acquires antigenic changes that allow replication in eggs, particularly in the hemagglutinin protein, which mediates receptor binding. This process of egg adaptation may cause antigenic changes that decrease vaccine effectiveness against circulating viruses.

The cell-based baculovirus influenza vaccine grown in dog kidney cells has higher antigenic content and is not subject to the limitations of egg-based vaccine, although it still requires annual updates. A recombinant influenza vaccine reduces the probability of influenza-like illness by 30% compared with the egg-based influenza vaccine, but also still requires annual updates.⁴⁶ The market share of these non-egg-based vaccines is small, and thus their effectiveness has yet to be demonstrated.

The US Department of Defense administered the cell-based influenza vaccine to about one-third of Armed Forces personnel, their families, and retirees in the 2017–2018 influenza seasons, and data on its effectiveness are expected in the near future.⁴⁷

A universal vaccine would be ideal

The quest continues for a universal influenza vaccine, one that remains protective for several years and does not require annual updates.⁴⁸ Such a vaccine would protect against seasonal epidemic influenza drift variants and pandemic strains. More people could likely be persuaded to be vaccinated once rather than every year.

David Schumick, medical illustrator; adapted from reference 49; illustration of the influenza hemagglutinin created in part with data from reference 50.

An ideal universal vaccine would be suitable for all age groups, at least 75% effective against symptomatic influenza virus infection, protective against all influenza A viruses (influenza A, not B, causes pandemics and seasonal epidemics), and durable through multiple influenza seasons.⁵¹

Research and production of such a vaccine are expected to require funding of about $1 billion over the next 5 years.

Boosting effectiveness

Estimates of influenza vaccine effectiveness range from 40% to 60% in years when the vaccine viruses closely match the circulating viruses, and variably lower when they do not match. The efficacy of most other vaccines given to prevent other infections is much higher.

New technologies to improve influenza vaccine effectiveness are needed, particularly for influenza A(H3N2) viruses, which are rapidly evolving and are highly susceptible to egg-adaptive mutations in the manufacturing process.

In one study, a nanoparticle vaccine formulated with a saponin-based adjuvant induced hemagglutination inhibition responses that were even greater than those induced by the high-dose vaccine.⁵²

Immunoglobulin A (IgA) may be a more effective vaccine target than traditional influenza vaccines that target IgG, since different parts of IgA may engage the influenza virus simultaneously.⁵³

Vaccines can be developed more quickly than in the past. The timeline from viral sequencing to human studies with deoxyribonucleic acid plasmid vaccines decreased from 20 months in 2003 for the severe acquired respiratory syndrome coronavirus to 11 months in 2006 for influenza A/Indonesia/2006 (H5), to 4 months in 2009 for influenza A/California/2009 (H1), to 3.5 months in 2016 for Zika virus.⁵⁴ This is because it is possible today to sequence a virus and insert the genetic material into a vaccine platform without ever having to grow the virus.

Numerous studies have found anti-influenza medications to be effective. Nevertheless, in an analysis of the 2011–2016 influenza seasons, only 15% of high-risk patients were prescribed anti-influenza medications within 2 days of symptom onset, including 37% in those with laboratory-confirmed influenza.⁵⁵ Fever was associated with an increased rate of antiviral treatment, but 25% of high-risk outpatients were afebrile. Empiric treatment of 4 high-risk outpatients with acute respiratory illness was needed to treat 1 patient with influenza.⁵⁵

Treatment with a neuraminidase inhibitor within 2 days of illness has recently been shown to improve survival and shorten duration of viral shedding in patients with avian influenza A(H7N9) infection.⁵⁶ Antiviral treatment within 2 days of illness is associated with improved outcomes in transplant recipients⁵⁷ and with a lower risk of otitis media in children.⁵⁸

Appropriate anti-influenza treatment is as important as avoiding unnecessary antibiotics. Regrettably, as many as one-third of patients with laboratory-confirmed influenza are prescribed antibiotics.⁵⁹

The US Food and Drug Administration warns against fraudulent unapproved over-the-counter influenza products.⁶⁰

Baloxavir marboxil

Baloxavir marboxil is a new anti-influenza medication approved in Japan in February 2018 and anticipated to be available in the United States sometime in 2019.

This prodrug is hydrolyzed in vivo to the active metabolite, which selectively inhibits cap-dependent endonuclease enzyme, a key enzyme in initiation of messenger ribonucleic acid synthesis required for influenza viral replication.⁶¹

In a double-blind phase 3 trial, the median time to alleviation of influenza symptoms is 26.5 hours shorter with baloxavir marboxil than with placebo. One tablet was as effective as 5 days of the neuraminidase inhibitor oseltamivir and was associated with greater reduction in viral load 1 day after initiation, and similar side effects.⁶² Of concern is the emergence of nucleic acid substitutions conferring resistance to baloxavir; this occurred in 2.2% and 9.7% of baloxavir recipients in the phase 2 and 3 trials, respectively.

CLOSING THE GAPS

Several gaps in the management of influenza persist since the 1918 pandemic.¹ These include gaps in epidemiology, prevention, diagnosis, treatment, and prognosis.

• Global networks wider than current ones are needed to address this global disease and to prioritize coordination efforts.
• Establishing and strengthening clinical capacity is needed in limited resource settings. New technologies are needed to expedite vaccine development and to achieve progress toward a universal vaccine.
• Current diagnostic tests do not distinguish between seasonal and novel influenza A viruses of zoonotic origin, which are expected to cause the next pandemic.
• Current antivirals have been shown to shorten duration of illness in outpatients with uncomplicated influenza, but the benefit in hospitalized patients has been less well established.
• In 2007, resistance of seasonal influenza A(H1N1) to oseltamivir became widespread. In 2009, pandemic influenza A(H1N1), which is highly susceptible to oseltamivir, replaced the seasonal virus and remains the predominantly circulating A(H1N1) strain.
• A small-molecule fragment, N-cyclohexyaltaurine, binds to the conserved hemagglutinin receptor-binding site in a manner that mimics the binding mode of the natural receptor sialic acid. This can serve as a template to guide the development of novel broad-spectrum small-molecule anti-influenza drugs.⁶³
• Biomarkers that can accurately predict development of severe disease in patients with influenza are needed.

A plethora of studies are under way in the field of perioperative medicine. As a result, evidence-based care of surgical patients is evolving at an exponential rate.

We performed a literature search and, using consensus, identified recent articles we believe will have a great impact on perioperative cardiovascular medicine. These articles report studies that were presented at national meetings in 2018, including the Perioperative Medicine Summit, Society of General Internal Medicine, and Society of Hospital Medicine. These articles are grouped under 5 questions that will help guide clinical practice in perioperative cardiovascular medicine.

SHOULD ASPIRIN BE CONTINUED PERIOPERATIVELY IN PATIENTS WITH A CORONARY STENT?

The Perioperative Ischemic Evaluation 2 (POISE-2) trial¹ found that giving aspirin before surgery and throughout the early postoperative period had no significant effect on the rate of a composite of death or nonfatal myocardial infarction; moreover, aspirin increased the risk of major bleeding. However, many experts felt uncomfortable stopping aspirin preoperatively in patients taking it for secondary prophylaxis, particularly patients with a coronary stent.

[Graham MM, Sessler DI, Parlow JL, et al. Aspirin in patients with previous percutaneous coronary intervention undergoing noncardiac surgery. Ann Intern Med 2018; 168(4):237–244.]

This post hoc subgroup analysis² of POISE-2 evaluated the benefit and harm of perioperative aspirin in patients who had previously undergone percutaneous coronary intervention, more than 90% of whom had received a stent. Patients were age 45 or older with atherosclerotic heart disease or risk factors for it who had previously undergone percutaneous coronary intervention and were now undergoing noncardiac surgery.

Patients who had received a bare-metal stent within the previous 6 weeks or a drug-eluting stent within 12 months before surgery were excluded because guidelines at that time said to continue dual antiplatelet therapy for that long. Recommendations have since changed; the optimal duration for dual antiplatelet therapy with drug-eluting stents is now 6 months. Second-generation drug-eluting stents pose a lower risk of stent thrombosis and require a shorter duration of dual antiplatelet therapy than first-generation drug-eluting stents. Approximately 25% of the percutaneous coronary intervention subgroup had a drug-eluting stent, but the authors did not specify the type of drug-eluting stent.

The post hoc analysis² included a subgroup of 234 of 4,998 patients receiving aspirin and 236 of 5,012 patients receiving placebo initiated within 4 hours before surgery and continued postoperatively. The primary outcome measured was the rate of death or nonfatal myocardial infarction within 30 days after surgery, and bleeding was a secondary outcome.

Findings. Although the overall POISE-2 study found no benefit from aspirin, in the subgroup who had previously undergone percutaneous coronary intervention, aspirin significantly reduced the risk of the primary outcome, which occurred in 6% vs 11.5% of the patients:

• Absolute risk reduction 5.5% (95% confidence interval 0.4%–10.5%)
• Hazard ratio 0.50 (0.26–0.95).

The reduction was primarily due to fewer myocardial infarctions:

• Absolute risk reduction 5.9% (1.0%–10.8%)
• Hazard ratio 0.44 (0.22–0.87).

The type of stent had no effect on the primary outcome, although this subgroup analysis had limited power. In the nonpercutaneous coronary intervention subgroup, there was no significant difference in outcomes between the aspirin and placebo groups. This subgroup analysis was underpowered to evaluate the effect of aspirin on the composite of major and life-threatening bleeding in patients with prior percutaneous coronary intervention, which was reported as “uncertain” due to wide confidence intervals (absolute risk increase 1.3%, 95% confidence interval –2.6% to 5.2%), but the increased risk of major or life-threatening bleeding with aspirin demonstrated in the overall POISE-2 study population likely applies:

• Absolute risk increase 0.8% (0.1%–1.6%)
• Hazard ratio 1.22 (1.01–1.48).

Limitations. This was a nonspecified subgroup analysis that was underpowered and had a relatively small sample size with few events.

Conclusion. In the absence of a very high bleeding risk, continuing aspirin perioperatively in patients with prior percutaneous coronary intervention undergoing noncardiac surgery is more likely to result in benefit than harm. This finding is in agreement with current recommendations from the American College Cardiology and American Heart Association (class I; level of evidence C).³

Despite advances in anesthesia and surgical techniques, about 1% of patients over age 45 die within 30 days of noncardiac surgery.⁴ Studies have demonstrated a high mortality rate in patients who experience myocardial injury after noncardiac surgery (MINS), defined as elevations of troponin T with or without ischemic symptoms or electrocardiographic changes.⁵ Most of these studies used earlier, “non-high-sensitivity” troponin T assays. Fifth-generation, highly sensitive troponin T assays are now available that can detect troponin T at lower concentrations, but their utility in predicting postoperative outcomes remains uncertain. Two recent studies provide further insight into these issues.

[Writing Committee for the VISION Study Investigators, Devereaux PJ, Biccard BM, Sigamani A, et al. Association of postoperative high-sensitivity troponin levels with myocardial injury and 30-day mortality among patients undergoing noncardiac surgery. JAMA 2017; 317(16):1642–1651.]

The Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study⁵ was an international, prospective cohort study that initially evaluated the association between MINS and the 30-day mortality rate using a non-high-sensitivity troponin T assay (Roche fourth-generation Elecsys TnT assay) in patients age 45 or older undergoing noncardiac surgery and requiring hospital admission for at least 1 night. After the first 15,000 patients, the study switched to the Roche fifth-generation assay, with measurements at 6 to 12 hours after surgery and on postoperative days 1, 2, and 3.

A 2017 analysis by Devereaux et al⁶ included only these later-enrolled patients and correlated their high-sensitivity troponin T levels with 30-day mortality rates. Patients with a level 14 ng/L or higher, the upper limit of normal in this study, were also assessed for ischemic symptoms and electrocardiographic changes. Although not required by the study, more than 7,800 patients had their troponin T levels measured before surgery, and the absolute change was also analyzed for an association with the 30-day mortality rate.

Findings. Of the 21,842 patients, about two-thirds underwent some form of major surgery; some of them had more than 1 type. A total of 1.2% of the patients died within 30 days of surgery.

Based on their analysis, the authors proposed that MINS be defined as:

• A postoperative troponin T level of 65 ng/L or higher, or
• A level in the range of 20 ng/L to less than 65 ng/L with an absolute increase from the preoperative level at least 5 ng/L, not attributable to a nonischemic cause.

Seventeen percent of the study patients met these criteria, and of these, 21.7% met the universal definition of myocardial infarction, although only 6.9% had symptoms of it.

Limitations. Only 40.4% of the patients had a preoperative high-sensitivity troponin T measurement for comparison, and in 13.8% of patients who had an elevated perioperative measurement, their preoperative value was the same or higher than their postoperative one. Thus, the incidence of MINS may have been overestimated if patients were otherwise not known to have troponin T elevations before surgery.

[Puelacher C, Lurati Buse G, Seeberger D, et al. Perioperative myocardial injury after noncardiac surgery: incidence, mortality, and characterization. Circulation 2018; 137(12):1221–1232.]

Puelacher et al⁷ investigated the prevalence of MINS in 2,018 patients at increased cardiovascular risk (age ≥ 65, or age ≥ 45 with a history of coronary artery disease, peripheral vascular disease, or stroke) who underwent major noncardiac surgery (planned overnight stay ≥ 24 hours) at a university hospital in Switzerland. Patients had their troponin T measured with a high-sensitivity assay within 30 days before surgery and on postoperative days 1 and 2.

Instead of MINS, the investigators used the term “perioperative myocardial injury” (PMI), defined as an absolute increase in troponin T of at least 14 ng/L from before surgery to the peak postoperative reading. Similar to MINS, PMI did not require ischemic features, but in this study, noncardiac triggers (sepsis, stroke, or pulmonary embolus) were not excluded.

Findings. PMI occurred in 16% of surgeries, and of the patients with PMI, 6% had typical chest pain and 18% had any ischemic symptoms. Unlike in the POISE-2 study discussed above, PMI triggered an automatic referral to a cardiologist.

The unadjusted 30-day mortality rate was 8.9% among patients with PMI and 1.5% in those without. Multivariable logistic regression analysis showed an adjusted hazard ratio for 30-day mortality of 2.7 (95% CI 1.5–4.8) for those with PMI vs without, and this difference persisted for at least 1 year.

In patients with PMI, the authors compared the 30-day mortality rate of those with no ischemic signs or symptoms (71% of the patients) with those who met the criteria for myocardial infarction and found no difference. Patients with PMI triggered by a noncardiac event had a worse prognosis than those with a presumed cardiac etiology.

Limitations. Despite the multivariate analysis that included adjustment for age, nonelective surgery, and Revised Cardiac Risk Index (RCRI), the increased risk associated with PMI could simply reflect higher risk at baseline. Although PMI resulted in automatic referral to a cardiologist, only 10% of patients eventually underwent coronary angiography; a similar percentage were discharged with additional medical therapy such as aspirin, a statin, or a beta-blocker. The effect of these interventions is not known.

Conclusions. MINS is common and has a strong association with mortality risk proportional to the degree of troponin T elevation using high-sensitivity assays, consistent with data from previous studies of earlier assays. Because the mechanism of MINS may differ from that of myocardial infarction, its prevention and treatment may differ, and it remains unclear how serial measurement in postoperative patients should change clinical practice.

The recently published Dabigatran in Patients With Myocardial Injury After Non-cardiac Surgery (MANAGE) trial⁸ suggests that dabigatran may reduce arterial and venous complications in patients with MINS, but the study had a number of limitations that may restrict the clinical applicability of this finding.

While awaiting further clinical outcomes data, pre- and postoperative troponin T measurement may be beneficial in higher-risk patients (such as those with cardiovascular disease or multiple RCRI risk factors) if the information will change perioperative management.

Intraoperative hypotension is associated with organ ischemia, which may cause postoperative myocardial infarction, myocardial injury, and acute kidney injury.⁹ Traditional anesthesia practice is to maintain intraoperative blood pressure within 20% of the preoperative baseline, based on the notion that hypertensive patients require higher perfusion pressures.

[Futier E, Lefrant J-Y, Guinot P-G, et al. Effect of individualized vs standard blood pressure management strategies on postoperative organ dysfunction among high-risk patients undergoing major surgery: a randomized clinical trial. JAMA 2017; 318(14):1346–1357.]

Futier et al¹⁰ sought to address uncertainty in intraoperative and immediate postoperative management of systolic blood pressure. In this multicenter, randomized, parallel-group trial, 298 patients at increased risk of postoperative renal complications were randomized to blood pressure management that was either “individualized” (within 10% of resting systolic pressure) or “standard” (≥ 80 mm Hg or ≥ 40% of resting systolic pressure) from induction to 4 hours postoperatively.

Blood pressure was monitored using radial arterial lines and maintained using a combination of intravenous fluids, norepinephrine (the first-line agent for the individualized group), and ephedrine (in the standard treatment group only). The primary outcome was a composite of systemic inflammatory response syndrome (SIRS) and organ dysfunction affecting at least 1 organ system (cardiovascular, respiratory, renal, hematologic, or neurologic).

Findings. Data on the primary outcome were available for 292 of 298 patients enrolled. The mean age was 70 years, 15% were women, and 82% had previously diagnosed hypertension. Despite the requirement for an elevated risk of acute kidney injury, only 13% of the patients had a baseline estimated glomerular filtration rate of less than 60 mL/min/1.73 m², and the median was 88 mL/min/1.73 m². Ninety-five percent of patients underwent abdominal surgery, and 50% of the surgeries were elective.

The mean systolic blood pressure was 123 mm Hg in the individualized treatment group compared with 116 mm Hg in the standard treatment group. Despite this small difference, 96% of individualized treatment patients received norepinephrine, compared with 26% in the standard treatment group.

The primary outcome of SIRS with organ dysfunction occurred in 38.1% of patients in the individualized treatment group and 51.7% of those in the standard treatment group. After adjusting for center, surgical urgency, surgical site, and acute kidney injury risk index, the relative risk of developing SIRS in those receiving individualized management was 0.73 (P = .02). Renal dysfunction (based on Acute Dialysis Quality Initiative criteria¹¹) occurred in 32.7% of individualized treatment patients and 49% of standardized treatment patients.  

Limitations of this study included differences in pharmacologic approach to maintain blood pressure in the 2 protocols (ephedrine and fluids vs norepinephrine) and a modest sample size.

Conclusions. Despite this, the difference in organ dysfunction was striking, with a number needed to treat of only 7 patients. This intervention extended 4 hours postoperatively, a time when many of these patients have left the postanesthesia care unit and have returned to hospitalist care on inpatient wards.

While optimal management of intraoperative and immediate postoperative blood pressure may not be settled, this study suggests that even mild relative hypotension may justify immediate action. Further studies may be useful to delineate high- and low-risk populations, the timing of greatest risk, and indications for intraarterial blood pressure monitoring.

[Salmasi V, Maheswari K, Yang D, et al. Relationship between intraoperative hypotension, defined by either reduction from baseline or absolute thresholds, and acute kidney and myocardial injury after noncardiac surgery: a retrospective cohort analysis. Anesthesiology 2017; 126(1):47–65.]

This retrospective cohort study¹² assessed the association between myocardial or kidney injury and absolute or relative thresholds of intraoperative mean arterial pressure. It included 57,315 adults who underwent inpatient noncardiac surgery, had a preoperative and at least 1 postoperative serum creatinine measurement within 7 days, and had blood pressure recorded in preoperative appointments within 6 months. Patients with chronic kidney disease (glomerular filtration rate < 60 mL/min/1.73 m²) and those on dialysis were excluded. The outcomes were MINS⁵ and acute kidney injury as defined by the Acute Kidney Injury Network.⁹

Findings. A mean arterial pressure below an absolute threshold of 65 mm Hg or a relative threshold of 20% lower than baseline value was associated with myocardial and kidney injury. At each threshold, prolonged periods of hypotension were associated with progressively increased risk.

An important conclusion of the study was that relative thresholds of mean arterial pressure were not any more predictive than absolute thresholds. Absolute thresholds are easier to use intraoperatively, especially when baseline values are not available. The authors did not find a clinically significant interaction between baseline blood pressure and the association of hypotension and myocardial and kidney injury.

Limitations included use of cardiac enzymes postoperatively to define MINS. Since these were not routinely collected, clinically silent myocardial injury may have been missed. Baseline blood pressure may have important implications in other forms of organ injury (ie, cerebral ischemia) that were not studied.

Summary. The lowest absolute mean arterial pressure is as predictive of postoperative myocardial and kidney injury as the relative pressure reduction, at least in patients with normal renal function. Limiting exposure to intraoperative hypotension is important. Baseline blood pressure values may have limited utility for intraoperative management.

In combination, these studies confirm that intraoperative hypotension is a predictor of postoperative organ dysfunction, but the definition and management remain unclear. While aggressive intraoperative management is likely beneficial, how to manage the anti­hypertensive therapy the patient has been taking as an outpatient when he or she comes into the hospital for surgery remains uncertain.

Perioperative stroke is an uncommon, severe complication of noncardiac surgery. The pathophysiology has been better defined in cardiac than in noncardiac surgeries. In nonsurgical patients, patent foramen ovale (PFO) is associated with stroke, even in patients considered to be at low risk.¹³ Perioperative patients have additional risk for venous thromboembolism and may have periprocedural antithrombotic medications altered, increasing their risk of paradoxical embolism through the PFO.

[Ng PY, Ng AK, Subramaniam B, et al. Association of preoperatively diagnosed patent foramen ovale with perioperative ischemic stroke. JAMA 2018; 319(5):452–462.]

This retrospective cohort study of noncardiac surgery patients at 3 hospitals¹⁴ sought to determine the association of preoperatively diagnosed PFO with the risk of perioperative ischemic stroke identified by International Classification of Diseases diagnoses.

Of 150,198 patients, 1.0% had a preoperative diagnosis of PFO, and at baseline, those with PFO had significantly more comorbidities than those without PFO. Stroke occurred in 3.2% of patients with PFO vs 0.5% of those without. Patients known to have a PFO were much more likely to have cardiovascular and thromboembolic risk factors for stroke. In the adjusted analysis, the absolute risk difference between groups was 0.4% (95% CI 0.2–0.6%), with an estimated perioperative stroke risk of 5.9 per 1,000 in patients with known patent foramen ovale and 2.2 per 1,000 in those without. A diagnosis of PFO was also associated with increased risk of large-vessel-territory stroke and more severe neurologic deficit.

Further attempts to adjust for baseline risk factors and other potential bias, including a propensity score-matched cohort analysis and an analysis limited to patients who had echocardiography performed in the same healthcare system, still showed a higher risk of perioperative stroke among patients with preoperatively detected patent foramen ovale.

Limitations. The study was retrospective and observational, used administrative data, and had a low rate of PFO diagnosis (1%), compared with about 25% in population-based studies.¹⁵ Indications for preoperative echocardiography are unknown. In addition, the study specifically examined preoperatively diagnosed PFO, rather than including those diagnosed in the postoperative period.

Discussion. How does this study affect clinical practice? The absolute stroke risk was increased by 0.4% in patients with PFO compared with those without. Although this is a relatively small increase, millions of patients undergo noncardiac surgery annually. The risks of therapeutic anticoagulation or PFO closure are likely too high in this context; however, clinicians may approach the perioperative management of antiplatelet agents and venous thromboembolism prophylaxis in patients with known PFO with additional caution.

HOW DOES TIMING OF EMERGENCY SURGERY AFTER PRIOR STROKE AFFECT OUTCOMES?

A history of stroke or transient ischemic attack is a known risk factor for perioperative vascular complications. A recent large cohort study demonstrated that a history of stroke within 9 months of elective surgery was associated with increased adverse outcomes.¹⁶ Little is known, however, of the perioperative risk in patients with a history of stroke who undergo emergency surgery.

[Christiansen MN, Andersson C, Gislason GH, et al. Risks of cardiovascular adverse events and death in patients with previous stroke undergoing emergency noncardiac, nonintracranial surgery: the importance of operative timing. Anesthesiology 2017; 127(1):9–19.]

In this study,¹⁷ all emergency noncardiac and nonintracranial surgeries from 2005 to 2011 were analyzed using multiple national patient registries in Denmark according to time elapsed between previous stroke and surgery. Primary outcomes were 30-day all-cause mortality and 30-day major adverse cardiac events (MACE), defined as nonfatal ischemic stroke, nonfatal myocardial infarction, and cardiovascular death. Statistical analysis to assess the risk of adverse outcomes included logistic regression models, spline analyses, and propensity-score matching.

Findings. The authors identified 146,694 emergency surgeries, with 7,861 patients (5.4%) having had a previous stroke (transient ischemic attacks and hemorrhagic strokes were not included). Rates of postoperative stroke were as follows:

• 9.9% in patents with a history of ischemic stroke within 3 months of surgery
• 2.8% in patients with a history of stroke 3 to 9 months before surgery
• 0.3% in patients with no previous stroke.

The risk plateaued when the time between stroke and surgery exceeded 4 to 5 months.¹⁵

Interestingly, in patients who underwent emergency surgery within 14 days of stroke, the risk of MACE was significantly lower immediately after surgery (1–3 days after stroke) compared with surgery that took place 4 to 14 days after stroke. The authors hypothesized that because cerebral autoregulation does not become compromised until approximately 5 days after a stroke, the risk was lower 1 to 3 days after surgery and increased thereafter.

Limitations of this study included the possibility of residual confounding, given its retrospective design using administrative data, not accounting for preoperative antithrombotic and anticoagulation therapy, and lack of information regarding the etiology of recurrent stroke (eg, thromboembolic, atherothrombotic, hypoperfusion).

Conclusions. Although it would be impractical to postpone emergency surgery in a patient who recently had a stroke, this study shows that the incidence rates of postoperative recurrent stroke and MACE are high. Therefore, it is important that the patient and perioperative team be aware of the risk. Further research is needed to confirm these estimates of postoperative adverse events in more diverse patient populations.

A plethora of studies are under way in the field of perioperative medicine. As a result, evidence-based care of surgical patients is evolving at an exponential rate.

We performed a literature search and, using consensus, identified recent articles we believe will have a great impact on perioperative cardiovascular medicine. These articles report studies that were presented at national meetings in 2018, including the Perioperative Medicine Summit, Society of General Internal Medicine, and Society of Hospital Medicine. These articles are grouped under 5 questions that will help guide clinical practice in perioperative cardiovascular medicine.

SHOULD ASPIRIN BE CONTINUED PERIOPERATIVELY IN PATIENTS WITH A CORONARY STENT?

The Perioperative Ischemic Evaluation 2 (POISE-2) trial¹ found that giving aspirin before surgery and throughout the early postoperative period had no significant effect on the rate of a composite of death or nonfatal myocardial infarction; moreover, aspirin increased the risk of major bleeding. However, many experts felt uncomfortable stopping aspirin preoperatively in patients taking it for secondary prophylaxis, particularly patients with a coronary stent.

[Graham MM, Sessler DI, Parlow JL, et al. Aspirin in patients with previous percutaneous coronary intervention undergoing noncardiac surgery. Ann Intern Med 2018; 168(4):237–244.]

This post hoc subgroup analysis² of POISE-2 evaluated the benefit and harm of perioperative aspirin in patients who had previously undergone percutaneous coronary intervention, more than 90% of whom had received a stent. Patients were age 45 or older with atherosclerotic heart disease or risk factors for it who had previously undergone percutaneous coronary intervention and were now undergoing noncardiac surgery.

Patients who had received a bare-metal stent within the previous 6 weeks or a drug-eluting stent within 12 months before surgery were excluded because guidelines at that time said to continue dual antiplatelet therapy for that long. Recommendations have since changed; the optimal duration for dual antiplatelet therapy with drug-eluting stents is now 6 months. Second-generation drug-eluting stents pose a lower risk of stent thrombosis and require a shorter duration of dual antiplatelet therapy than first-generation drug-eluting stents. Approximately 25% of the percutaneous coronary intervention subgroup had a drug-eluting stent, but the authors did not specify the type of drug-eluting stent.

The post hoc analysis² included a subgroup of 234 of 4,998 patients receiving aspirin and 236 of 5,012 patients receiving placebo initiated within 4 hours before surgery and continued postoperatively. The primary outcome measured was the rate of death or nonfatal myocardial infarction within 30 days after surgery, and bleeding was a secondary outcome.

Findings. Although the overall POISE-2 study found no benefit from aspirin, in the subgroup who had previously undergone percutaneous coronary intervention, aspirin significantly reduced the risk of the primary outcome, which occurred in 6% vs 11.5% of the patients:

• Absolute risk reduction 5.5% (95% confidence interval 0.4%–10.5%)
• Hazard ratio 0.50 (0.26–0.95).

The reduction was primarily due to fewer myocardial infarctions:

• Absolute risk reduction 5.9% (1.0%–10.8%)
• Hazard ratio 0.44 (0.22–0.87).

The type of stent had no effect on the primary outcome, although this subgroup analysis had limited power. In the nonpercutaneous coronary intervention subgroup, there was no significant difference in outcomes between the aspirin and placebo groups. This subgroup analysis was underpowered to evaluate the effect of aspirin on the composite of major and life-threatening bleeding in patients with prior percutaneous coronary intervention, which was reported as “uncertain” due to wide confidence intervals (absolute risk increase 1.3%, 95% confidence interval –2.6% to 5.2%), but the increased risk of major or life-threatening bleeding with aspirin demonstrated in the overall POISE-2 study population likely applies:

• Absolute risk increase 0.8% (0.1%–1.6%)
• Hazard ratio 1.22 (1.01–1.48).

Limitations. This was a nonspecified subgroup analysis that was underpowered and had a relatively small sample size with few events.

Conclusion. In the absence of a very high bleeding risk, continuing aspirin perioperatively in patients with prior percutaneous coronary intervention undergoing noncardiac surgery is more likely to result in benefit than harm. This finding is in agreement with current recommendations from the American College Cardiology and American Heart Association (class I; level of evidence C).³

Despite advances in anesthesia and surgical techniques, about 1% of patients over age 45 die within 30 days of noncardiac surgery.⁴ Studies have demonstrated a high mortality rate in patients who experience myocardial injury after noncardiac surgery (MINS), defined as elevations of troponin T with or without ischemic symptoms or electrocardiographic changes.⁵ Most of these studies used earlier, “non-high-sensitivity” troponin T assays. Fifth-generation, highly sensitive troponin T assays are now available that can detect troponin T at lower concentrations, but their utility in predicting postoperative outcomes remains uncertain. Two recent studies provide further insight into these issues.

[Writing Committee for the VISION Study Investigators, Devereaux PJ, Biccard BM, Sigamani A, et al. Association of postoperative high-sensitivity troponin levels with myocardial injury and 30-day mortality among patients undergoing noncardiac surgery. JAMA 2017; 317(16):1642–1651.]

The Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study⁵ was an international, prospective cohort study that initially evaluated the association between MINS and the 30-day mortality rate using a non-high-sensitivity troponin T assay (Roche fourth-generation Elecsys TnT assay) in patients age 45 or older undergoing noncardiac surgery and requiring hospital admission for at least 1 night. After the first 15,000 patients, the study switched to the Roche fifth-generation assay, with measurements at 6 to 12 hours after surgery and on postoperative days 1, 2, and 3.

A 2017 analysis by Devereaux et al⁶ included only these later-enrolled patients and correlated their high-sensitivity troponin T levels with 30-day mortality rates. Patients with a level 14 ng/L or higher, the upper limit of normal in this study, were also assessed for ischemic symptoms and electrocardiographic changes. Although not required by the study, more than 7,800 patients had their troponin T levels measured before surgery, and the absolute change was also analyzed for an association with the 30-day mortality rate.

Findings. Of the 21,842 patients, about two-thirds underwent some form of major surgery; some of them had more than 1 type. A total of 1.2% of the patients died within 30 days of surgery.

Based on their analysis, the authors proposed that MINS be defined as:

• A postoperative troponin T level of 65 ng/L or higher, or
• A level in the range of 20 ng/L to less than 65 ng/L with an absolute increase from the preoperative level at least 5 ng/L, not attributable to a nonischemic cause.

Seventeen percent of the study patients met these criteria, and of these, 21.7% met the universal definition of myocardial infarction, although only 6.9% had symptoms of it.

Limitations. Only 40.4% of the patients had a preoperative high-sensitivity troponin T measurement for comparison, and in 13.8% of patients who had an elevated perioperative measurement, their preoperative value was the same or higher than their postoperative one. Thus, the incidence of MINS may have been overestimated if patients were otherwise not known to have troponin T elevations before surgery.

[Puelacher C, Lurati Buse G, Seeberger D, et al. Perioperative myocardial injury after noncardiac surgery: incidence, mortality, and characterization. Circulation 2018; 137(12):1221–1232.]

Puelacher et al⁷ investigated the prevalence of MINS in 2,018 patients at increased cardiovascular risk (age ≥ 65, or age ≥ 45 with a history of coronary artery disease, peripheral vascular disease, or stroke) who underwent major noncardiac surgery (planned overnight stay ≥ 24 hours) at a university hospital in Switzerland. Patients had their troponin T measured with a high-sensitivity assay within 30 days before surgery and on postoperative days 1 and 2.

Instead of MINS, the investigators used the term “perioperative myocardial injury” (PMI), defined as an absolute increase in troponin T of at least 14 ng/L from before surgery to the peak postoperative reading. Similar to MINS, PMI did not require ischemic features, but in this study, noncardiac triggers (sepsis, stroke, or pulmonary embolus) were not excluded.

Findings. PMI occurred in 16% of surgeries, and of the patients with PMI, 6% had typical chest pain and 18% had any ischemic symptoms. Unlike in the POISE-2 study discussed above, PMI triggered an automatic referral to a cardiologist.

The unadjusted 30-day mortality rate was 8.9% among patients with PMI and 1.5% in those without. Multivariable logistic regression analysis showed an adjusted hazard ratio for 30-day mortality of 2.7 (95% CI 1.5–4.8) for those with PMI vs without, and this difference persisted for at least 1 year.

In patients with PMI, the authors compared the 30-day mortality rate of those with no ischemic signs or symptoms (71% of the patients) with those who met the criteria for myocardial infarction and found no difference. Patients with PMI triggered by a noncardiac event had a worse prognosis than those with a presumed cardiac etiology.

Limitations. Despite the multivariate analysis that included adjustment for age, nonelective surgery, and Revised Cardiac Risk Index (RCRI), the increased risk associated with PMI could simply reflect higher risk at baseline. Although PMI resulted in automatic referral to a cardiologist, only 10% of patients eventually underwent coronary angiography; a similar percentage were discharged with additional medical therapy such as aspirin, a statin, or a beta-blocker. The effect of these interventions is not known.

Conclusions. MINS is common and has a strong association with mortality risk proportional to the degree of troponin T elevation using high-sensitivity assays, consistent with data from previous studies of earlier assays. Because the mechanism of MINS may differ from that of myocardial infarction, its prevention and treatment may differ, and it remains unclear how serial measurement in postoperative patients should change clinical practice.

The recently published Dabigatran in Patients With Myocardial Injury After Non-cardiac Surgery (MANAGE) trial⁸ suggests that dabigatran may reduce arterial and venous complications in patients with MINS, but the study had a number of limitations that may restrict the clinical applicability of this finding.

While awaiting further clinical outcomes data, pre- and postoperative troponin T measurement may be beneficial in higher-risk patients (such as those with cardiovascular disease or multiple RCRI risk factors) if the information will change perioperative management.

Intraoperative hypotension is associated with organ ischemia, which may cause postoperative myocardial infarction, myocardial injury, and acute kidney injury.⁹ Traditional anesthesia practice is to maintain intraoperative blood pressure within 20% of the preoperative baseline, based on the notion that hypertensive patients require higher perfusion pressures.

[Futier E, Lefrant J-Y, Guinot P-G, et al. Effect of individualized vs standard blood pressure management strategies on postoperative organ dysfunction among high-risk patients undergoing major surgery: a randomized clinical trial. JAMA 2017; 318(14):1346–1357.]

Futier et al¹⁰ sought to address uncertainty in intraoperative and immediate postoperative management of systolic blood pressure. In this multicenter, randomized, parallel-group trial, 298 patients at increased risk of postoperative renal complications were randomized to blood pressure management that was either “individualized” (within 10% of resting systolic pressure) or “standard” (≥ 80 mm Hg or ≥ 40% of resting systolic pressure) from induction to 4 hours postoperatively.

Blood pressure was monitored using radial arterial lines and maintained using a combination of intravenous fluids, norepinephrine (the first-line agent for the individualized group), and ephedrine (in the standard treatment group only). The primary outcome was a composite of systemic inflammatory response syndrome (SIRS) and organ dysfunction affecting at least 1 organ system (cardiovascular, respiratory, renal, hematologic, or neurologic).

Findings. Data on the primary outcome were available for 292 of 298 patients enrolled. The mean age was 70 years, 15% were women, and 82% had previously diagnosed hypertension. Despite the requirement for an elevated risk of acute kidney injury, only 13% of the patients had a baseline estimated glomerular filtration rate of less than 60 mL/min/1.73 m², and the median was 88 mL/min/1.73 m². Ninety-five percent of patients underwent abdominal surgery, and 50% of the surgeries were elective.

The mean systolic blood pressure was 123 mm Hg in the individualized treatment group compared with 116 mm Hg in the standard treatment group. Despite this small difference, 96% of individualized treatment patients received norepinephrine, compared with 26% in the standard treatment group.

The primary outcome of SIRS with organ dysfunction occurred in 38.1% of patients in the individualized treatment group and 51.7% of those in the standard treatment group. After adjusting for center, surgical urgency, surgical site, and acute kidney injury risk index, the relative risk of developing SIRS in those receiving individualized management was 0.73 (P = .02). Renal dysfunction (based on Acute Dialysis Quality Initiative criteria¹¹) occurred in 32.7% of individualized treatment patients and 49% of standardized treatment patients.  

Limitations of this study included differences in pharmacologic approach to maintain blood pressure in the 2 protocols (ephedrine and fluids vs norepinephrine) and a modest sample size.

Conclusions. Despite this, the difference in organ dysfunction was striking, with a number needed to treat of only 7 patients. This intervention extended 4 hours postoperatively, a time when many of these patients have left the postanesthesia care unit and have returned to hospitalist care on inpatient wards.

While optimal management of intraoperative and immediate postoperative blood pressure may not be settled, this study suggests that even mild relative hypotension may justify immediate action. Further studies may be useful to delineate high- and low-risk populations, the timing of greatest risk, and indications for intraarterial blood pressure monitoring.

[Salmasi V, Maheswari K, Yang D, et al. Relationship between intraoperative hypotension, defined by either reduction from baseline or absolute thresholds, and acute kidney and myocardial injury after noncardiac surgery: a retrospective cohort analysis. Anesthesiology 2017; 126(1):47–65.]

This retrospective cohort study¹² assessed the association between myocardial or kidney injury and absolute or relative thresholds of intraoperative mean arterial pressure. It included 57,315 adults who underwent inpatient noncardiac surgery, had a preoperative and at least 1 postoperative serum creatinine measurement within 7 days, and had blood pressure recorded in preoperative appointments within 6 months. Patients with chronic kidney disease (glomerular filtration rate < 60 mL/min/1.73 m²) and those on dialysis were excluded. The outcomes were MINS⁵ and acute kidney injury as defined by the Acute Kidney Injury Network.⁹

Findings. A mean arterial pressure below an absolute threshold of 65 mm Hg or a relative threshold of 20% lower than baseline value was associated with myocardial and kidney injury. At each threshold, prolonged periods of hypotension were associated with progressively increased risk.

An important conclusion of the study was that relative thresholds of mean arterial pressure were not any more predictive than absolute thresholds. Absolute thresholds are easier to use intraoperatively, especially when baseline values are not available. The authors did not find a clinically significant interaction between baseline blood pressure and the association of hypotension and myocardial and kidney injury.

Limitations included use of cardiac enzymes postoperatively to define MINS. Since these were not routinely collected, clinically silent myocardial injury may have been missed. Baseline blood pressure may have important implications in other forms of organ injury (ie, cerebral ischemia) that were not studied.

Summary. The lowest absolute mean arterial pressure is as predictive of postoperative myocardial and kidney injury as the relative pressure reduction, at least in patients with normal renal function. Limiting exposure to intraoperative hypotension is important. Baseline blood pressure values may have limited utility for intraoperative management.

In combination, these studies confirm that intraoperative hypotension is a predictor of postoperative organ dysfunction, but the definition and management remain unclear. While aggressive intraoperative management is likely beneficial, how to manage the anti­hypertensive therapy the patient has been taking as an outpatient when he or she comes into the hospital for surgery remains uncertain.

Perioperative stroke is an uncommon, severe complication of noncardiac surgery. The pathophysiology has been better defined in cardiac than in noncardiac surgeries. In nonsurgical patients, patent foramen ovale (PFO) is associated with stroke, even in patients considered to be at low risk.¹³ Perioperative patients have additional risk for venous thromboembolism and may have periprocedural antithrombotic medications altered, increasing their risk of paradoxical embolism through the PFO.

[Ng PY, Ng AK, Subramaniam B, et al. Association of preoperatively diagnosed patent foramen ovale with perioperative ischemic stroke. JAMA 2018; 319(5):452–462.]

This retrospective cohort study of noncardiac surgery patients at 3 hospitals¹⁴ sought to determine the association of preoperatively diagnosed PFO with the risk of perioperative ischemic stroke identified by International Classification of Diseases diagnoses.

Of 150,198 patients, 1.0% had a preoperative diagnosis of PFO, and at baseline, those with PFO had significantly more comorbidities than those without PFO. Stroke occurred in 3.2% of patients with PFO vs 0.5% of those without. Patients known to have a PFO were much more likely to have cardiovascular and thromboembolic risk factors for stroke. In the adjusted analysis, the absolute risk difference between groups was 0.4% (95% CI 0.2–0.6%), with an estimated perioperative stroke risk of 5.9 per 1,000 in patients with known patent foramen ovale and 2.2 per 1,000 in those without. A diagnosis of PFO was also associated with increased risk of large-vessel-territory stroke and more severe neurologic deficit.

Further attempts to adjust for baseline risk factors and other potential bias, including a propensity score-matched cohort analysis and an analysis limited to patients who had echocardiography performed in the same healthcare system, still showed a higher risk of perioperative stroke among patients with preoperatively detected patent foramen ovale.

Limitations. The study was retrospective and observational, used administrative data, and had a low rate of PFO diagnosis (1%), compared with about 25% in population-based studies.¹⁵ Indications for preoperative echocardiography are unknown. In addition, the study specifically examined preoperatively diagnosed PFO, rather than including those diagnosed in the postoperative period.

Discussion. How does this study affect clinical practice? The absolute stroke risk was increased by 0.4% in patients with PFO compared with those without. Although this is a relatively small increase, millions of patients undergo noncardiac surgery annually. The risks of therapeutic anticoagulation or PFO closure are likely too high in this context; however, clinicians may approach the perioperative management of antiplatelet agents and venous thromboembolism prophylaxis in patients with known PFO with additional caution.

HOW DOES TIMING OF EMERGENCY SURGERY AFTER PRIOR STROKE AFFECT OUTCOMES?

A history of stroke or transient ischemic attack is a known risk factor for perioperative vascular complications. A recent large cohort study demonstrated that a history of stroke within 9 months of elective surgery was associated with increased adverse outcomes.¹⁶ Little is known, however, of the perioperative risk in patients with a history of stroke who undergo emergency surgery.

[Christiansen MN, Andersson C, Gislason GH, et al. Risks of cardiovascular adverse events and death in patients with previous stroke undergoing emergency noncardiac, nonintracranial surgery: the importance of operative timing. Anesthesiology 2017; 127(1):9–19.]

In this study,¹⁷ all emergency noncardiac and nonintracranial surgeries from 2005 to 2011 were analyzed using multiple national patient registries in Denmark according to time elapsed between previous stroke and surgery. Primary outcomes were 30-day all-cause mortality and 30-day major adverse cardiac events (MACE), defined as nonfatal ischemic stroke, nonfatal myocardial infarction, and cardiovascular death. Statistical analysis to assess the risk of adverse outcomes included logistic regression models, spline analyses, and propensity-score matching.

Findings. The authors identified 146,694 emergency surgeries, with 7,861 patients (5.4%) having had a previous stroke (transient ischemic attacks and hemorrhagic strokes were not included). Rates of postoperative stroke were as follows:

• 9.9% in patents with a history of ischemic stroke within 3 months of surgery
• 2.8% in patients with a history of stroke 3 to 9 months before surgery
• 0.3% in patients with no previous stroke.

The risk plateaued when the time between stroke and surgery exceeded 4 to 5 months.¹⁵

Interestingly, in patients who underwent emergency surgery within 14 days of stroke, the risk of MACE was significantly lower immediately after surgery (1–3 days after stroke) compared with surgery that took place 4 to 14 days after stroke. The authors hypothesized that because cerebral autoregulation does not become compromised until approximately 5 days after a stroke, the risk was lower 1 to 3 days after surgery and increased thereafter.

Limitations of this study included the possibility of residual confounding, given its retrospective design using administrative data, not accounting for preoperative antithrombotic and anticoagulation therapy, and lack of information regarding the etiology of recurrent stroke (eg, thromboembolic, atherothrombotic, hypoperfusion).

Conclusions. Although it would be impractical to postpone emergency surgery in a patient who recently had a stroke, this study shows that the incidence rates of postoperative recurrent stroke and MACE are high. Therefore, it is important that the patient and perioperative team be aware of the risk. Further research is needed to confirm these estimates of postoperative adverse events in more diverse patient populations.

A plethora of studies are under way in the field of perioperative medicine. As a result, evidence-based care of surgical patients is evolving at an exponential rate.

We performed a literature search and, using consensus, identified recent articles we believe will have a great impact on perioperative cardiovascular medicine. These articles report studies that were presented at national meetings in 2018, including the Perioperative Medicine Summit, Society of General Internal Medicine, and Society of Hospital Medicine. These articles are grouped under 5 questions that will help guide clinical practice in perioperative cardiovascular medicine.

SHOULD ASPIRIN BE CONTINUED PERIOPERATIVELY IN PATIENTS WITH A CORONARY STENT?

The Perioperative Ischemic Evaluation 2 (POISE-2) trial¹ found that giving aspirin before surgery and throughout the early postoperative period had no significant effect on the rate of a composite of death or nonfatal myocardial infarction; moreover, aspirin increased the risk of major bleeding. However, many experts felt uncomfortable stopping aspirin preoperatively in patients taking it for secondary prophylaxis, particularly patients with a coronary stent.

[Graham MM, Sessler DI, Parlow JL, et al. Aspirin in patients with previous percutaneous coronary intervention undergoing noncardiac surgery. Ann Intern Med 2018; 168(4):237–244.]

This post hoc subgroup analysis² of POISE-2 evaluated the benefit and harm of perioperative aspirin in patients who had previously undergone percutaneous coronary intervention, more than 90% of whom had received a stent. Patients were age 45 or older with atherosclerotic heart disease or risk factors for it who had previously undergone percutaneous coronary intervention and were now undergoing noncardiac surgery.

Patients who had received a bare-metal stent within the previous 6 weeks or a drug-eluting stent within 12 months before surgery were excluded because guidelines at that time said to continue dual antiplatelet therapy for that long. Recommendations have since changed; the optimal duration for dual antiplatelet therapy with drug-eluting stents is now 6 months. Second-generation drug-eluting stents pose a lower risk of stent thrombosis and require a shorter duration of dual antiplatelet therapy than first-generation drug-eluting stents. Approximately 25% of the percutaneous coronary intervention subgroup had a drug-eluting stent, but the authors did not specify the type of drug-eluting stent.

The post hoc analysis² included a subgroup of 234 of 4,998 patients receiving aspirin and 236 of 5,012 patients receiving placebo initiated within 4 hours before surgery and continued postoperatively. The primary outcome measured was the rate of death or nonfatal myocardial infarction within 30 days after surgery, and bleeding was a secondary outcome.

Findings. Although the overall POISE-2 study found no benefit from aspirin, in the subgroup who had previously undergone percutaneous coronary intervention, aspirin significantly reduced the risk of the primary outcome, which occurred in 6% vs 11.5% of the patients:

• Absolute risk reduction 5.5% (95% confidence interval 0.4%–10.5%)
• Hazard ratio 0.50 (0.26–0.95).

The reduction was primarily due to fewer myocardial infarctions:

• Absolute risk reduction 5.9% (1.0%–10.8%)
• Hazard ratio 0.44 (0.22–0.87).

The type of stent had no effect on the primary outcome, although this subgroup analysis had limited power. In the nonpercutaneous coronary intervention subgroup, there was no significant difference in outcomes between the aspirin and placebo groups. This subgroup analysis was underpowered to evaluate the effect of aspirin on the composite of major and life-threatening bleeding in patients with prior percutaneous coronary intervention, which was reported as “uncertain” due to wide confidence intervals (absolute risk increase 1.3%, 95% confidence interval –2.6% to 5.2%), but the increased risk of major or life-threatening bleeding with aspirin demonstrated in the overall POISE-2 study population likely applies:

• Absolute risk increase 0.8% (0.1%–1.6%)
• Hazard ratio 1.22 (1.01–1.48).

Limitations. This was a nonspecified subgroup analysis that was underpowered and had a relatively small sample size with few events.

Conclusion. In the absence of a very high bleeding risk, continuing aspirin perioperatively in patients with prior percutaneous coronary intervention undergoing noncardiac surgery is more likely to result in benefit than harm. This finding is in agreement with current recommendations from the American College Cardiology and American Heart Association (class I; level of evidence C).³

Despite advances in anesthesia and surgical techniques, about 1% of patients over age 45 die within 30 days of noncardiac surgery.⁴ Studies have demonstrated a high mortality rate in patients who experience myocardial injury after noncardiac surgery (MINS), defined as elevations of troponin T with or without ischemic symptoms or electrocardiographic changes.⁵ Most of these studies used earlier, “non-high-sensitivity” troponin T assays. Fifth-generation, highly sensitive troponin T assays are now available that can detect troponin T at lower concentrations, but their utility in predicting postoperative outcomes remains uncertain. Two recent studies provide further insight into these issues.

[Writing Committee for the VISION Study Investigators, Devereaux PJ, Biccard BM, Sigamani A, et al. Association of postoperative high-sensitivity troponin levels with myocardial injury and 30-day mortality among patients undergoing noncardiac surgery. JAMA 2017; 317(16):1642–1651.]

The Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study⁵ was an international, prospective cohort study that initially evaluated the association between MINS and the 30-day mortality rate using a non-high-sensitivity troponin T assay (Roche fourth-generation Elecsys TnT assay) in patients age 45 or older undergoing noncardiac surgery and requiring hospital admission for at least 1 night. After the first 15,000 patients, the study switched to the Roche fifth-generation assay, with measurements at 6 to 12 hours after surgery and on postoperative days 1, 2, and 3.

A 2017 analysis by Devereaux et al⁶ included only these later-enrolled patients and correlated their high-sensitivity troponin T levels with 30-day mortality rates. Patients with a level 14 ng/L or higher, the upper limit of normal in this study, were also assessed for ischemic symptoms and electrocardiographic changes. Although not required by the study, more than 7,800 patients had their troponin T levels measured before surgery, and the absolute change was also analyzed for an association with the 30-day mortality rate.

Findings. Of the 21,842 patients, about two-thirds underwent some form of major surgery; some of them had more than 1 type. A total of 1.2% of the patients died within 30 days of surgery.

Based on their analysis, the authors proposed that MINS be defined as:

• A postoperative troponin T level of 65 ng/L or higher, or
• A level in the range of 20 ng/L to less than 65 ng/L with an absolute increase from the preoperative level at least 5 ng/L, not attributable to a nonischemic cause.

Seventeen percent of the study patients met these criteria, and of these, 21.7% met the universal definition of myocardial infarction, although only 6.9% had symptoms of it.

Limitations. Only 40.4% of the patients had a preoperative high-sensitivity troponin T measurement for comparison, and in 13.8% of patients who had an elevated perioperative measurement, their preoperative value was the same or higher than their postoperative one. Thus, the incidence of MINS may have been overestimated if patients were otherwise not known to have troponin T elevations before surgery.

[Puelacher C, Lurati Buse G, Seeberger D, et al. Perioperative myocardial injury after noncardiac surgery: incidence, mortality, and characterization. Circulation 2018; 137(12):1221–1232.]

Puelacher et al⁷ investigated the prevalence of MINS in 2,018 patients at increased cardiovascular risk (age ≥ 65, or age ≥ 45 with a history of coronary artery disease, peripheral vascular disease, or stroke) who underwent major noncardiac surgery (planned overnight stay ≥ 24 hours) at a university hospital in Switzerland. Patients had their troponin T measured with a high-sensitivity assay within 30 days before surgery and on postoperative days 1 and 2.

Instead of MINS, the investigators used the term “perioperative myocardial injury” (PMI), defined as an absolute increase in troponin T of at least 14 ng/L from before surgery to the peak postoperative reading. Similar to MINS, PMI did not require ischemic features, but in this study, noncardiac triggers (sepsis, stroke, or pulmonary embolus) were not excluded.

Findings. PMI occurred in 16% of surgeries, and of the patients with PMI, 6% had typical chest pain and 18% had any ischemic symptoms. Unlike in the POISE-2 study discussed above, PMI triggered an automatic referral to a cardiologist.

The unadjusted 30-day mortality rate was 8.9% among patients with PMI and 1.5% in those without. Multivariable logistic regression analysis showed an adjusted hazard ratio for 30-day mortality of 2.7 (95% CI 1.5–4.8) for those with PMI vs without, and this difference persisted for at least 1 year.

In patients with PMI, the authors compared the 30-day mortality rate of those with no ischemic signs or symptoms (71% of the patients) with those who met the criteria for myocardial infarction and found no difference. Patients with PMI triggered by a noncardiac event had a worse prognosis than those with a presumed cardiac etiology.

Limitations. Despite the multivariate analysis that included adjustment for age, nonelective surgery, and Revised Cardiac Risk Index (RCRI), the increased risk associated with PMI could simply reflect higher risk at baseline. Although PMI resulted in automatic referral to a cardiologist, only 10% of patients eventually underwent coronary angiography; a similar percentage were discharged with additional medical therapy such as aspirin, a statin, or a beta-blocker. The effect of these interventions is not known.

Conclusions. MINS is common and has a strong association with mortality risk proportional to the degree of troponin T elevation using high-sensitivity assays, consistent with data from previous studies of earlier assays. Because the mechanism of MINS may differ from that of myocardial infarction, its prevention and treatment may differ, and it remains unclear how serial measurement in postoperative patients should change clinical practice.

The recently published Dabigatran in Patients With Myocardial Injury After Non-cardiac Surgery (MANAGE) trial⁸ suggests that dabigatran may reduce arterial and venous complications in patients with MINS, but the study had a number of limitations that may restrict the clinical applicability of this finding.

While awaiting further clinical outcomes data, pre- and postoperative troponin T measurement may be beneficial in higher-risk patients (such as those with cardiovascular disease or multiple RCRI risk factors) if the information will change perioperative management.

Intraoperative hypotension is associated with organ ischemia, which may cause postoperative myocardial infarction, myocardial injury, and acute kidney injury.⁹ Traditional anesthesia practice is to maintain intraoperative blood pressure within 20% of the preoperative baseline, based on the notion that hypertensive patients require higher perfusion pressures.

[Futier E, Lefrant J-Y, Guinot P-G, et al. Effect of individualized vs standard blood pressure management strategies on postoperative organ dysfunction among high-risk patients undergoing major surgery: a randomized clinical trial. JAMA 2017; 318(14):1346–1357.]

Futier et al¹⁰ sought to address uncertainty in intraoperative and immediate postoperative management of systolic blood pressure. In this multicenter, randomized, parallel-group trial, 298 patients at increased risk of postoperative renal complications were randomized to blood pressure management that was either “individualized” (within 10% of resting systolic pressure) or “standard” (≥ 80 mm Hg or ≥ 40% of resting systolic pressure) from induction to 4 hours postoperatively.

Blood pressure was monitored using radial arterial lines and maintained using a combination of intravenous fluids, norepinephrine (the first-line agent for the individualized group), and ephedrine (in the standard treatment group only). The primary outcome was a composite of systemic inflammatory response syndrome (SIRS) and organ dysfunction affecting at least 1 organ system (cardiovascular, respiratory, renal, hematologic, or neurologic).

Findings. Data on the primary outcome were available for 292 of 298 patients enrolled. The mean age was 70 years, 15% were women, and 82% had previously diagnosed hypertension. Despite the requirement for an elevated risk of acute kidney injury, only 13% of the patients had a baseline estimated glomerular filtration rate of less than 60 mL/min/1.73 m², and the median was 88 mL/min/1.73 m². Ninety-five percent of patients underwent abdominal surgery, and 50% of the surgeries were elective.

The mean systolic blood pressure was 123 mm Hg in the individualized treatment group compared with 116 mm Hg in the standard treatment group. Despite this small difference, 96% of individualized treatment patients received norepinephrine, compared with 26% in the standard treatment group.

The primary outcome of SIRS with organ dysfunction occurred in 38.1% of patients in the individualized treatment group and 51.7% of those in the standard treatment group. After adjusting for center, surgical urgency, surgical site, and acute kidney injury risk index, the relative risk of developing SIRS in those receiving individualized management was 0.73 (P = .02). Renal dysfunction (based on Acute Dialysis Quality Initiative criteria¹¹) occurred in 32.7% of individualized treatment patients and 49% of standardized treatment patients.  

Limitations of this study included differences in pharmacologic approach to maintain blood pressure in the 2 protocols (ephedrine and fluids vs norepinephrine) and a modest sample size.

Conclusions. Despite this, the difference in organ dysfunction was striking, with a number needed to treat of only 7 patients. This intervention extended 4 hours postoperatively, a time when many of these patients have left the postanesthesia care unit and have returned to hospitalist care on inpatient wards.

While optimal management of intraoperative and immediate postoperative blood pressure may not be settled, this study suggests that even mild relative hypotension may justify immediate action. Further studies may be useful to delineate high- and low-risk populations, the timing of greatest risk, and indications for intraarterial blood pressure monitoring.

[Salmasi V, Maheswari K, Yang D, et al. Relationship between intraoperative hypotension, defined by either reduction from baseline or absolute thresholds, and acute kidney and myocardial injury after noncardiac surgery: a retrospective cohort analysis. Anesthesiology 2017; 126(1):47–65.]

This retrospective cohort study¹² assessed the association between myocardial or kidney injury and absolute or relative thresholds of intraoperative mean arterial pressure. It included 57,315 adults who underwent inpatient noncardiac surgery, had a preoperative and at least 1 postoperative serum creatinine measurement within 7 days, and had blood pressure recorded in preoperative appointments within 6 months. Patients with chronic kidney disease (glomerular filtration rate < 60 mL/min/1.73 m²) and those on dialysis were excluded. The outcomes were MINS⁵ and acute kidney injury as defined by the Acute Kidney Injury Network.⁹

Findings. A mean arterial pressure below an absolute threshold of 65 mm Hg or a relative threshold of 20% lower than baseline value was associated with myocardial and kidney injury. At each threshold, prolonged periods of hypotension were associated with progressively increased risk.

An important conclusion of the study was that relative thresholds of mean arterial pressure were not any more predictive than absolute thresholds. Absolute thresholds are easier to use intraoperatively, especially when baseline values are not available. The authors did not find a clinically significant interaction between baseline blood pressure and the association of hypotension and myocardial and kidney injury.

Limitations included use of cardiac enzymes postoperatively to define MINS. Since these were not routinely collected, clinically silent myocardial injury may have been missed. Baseline blood pressure may have important implications in other forms of organ injury (ie, cerebral ischemia) that were not studied.

Summary. The lowest absolute mean arterial pressure is as predictive of postoperative myocardial and kidney injury as the relative pressure reduction, at least in patients with normal renal function. Limiting exposure to intraoperative hypotension is important. Baseline blood pressure values may have limited utility for intraoperative management.

In combination, these studies confirm that intraoperative hypotension is a predictor of postoperative organ dysfunction, but the definition and management remain unclear. While aggressive intraoperative management is likely beneficial, how to manage the anti­hypertensive therapy the patient has been taking as an outpatient when he or she comes into the hospital for surgery remains uncertain.

Perioperative stroke is an uncommon, severe complication of noncardiac surgery. The pathophysiology has been better defined in cardiac than in noncardiac surgeries. In nonsurgical patients, patent foramen ovale (PFO) is associated with stroke, even in patients considered to be at low risk.¹³ Perioperative patients have additional risk for venous thromboembolism and may have periprocedural antithrombotic medications altered, increasing their risk of paradoxical embolism through the PFO.

[Ng PY, Ng AK, Subramaniam B, et al. Association of preoperatively diagnosed patent foramen ovale with perioperative ischemic stroke. JAMA 2018; 319(5):452–462.]

This retrospective cohort study of noncardiac surgery patients at 3 hospitals¹⁴ sought to determine the association of preoperatively diagnosed PFO with the risk of perioperative ischemic stroke identified by International Classification of Diseases diagnoses.

Of 150,198 patients, 1.0% had a preoperative diagnosis of PFO, and at baseline, those with PFO had significantly more comorbidities than those without PFO. Stroke occurred in 3.2% of patients with PFO vs 0.5% of those without. Patients known to have a PFO were much more likely to have cardiovascular and thromboembolic risk factors for stroke. In the adjusted analysis, the absolute risk difference between groups was 0.4% (95% CI 0.2–0.6%), with an estimated perioperative stroke risk of 5.9 per 1,000 in patients with known patent foramen ovale and 2.2 per 1,000 in those without. A diagnosis of PFO was also associated with increased risk of large-vessel-territory stroke and more severe neurologic deficit.

Further attempts to adjust for baseline risk factors and other potential bias, including a propensity score-matched cohort analysis and an analysis limited to patients who had echocardiography performed in the same healthcare system, still showed a higher risk of perioperative stroke among patients with preoperatively detected patent foramen ovale.

Limitations. The study was retrospective and observational, used administrative data, and had a low rate of PFO diagnosis (1%), compared with about 25% in population-based studies.¹⁵ Indications for preoperative echocardiography are unknown. In addition, the study specifically examined preoperatively diagnosed PFO, rather than including those diagnosed in the postoperative period.

Discussion. How does this study affect clinical practice? The absolute stroke risk was increased by 0.4% in patients with PFO compared with those without. Although this is a relatively small increase, millions of patients undergo noncardiac surgery annually. The risks of therapeutic anticoagulation or PFO closure are likely too high in this context; however, clinicians may approach the perioperative management of antiplatelet agents and venous thromboembolism prophylaxis in patients with known PFO with additional caution.

HOW DOES TIMING OF EMERGENCY SURGERY AFTER PRIOR STROKE AFFECT OUTCOMES?

A history of stroke or transient ischemic attack is a known risk factor for perioperative vascular complications. A recent large cohort study demonstrated that a history of stroke within 9 months of elective surgery was associated with increased adverse outcomes.¹⁶ Little is known, however, of the perioperative risk in patients with a history of stroke who undergo emergency surgery.

[Christiansen MN, Andersson C, Gislason GH, et al. Risks of cardiovascular adverse events and death in patients with previous stroke undergoing emergency noncardiac, nonintracranial surgery: the importance of operative timing. Anesthesiology 2017; 127(1):9–19.]

In this study,¹⁷ all emergency noncardiac and nonintracranial surgeries from 2005 to 2011 were analyzed using multiple national patient registries in Denmark according to time elapsed between previous stroke and surgery. Primary outcomes were 30-day all-cause mortality and 30-day major adverse cardiac events (MACE), defined as nonfatal ischemic stroke, nonfatal myocardial infarction, and cardiovascular death. Statistical analysis to assess the risk of adverse outcomes included logistic regression models, spline analyses, and propensity-score matching.

Findings. The authors identified 146,694 emergency surgeries, with 7,861 patients (5.4%) having had a previous stroke (transient ischemic attacks and hemorrhagic strokes were not included). Rates of postoperative stroke were as follows:

• 9.9% in patents with a history of ischemic stroke within 3 months of surgery
• 2.8% in patients with a history of stroke 3 to 9 months before surgery
• 0.3% in patients with no previous stroke.

The risk plateaued when the time between stroke and surgery exceeded 4 to 5 months.¹⁵

Interestingly, in patients who underwent emergency surgery within 14 days of stroke, the risk of MACE was significantly lower immediately after surgery (1–3 days after stroke) compared with surgery that took place 4 to 14 days after stroke. The authors hypothesized that because cerebral autoregulation does not become compromised until approximately 5 days after a stroke, the risk was lower 1 to 3 days after surgery and increased thereafter.

Limitations of this study included the possibility of residual confounding, given its retrospective design using administrative data, not accounting for preoperative antithrombotic and anticoagulation therapy, and lack of information regarding the etiology of recurrent stroke (eg, thromboembolic, atherothrombotic, hypoperfusion).

Conclusions. Although it would be impractical to postpone emergency surgery in a patient who recently had a stroke, this study shows that the incidence rates of postoperative recurrent stroke and MACE are high. Therefore, it is important that the patient and perioperative team be aware of the risk. Further research is needed to confirm these estimates of postoperative adverse events in more diverse patient populations.

Primary care physicians are tasked with a wide variety of issues affecting men. This article reviews the latest research in 4 areas of men’s health commonly addressed in primary care:

• Medical management of benign prostatic hyperplasia (BPH)
• Prostate cancer screening and treatment
• Medical management of erectile dysfunction
• Use of supplements.

See related commentary

MEDICAL MANAGEMENT OF BPH

An 84-year-old man with a history of hypertension, type 2 diabetes, hyperlipidemia, BPH, mild cognitive impairment, and osteoarthritis presents for a 6-month follow-up, accompanied by his son.

Two years ago he was started on a 5-alpha reductase inhibitor and an alpha-blocker for worsening BPH symptoms. His BPH symptoms are currently under control, with an American Urological Association (AUA) symptom index score of 7 of a possible 35 (higher scores being worse).

However, both the patient and son are concerned about the number of medications he is on and wonder if some could be eliminated.

Assessment tools

BPH is a common cause of lower urinary tract symptoms in older men. Evidence-based tools to help the clinician and patient decide on when to consider treatment for symptoms are:

• The AUA symptom index¹
• The International Prostate Symptom Score (IPSS).²

An AUA symptom index score or IPSS score of 8 through 19 of a possible 35 is consistent with moderate symptoms, while a score of 20 or higher indicates severe symptoms.

Combination therapy or monotherapy?

Monotherapy with an alpha-blocker or a 5-alpha reductase inhibitor is often the first-line treatment for BPH-related lower urinary tract symptoms.³ However, combination therapy with both an alpha-blocker and a 5-alpha reductase inhibitor is another evidence-based option.

The Medical Therapy of Prostatic Symptoms study,⁴ a randomized controlled trial, reported that long-term combination therapy reduced the risk of BPH clinical progression better than monotherapy. The same trial also found that either combination therapy or finasteride alone (a 5-alpha reductase inhibitor) reduced the risk of acute urinary retention and the future need for invasive therapy.

Monotherapy after a period of combination therapy?

There is also evidence to support switching from combination to monotherapy after an initial treatment period.

Matsukawa et al⁵ examined the effects of withdrawing the alpha-blocker from BPH combination therapy in a study in 140 patients. For 12 months, all patients received the alpha-blocker silodosin and the 5-alpha reductase inhibitor dutasteride. At 12 months, the remaining 132 patients (8 patients had been lost to follow-up) were randomized to continue combination therapy or to take dutasteride alone for another 12 months. They were evaluated at 0, 12, and 24 months by questionnaires (the IPSS and Overactive Bladder Symptom Score) and urodynamic testing (uroflowmetry, cystometrography, and pressure-flow studies).

There were no significant differen­ces in subjective symptoms and bladder outlet obstruction between patients who continued combination therapy and those who switched to dutasteride monotherapy. In the monotherapy group, those whose symptoms worsened weighed more (68.8 kg vs 62.6 kg, P =.002) and had a higher body mass index (BMI) (26.2 kg/m² vs 22.8 kg/m², P < .001) than those whose symptoms stayed the same or got better.

These findings of successful alpha-blocker withdrawal were consistent with those of other studies.

The Symptom Management After Reducing Therapy study⁶ showed that 80% of men with an IPSS score less than 20 who changed to dutasteride monotherapy did not have a noticeable worsening of their symptoms.

Baldwin et al⁷ noted similar success after withdrawing the alpha-blocker doxazosin in patients on finasteride.

Review all medications

The National Health and Nutrition Examination Survey noted that the estimated prevalence of polypharmacy increased from 8% in 1999 to 15% in 2011.⁸ Many commonly used medications, such as decongestants, antihistamines, and anticholinergic agents, can worsen BPH symptoms,⁹ so it is reasonable to consistently review the patient’s medications to weigh the risks and benefits and determine which ones align with the patient’s personal care goals.

BPH: Take-home points

• Combination therapy with an alpha-blocker and a 5-alpha reductase inhibitor is an effective regimen for BPH.
• Polypharmacy is a significant problem in the elderly.
• Withdrawing the alpha-blocker component from BPH combination therapy can be considered after 1 year of combination therapy in patients whose symptoms have been well controlled.

A 60-year-old patient calls you after receiving his laboratory testing report from his insurance physical. His prostate-specific antigen (PSA) level is 5.1 ng/mL, and he has several questions:

• Should he have agreed to the screening?
• How effective is the screening?
• What are the next steps?

Is PSA screening useful?

Over the last few years, there has been great debate as to the utility of screening for prostate cancer.

The US Centers for Disease Control and Prevention¹⁰ reported that in 2014, an estimated 172,258 men in the United States were diagnosed with prostate cancer, but only 28,343 men died of it. These statistics support the notion that screening programs may be detecting what might otherwise be a silent disease.

The US Preventive Services Task Force (USPSTF)¹¹ recommends against blanket PSA screening, in view of the low probability that it reduces the risk of death from prostate cancer. For men ages 55 through 69, current guidelines give a grade C recommendation to PSA screening, meaning there is moderate agreement that the benefit is likely small, and screening should be selectively offered based on professional judgment and patient preference. In men ages 70 and older who are not at high risk, the guideline gives screening a grade D recommendation, meaning there is moderate evidence that there is no benefit from the practice. This is a change from the 2012 USPSTF guidelines,¹² which gave a grade D recommendation to PSA screening for all ages.

The American Urological Association¹³ recommends against PSA screening in men under age 40 or ages 70 and older. It does not recommend routine screening in those ages 40 to 54 at average risk, but it says the decision should be individualized in this age group in those at higher risk (eg, with a positive family history, African American). At ages 55 through 69, it recommends shared decision-making, taking into account cancer risk and life expectancy. In those who opt for screening, an interval of 2 years or more may be preferred over annual screening to reduce the risk of overdiagnosis.

The USPSTF recommendations rely heavily on data from 2 trials: the European Randomized Study of Screening for Prostate Cancer (ERSPC)¹⁴ and the Prostate, Lung, Colorectal, and Ovarian Screening (PLCO) trial.¹⁵

The ERSPC¹⁴ demonstrated that screening for prostate cancer reduced deaths from prostate cancer by 20%, with an absolute risk difference of 0.71 deaths per 1,000 men; 1,410 men would need to be screened and 48 additional cases of prostate cancer would need to be treated to prevent 1 death from prostate cancer. Screening also decreased the risk of developing metastatic disease by 30%.¹⁶ On the negative side, screening increased the risk of overdiagnosis and other harms such as bleeding, sepsis, and incontinence.

The PLCO trial,¹⁵ in contrast, found no difference in death rates between men randomly assigned to annual screening and those assigned to usual care. Differences between the trial results were thought to be due to different practice settings as well as study implementation and compliance.

Tsodikov et al¹⁷ reanalyzed data from the ERSPC and the PLCO trial using 3 different mathematical models to estimate the effects of screening in both trials compared with no screening. The analysis found no evidence that the effects of screening vs not screening differed between the 2 trials, ultimately concluding that PSA screening reduced prostate cancer deaths by 25% to 32%, which the authors inferred was primarily a result of earlier detection of cancer.

The Cluster Randomized Trial of PSA Testing for Prostate Cancer,¹⁸ published in March 2018, explored the effect of single PSA screening vs no screening on prostate cancer mortality rates in 419,582 men ages 50 through 69. Although screening detected more cases of low-risk prostate cancer, there was no significant difference in prostate cancer mortality rates after a median follow-up of 10 years. However, 10% to 15% of the control group was estimated to have also been screened, and these results do not directly speak to the efficacy of serial PSA screening.

Extended follow-up of this trial is planned to report on long-term survival benefits and whether screening lowers the risk of metastasis.

Imaging-guided prostate biopsy

Once a patient is found to have an elevated PSA level, standard practice has been to perform transrectal ultrasonography to obtain 12 core biopsy samples. The results indicate whether the prostate contains cancer, how aggressive the cancer is (Gleason score), and whether there is extracapsular extension.

In the past, magnetic resonance imaging (MRI) of the prostate before biopsy was thought to be too costly, and many insurance plans do not currently cover it.

Pahwa et al,¹⁹ however, in a cost-effectiveness study using a decision-analysis model, found that using MRI to detect lesions and then guide biopsy by triaging patients into proper treatment pathways added health benefits in a cost-effective manner in 94.05% of simulations. These benefits were found across all age groups.

This study demonstrated that doctors could use MRI to better evaluate patients for potentially harmful lesions. If a focus of cancer is found, it can be biopsied; if no cancer is seen on MRI, the patient can avoid biopsy completely. Additionally, though MRI tended to miss low-risk cancers, these cancers are thought to disproportionately lead to higher healthcare costs through unnecessary treatment. Therefore, a negative MRI study was believed to be an excellent sign that the patient does not have aggressive prostate cancer. This approach led to a net gain of 0.251 additional quality-adjusted life years compared with the standard biopsy strategy.

The Prostate MRI Imaging Study²⁰ also found MRI to be effective in the prostate cancer workup. In this trial, 576 men who had never undergone biopsy underwent multiparametric MRI, transrectal ultrasonography-guided biopsy, and the reference standard, ie, transperineal template prostate mapping biopsy. Of those who underwent biopsy, 71% received a diagnosis of prostate cancer, and 40% had clinically significant disease. In patients with clinically significant disease, MRI was more sensitive than ultrasonography-guided biopsy (93% vs 48%, P < .0001) but less specific (41% vs 96%, P < .0001).

Based on these findings, if biopsy were performed only in those who had suspicious lesions on MRI, 27% of men with elevated PSA could avoid biopsy and its potential complications such as bleeding and sepsis, which occurred in 5.9% of the biopsy group.

The Prostate Evaluation for Clinically Important Disease: Sampling Using Image Guidance or Not? trial²¹ more recently studied MRI with or without targeted biopsy vs standard transrectal ultrasonography-guided biopsy in 500 men who had not undergone biopsy before, and reported similar results. MRI with or without biopsy led to fewer biopsies and less overdetection of clinically insignificant prostate cancers compared with the standard approach. Furthermore, those in the MRI-targeted biopsy group were 13% less likely to receive a diagnosis of clinically insignificant cancer than those who received the standard biopsy (adjusted difference −13 percentage points, 95% confidence interval [CI] −19 to −7, P < .001).

Together, these data provide another argument for adding multiparametric MRI to the workup of men with an elevated PSA level.

Surveillance vs treatment for prostate cancer

Once prostate cancer is diagnosed, surveillance is becoming an increasingly common management strategy.

The Prostate Cancer Intervention Versus Observation Trial (PIVOT),²² one of the largest and longest trials involving cancer patients, offered further evidence that active surveillance and less intervention for men with prostate cancer is a better approach in many cases. This trial compared prostatectomy and observation alone in a randomized fashion. Inclusion for the study required men to be medically fit for radical prostatectomy, along with having histologically confirmed localized prostate cancer (stage T1-T2NxM0 in the tumor-node-metastasis classification system) of any grade diagnosed within the last 12 months.

During 19.5 years of follow-up, 223 (61.3%) of the 364 men randomly assigned to radical prostatectomy died, compared with 245 (66.8%) of 367 men in the observation group; the difference was not statistically different (P = .06). Only 9.4% of the deaths were due to prostate cancer, 7.4% in the surgery group and 11.4% in the observation group (P = .06).

Surgery was associated with a lower all-cause mortality rate than observation in the subgroup of patients with intermediate-risk prostate cancer (defined as PSA 10–20 ng/mL and a Gleason score of 7). Surgery was also associated with less disease progression.²²

This finding is in line with previous data from the Scandinavian Prostate Cancer Group Study Number 4,²³ as well as the much larger Prostate Testing for Cancer and Treatment (ProtecT) trial,²⁴ both of which reported that metastasis was 1.5 and 2.6 times as common, respectively, in participants in the active surveillance groups. However, in the PIVOT trial, those in the surgery group were significantly more likely than those in the observation group to have erectile dysfunction and urinary incontinence at 10 years.

Therefore, in men with localized disease and in those with low-risk PSA levels, both the PIVOT and ProtecT trials suggest that death from prostate cancer is uncommon and that observation may be more appropriate.

Prostate cancer: Take-home points

• A new look at 2 large trials of PSA screening strengthened evidence that testing in the right patient population can reduce deaths from prostate cancer, but a third recently published trial that found no benefit from 1-time screening may reopen debate on the topic.
• MRI offers a better method than ultrasonography-guided biopsy to triage patients thought to be at high risk of prostate cancer and tends to limit costly overtreatment of disease that likely would not cause death.
• Surgery for prostate cancer may not prolong life but could reduce disease progression, at the risk of more adverse effects.
• Shared decision-making should be practiced when deciding whether to use active surveillance or active treatment of diagnosed prostate cancer.

A 62-year-old man with hypertension, hyperlipidemia, peripheral artery disease, and type 2 diabetes presents for a 6-month follow-up. His medications include aspirin, metformin, lisinopril, and atorvastatin, all of which he takes without problems. Over the past several months, he has noticed that his erections are not adequate for sexual intercourse. He recently heard that a generic version of sildenafil has just become available, and he wonders if it might benefit him.

Erectile dysfunction is common, associated with chronic diseases

Erectile dysfunction, ie, persistent inability to obtain and maintain an erection sufficient to permit satisfactory sexual intercourse,^25,26 is estimated to affect nearly 20% of men over the age of 20 and 75% of men over the age of 75.²⁷

In age-adjusted models, erectile dysfunction has been shown²⁸ to be associated with:

• History of cardiovascular disease (odds ratio [OR] 1.63, 95% CI 1.02–2.63)
• Diabetes (OR 3.90, 95% CI 2.16–7.04)
• Treated hypertension vs no hypertension (OR 2.22, 95% CI 1.30–3.80)
• Current smoking vs never smoking (OR 1.63, 95% CI 1.01–2.62)
• BMI greater than 30 kg/m² vs less than 25 kg/m² (OR 1.80, 95% CI 1.03–3.14).

Because of the strong association between cardiovascular disease and erectile dysfunction, the presence of one often suggests the need to screen for the other.²⁹ While tools such as the International Index of Erectile Function (IIEF-5) have been developed to evaluate erectile dysfunction, it is most often diagnosed on the basis of clinical impression, while validated assessment methods are reserved for clinical trials.²⁸

Multiple causes of erectile dysfunction

Erectile dysfunction arises from inadequate penile tissue response to a sexual signal. The response can be disrupted at several points. For example, damage to vascular smooth muscle cells (eg, from age or obesity) and endothelial cells (from smoking or diabetes) and narrowing of the vascular lumen (from atherosclerosis or hypertension) have all been shown to impair engorgement of the corpus cavernosum.³⁰ In addition, denervation from prostate surgery or spinal trauma and psychogenic causes should be recognized in discussions with patients.

Drugs for erectile dysfunction

Pharmacologic management of erectile dysfunction includes oral, sublingual, intracavernosal, and intraurethral therapies.³¹ Treatment in primary care settings usually includes addressing underlying chronic diseases³² and prescribing phosphodiesterase-5 inhibitors (sildenafil, tadalafil, vardenafil, and avanafil). These drugs work by increasing local concentrations of cyclic guanosine monophosphate in the corpus cavernosum to induce vasodilation.³³

While these 4 drugs are still patent-protected, a manufacturer has been allowed to introduce a generic version of sildenafil into US markets, and a generic version of tadalafil is expected to be available soon.

Sildenafil, tadalafil, and vardenafil have been studied and found to have some degree of effectiveness in erectile dysfunction caused by damage to the penile vasculature, denervation, and spinal cord injury.³⁴ All drugs of this class have adverse effects including headache, facial flushing, and nasal congestion, but the drugs are generally well tolerated.³⁵

Sildenafil and tadalafil improve IIEF-5 scores by a similar margin, raising scores on the erectile domain subsection from approximately 14 of a possible 30 to approximately 24 of 30 in a trial of both drugs.³⁶ However, multiple crossover studies comparing the 2 drugs have shown that nearly 75% of patients prefer tadalafil to sildenafil,^36,37 perhaps because of tadalafil’s longer duration of action.³⁴

There is little evidence to suggest that vardenafil is more effective or more often preferred by patients than tadalafil or sidenafil.^34,38 And though data on the newest drug on the market, avanafil, are limited, a meta-analysis concluded that it may be less effective than tadalafil and without significant differences in terms of safety.³⁹

Other treatments

Lifestyle modifications, especially smoking cessation and exercise, have been shown to reduce the risk of erectile dysfunction with varying effect sizes across studies.^40–42 Moreover, factors such as obesity, alcohol use, and smoking may cause irreversible harm, and thus a healthy lifestyle should be encouraged.⁴¹

While there is only weak evidence for the use of psychological interventions alone for treating most types of erectile dysfunction, one meta-analysis found that the combination of psychological intervention and a phosphodiesterase-5 inhibitor improved sexual satisfaction more than drug therapy alone.⁴³

Erectile dysfunction: Take-home points

• Erectile dysfunction is common, affecting nearly 20% of men over the age of 20 and over 75% of men over the age of 75.
• Erectile dysfunction is often associated with chronic disease and may suggest the need to screen for cardiovascular disease.
• Treating underlying chronic diseases may help, and phosphodiesterase-5 inhibitors are effective; tadalafil may be most often preferred.

A 68-year-old man with a history of hypertension, BPH, and erectile dysfunction presents for a 6-month follow-up. His medication use includes lisinopril, which he takes without problems. He denies any new physical symptoms. His physical examination is unremarkable. He says he has heard about supplements that might help with his sexual performance and hopes to discuss recommendations during the visit.

A burgeoning, unregulated industry

Since the passage of the Dietary Supplement and Health Education Act in 1994, a law that decreased oversight of the supplement industry, spending on supplements has skyrocketed to over $41.1 billion each year.⁴⁴ Advertisements for these products typically claim that they improve general mental and physical health, sexual and romantic performance, leanness, and muscularity.⁴⁵ A national survey of men ages 57 and older reported that the most popular products were aimed at nutrition (such as multivitamins), cardiovascular health (such as omega-3 fatty acids), and chronic conditions (such as saw palmetto for BPH).⁴⁶

Little evidence of efficacy

There is little evidence to support the use of most supplements to improve men’s health. For example, a study in 82,405 men found no association between mortality rates and multivitamin use (hazard ratio [HR] 1.07, 95% CI 0.96–1.19).⁴⁷ Even for specific uses, such as cognitive performance, randomized trials exploring the effects of multivitamins in men have been largely negative.⁴⁸

The positive trials that have been reported are often of low quality and are funded by supplement manufacturers. For example, one of the few trials that reported a positive association between multivitamin supplementation and cognition in men was underpowered (N = 51) and found improvement in only 1 of 19 cognitive domains.⁴⁹ Despite the poor design and results to the contrary, this industry-funded study nevertheless concluded that multivitamins may play a role in improving elements of memory.

Evidence of possible harm from antioxidants

While not always specific to men, many meta-analyses have explored the effects of antioxidant supplements on cardiovascular and mortality risk. Most of them concluded that antioxidant supplements have no benefit and that some may actually be harmful.

For example, multiple meta-analyses of vitamin E supplementation found no cardiovascular benefit but possible increases in all-cause mortality rates in those taking high doses (risk ratio 1.04, 95% CI 1.01–1.07).^50,51

Another meta-analysis of 180,938 partici­pants in high-quality studies found an increased risk of all-cause mortality associated with independent intake of several antioxidant vitamins, including beta-carotene (risk ratio 1.07, 95% CI 1.02–1.11) and vitamin A (risk ratio 1.16, 95% CI 1.10–1.24), while intake of vitamin C and selenium had no impact on mortality.⁵²

Similarly, although nearly 10% of US adults report taking omega-3 fatty acid supplements, a review of 24 randomized controlled trials and meta-analyses published between 2005 and 2012 concluded that only 2 supported the use of these supplements for any health benefit.⁵³

Can supplements improve sexual function, prostate health?

To improve sexual function. A 2015 narrative review of the ingredients in General Nutrition Center’s top 30 best-selling products targeted at improving men’s sexual performance (including improving libido and erectile dysfunction) found only poor evidence for any efficacy.⁵⁴ The few studies that did support the use of select supplements, including B vitamins in people with diabetes, L-arginine, and yohimbine, were deemed to be of poor quality or showed a smaller effect size compared with standard medical therapy.

To prevent prostate cancer. Studies of supplement use to improve prostate health have had mixed results. For example, multiple large case-control studies have suggested that taking vitamin D^55,56 or vitamin C⁵⁷ is not associated with prostate cancer risk, while increased vitamin A^58,59 and E^60,61 intake is associated with inconsistent increases in prostate cancer risk.

In the Selenium and Vitamin E Cancer Prevention Trial,⁶² a randomized controlled trial in 35,533 men, those assigned to receive vitamin E supplementation were 17% more likely to get prostate cancer than were those assigned to placebo (HR 1.17, 99% CI 1.004–1.36, P = .008).

However, there are plausible biologic links between nutraceuticals and prostate cancer. For example, studies have linked genetic polymorphisms in vitamin D receptors⁶³ as well as intake of natural androgen receptor modulators, such as the most active polyphenol in green tea,⁶⁴ to prostate cancer risk and aggressiveness in certain populations. This led a recent review to conclude that there is some biologic plausibility, but at present little epidemiologic evidence, to support any dietary supplement’s ability to broadly affect prostate cancer risk.⁶⁵

Interest continues in exploring the targeted use of nutraceuticals as adjuvant therapy in specific populations at risk of prostate cancer.^66,67

To treat BPH. There is a similar dearth of clinical or population-based evidence that supplements can broadly affect BPH symptoms. For example, in a 2012 Cochrane review of Serenoa repens (saw palmetto) utilizing only high-quality evidence, there was no evidence that supplement use significantly reduced lower urinary tract symptoms, nocturia, or peak urine flow in BPH patients, and this was true even when the supplement was taken at triple-strength doses.⁶⁸

For other diseases. There is also limited evidence that supplements can affect other chronic diseases. For example, a meta-analysis of 3,803 patients found that glucosamine, chondroitin, and their combination had no impact on joint pain or joint space narrowing in patients with osteoarthritis of the knee or hip.⁶⁹

Even when there is some evidence to suggest benefit from supplementation, study heterogeneity and varying evidence quality limit confidence in the conclusions. For example, meta-analyses suggest garlic may improve blood pressure control in those with hypertension⁷⁰ and improve lipid and blood glucose control in type 2 diabetes.⁷¹ However, most of the trials included in those systematic reviews were underpowered, with samples as low as 10 patients, and many suffered from improper design, such as inadequate blinding of researchers. In addition, these meta-analyses often do not report adverse events, suggesting that higher quality studies would be needed to adequately measure event rates. As such, there is need for caution and a case-by-case review before recommending even a seemingly benign supplement like garlic to patients.

In total, there is only limited evidence to support the efficacy of supplements across many diseases and concerns common to men in primary care. This includes improving general health, cardiovascular health, sexual functioning, or other chronic diseases. While a supplement’s placebo effect may at times provide some benefit, supplements are much less strictly regulated since the passing of the 1994 act, and even vitamin supplementation has been shown to be associated with negative health outcomes. As such, a patient’s use of supplements requires careful consideration and shared decision-making.

Supplements: Take-home points

• Supplements are only loosely regulated by the federal government.
• There is some biologic but limited epidemiologic evidence for the use of multivitamins to improve cognition or mortality rates; for the use of antioxidant vitamins or omega-3 fatty acids to improve cardiovascular health; for the use of any of the top-selling sexual enhancement supplements to improve libido or erectile function; and for the use of vitamins or other supplements for improving BPH or reducing prostate cancer risk. Using supplements may in some cases be harmful.
• Given the heterogeneity of studies of supplements to manage chronic diseases and a lack of reporting of adverse events, careful consideration is needed when recommending supplements to patients.

Primary care physicians are tasked with a wide variety of issues affecting men. This article reviews the latest research in 4 areas of men’s health commonly addressed in primary care:

• Medical management of benign prostatic hyperplasia (BPH)
• Prostate cancer screening and treatment
• Medical management of erectile dysfunction
• Use of supplements.

See related commentary

MEDICAL MANAGEMENT OF BPH

An 84-year-old man with a history of hypertension, type 2 diabetes, hyperlipidemia, BPH, mild cognitive impairment, and osteoarthritis presents for a 6-month follow-up, accompanied by his son.

Two years ago he was started on a 5-alpha reductase inhibitor and an alpha-blocker for worsening BPH symptoms. His BPH symptoms are currently under control, with an American Urological Association (AUA) symptom index score of 7 of a possible 35 (higher scores being worse).

However, both the patient and son are concerned about the number of medications he is on and wonder if some could be eliminated.

Assessment tools

BPH is a common cause of lower urinary tract symptoms in older men. Evidence-based tools to help the clinician and patient decide on when to consider treatment for symptoms are:

• The AUA symptom index¹
• The International Prostate Symptom Score (IPSS).²

An AUA symptom index score or IPSS score of 8 through 19 of a possible 35 is consistent with moderate symptoms, while a score of 20 or higher indicates severe symptoms.

Combination therapy or monotherapy?

Monotherapy with an alpha-blocker or a 5-alpha reductase inhibitor is often the first-line treatment for BPH-related lower urinary tract symptoms.³ However, combination therapy with both an alpha-blocker and a 5-alpha reductase inhibitor is another evidence-based option.

The Medical Therapy of Prostatic Symptoms study,⁴ a randomized controlled trial, reported that long-term combination therapy reduced the risk of BPH clinical progression better than monotherapy. The same trial also found that either combination therapy or finasteride alone (a 5-alpha reductase inhibitor) reduced the risk of acute urinary retention and the future need for invasive therapy.

Monotherapy after a period of combination therapy?

There is also evidence to support switching from combination to monotherapy after an initial treatment period.

Matsukawa et al⁵ examined the effects of withdrawing the alpha-blocker from BPH combination therapy in a study in 140 patients. For 12 months, all patients received the alpha-blocker silodosin and the 5-alpha reductase inhibitor dutasteride. At 12 months, the remaining 132 patients (8 patients had been lost to follow-up) were randomized to continue combination therapy or to take dutasteride alone for another 12 months. They were evaluated at 0, 12, and 24 months by questionnaires (the IPSS and Overactive Bladder Symptom Score) and urodynamic testing (uroflowmetry, cystometrography, and pressure-flow studies).

There were no significant differen­ces in subjective symptoms and bladder outlet obstruction between patients who continued combination therapy and those who switched to dutasteride monotherapy. In the monotherapy group, those whose symptoms worsened weighed more (68.8 kg vs 62.6 kg, P =.002) and had a higher body mass index (BMI) (26.2 kg/m² vs 22.8 kg/m², P < .001) than those whose symptoms stayed the same or got better.

These findings of successful alpha-blocker withdrawal were consistent with those of other studies.

The Symptom Management After Reducing Therapy study⁶ showed that 80% of men with an IPSS score less than 20 who changed to dutasteride monotherapy did not have a noticeable worsening of their symptoms.

Baldwin et al⁷ noted similar success after withdrawing the alpha-blocker doxazosin in patients on finasteride.

Review all medications

The National Health and Nutrition Examination Survey noted that the estimated prevalence of polypharmacy increased from 8% in 1999 to 15% in 2011.⁸ Many commonly used medications, such as decongestants, antihistamines, and anticholinergic agents, can worsen BPH symptoms,⁹ so it is reasonable to consistently review the patient’s medications to weigh the risks and benefits and determine which ones align with the patient’s personal care goals.

BPH: Take-home points

• Combination therapy with an alpha-blocker and a 5-alpha reductase inhibitor is an effective regimen for BPH.
• Polypharmacy is a significant problem in the elderly.
• Withdrawing the alpha-blocker component from BPH combination therapy can be considered after 1 year of combination therapy in patients whose symptoms have been well controlled.

A 60-year-old patient calls you after receiving his laboratory testing report from his insurance physical. His prostate-specific antigen (PSA) level is 5.1 ng/mL, and he has several questions:

• Should he have agreed to the screening?
• How effective is the screening?
• What are the next steps?

Is PSA screening useful?

Over the last few years, there has been great debate as to the utility of screening for prostate cancer.

The US Centers for Disease Control and Prevention¹⁰ reported that in 2014, an estimated 172,258 men in the United States were diagnosed with prostate cancer, but only 28,343 men died of it. These statistics support the notion that screening programs may be detecting what might otherwise be a silent disease.

The US Preventive Services Task Force (USPSTF)¹¹ recommends against blanket PSA screening, in view of the low probability that it reduces the risk of death from prostate cancer. For men ages 55 through 69, current guidelines give a grade C recommendation to PSA screening, meaning there is moderate agreement that the benefit is likely small, and screening should be selectively offered based on professional judgment and patient preference. In men ages 70 and older who are not at high risk, the guideline gives screening a grade D recommendation, meaning there is moderate evidence that there is no benefit from the practice. This is a change from the 2012 USPSTF guidelines,¹² which gave a grade D recommendation to PSA screening for all ages.

The American Urological Association¹³ recommends against PSA screening in men under age 40 or ages 70 and older. It does not recommend routine screening in those ages 40 to 54 at average risk, but it says the decision should be individualized in this age group in those at higher risk (eg, with a positive family history, African American). At ages 55 through 69, it recommends shared decision-making, taking into account cancer risk and life expectancy. In those who opt for screening, an interval of 2 years or more may be preferred over annual screening to reduce the risk of overdiagnosis.

The USPSTF recommendations rely heavily on data from 2 trials: the European Randomized Study of Screening for Prostate Cancer (ERSPC)¹⁴ and the Prostate, Lung, Colorectal, and Ovarian Screening (PLCO) trial.¹⁵

The ERSPC¹⁴ demonstrated that screening for prostate cancer reduced deaths from prostate cancer by 20%, with an absolute risk difference of 0.71 deaths per 1,000 men; 1,410 men would need to be screened and 48 additional cases of prostate cancer would need to be treated to prevent 1 death from prostate cancer. Screening also decreased the risk of developing metastatic disease by 30%.¹⁶ On the negative side, screening increased the risk of overdiagnosis and other harms such as bleeding, sepsis, and incontinence.

The PLCO trial,¹⁵ in contrast, found no difference in death rates between men randomly assigned to annual screening and those assigned to usual care. Differences between the trial results were thought to be due to different practice settings as well as study implementation and compliance.

Tsodikov et al¹⁷ reanalyzed data from the ERSPC and the PLCO trial using 3 different mathematical models to estimate the effects of screening in both trials compared with no screening. The analysis found no evidence that the effects of screening vs not screening differed between the 2 trials, ultimately concluding that PSA screening reduced prostate cancer deaths by 25% to 32%, which the authors inferred was primarily a result of earlier detection of cancer.

The Cluster Randomized Trial of PSA Testing for Prostate Cancer,¹⁸ published in March 2018, explored the effect of single PSA screening vs no screening on prostate cancer mortality rates in 419,582 men ages 50 through 69. Although screening detected more cases of low-risk prostate cancer, there was no significant difference in prostate cancer mortality rates after a median follow-up of 10 years. However, 10% to 15% of the control group was estimated to have also been screened, and these results do not directly speak to the efficacy of serial PSA screening.

Extended follow-up of this trial is planned to report on long-term survival benefits and whether screening lowers the risk of metastasis.

Imaging-guided prostate biopsy

Once a patient is found to have an elevated PSA level, standard practice has been to perform transrectal ultrasonography to obtain 12 core biopsy samples. The results indicate whether the prostate contains cancer, how aggressive the cancer is (Gleason score), and whether there is extracapsular extension.

In the past, magnetic resonance imaging (MRI) of the prostate before biopsy was thought to be too costly, and many insurance plans do not currently cover it.

Pahwa et al,¹⁹ however, in a cost-effectiveness study using a decision-analysis model, found that using MRI to detect lesions and then guide biopsy by triaging patients into proper treatment pathways added health benefits in a cost-effective manner in 94.05% of simulations. These benefits were found across all age groups.

This study demonstrated that doctors could use MRI to better evaluate patients for potentially harmful lesions. If a focus of cancer is found, it can be biopsied; if no cancer is seen on MRI, the patient can avoid biopsy completely. Additionally, though MRI tended to miss low-risk cancers, these cancers are thought to disproportionately lead to higher healthcare costs through unnecessary treatment. Therefore, a negative MRI study was believed to be an excellent sign that the patient does not have aggressive prostate cancer. This approach led to a net gain of 0.251 additional quality-adjusted life years compared with the standard biopsy strategy.

The Prostate MRI Imaging Study²⁰ also found MRI to be effective in the prostate cancer workup. In this trial, 576 men who had never undergone biopsy underwent multiparametric MRI, transrectal ultrasonography-guided biopsy, and the reference standard, ie, transperineal template prostate mapping biopsy. Of those who underwent biopsy, 71% received a diagnosis of prostate cancer, and 40% had clinically significant disease. In patients with clinically significant disease, MRI was more sensitive than ultrasonography-guided biopsy (93% vs 48%, P < .0001) but less specific (41% vs 96%, P < .0001).

Based on these findings, if biopsy were performed only in those who had suspicious lesions on MRI, 27% of men with elevated PSA could avoid biopsy and its potential complications such as bleeding and sepsis, which occurred in 5.9% of the biopsy group.

The Prostate Evaluation for Clinically Important Disease: Sampling Using Image Guidance or Not? trial²¹ more recently studied MRI with or without targeted biopsy vs standard transrectal ultrasonography-guided biopsy in 500 men who had not undergone biopsy before, and reported similar results. MRI with or without biopsy led to fewer biopsies and less overdetection of clinically insignificant prostate cancers compared with the standard approach. Furthermore, those in the MRI-targeted biopsy group were 13% less likely to receive a diagnosis of clinically insignificant cancer than those who received the standard biopsy (adjusted difference −13 percentage points, 95% confidence interval [CI] −19 to −7, P < .001).

Together, these data provide another argument for adding multiparametric MRI to the workup of men with an elevated PSA level.

Surveillance vs treatment for prostate cancer

Once prostate cancer is diagnosed, surveillance is becoming an increasingly common management strategy.

The Prostate Cancer Intervention Versus Observation Trial (PIVOT),²² one of the largest and longest trials involving cancer patients, offered further evidence that active surveillance and less intervention for men with prostate cancer is a better approach in many cases. This trial compared prostatectomy and observation alone in a randomized fashion. Inclusion for the study required men to be medically fit for radical prostatectomy, along with having histologically confirmed localized prostate cancer (stage T1-T2NxM0 in the tumor-node-metastasis classification system) of any grade diagnosed within the last 12 months.

During 19.5 years of follow-up, 223 (61.3%) of the 364 men randomly assigned to radical prostatectomy died, compared with 245 (66.8%) of 367 men in the observation group; the difference was not statistically different (P = .06). Only 9.4% of the deaths were due to prostate cancer, 7.4% in the surgery group and 11.4% in the observation group (P = .06).

Surgery was associated with a lower all-cause mortality rate than observation in the subgroup of patients with intermediate-risk prostate cancer (defined as PSA 10–20 ng/mL and a Gleason score of 7). Surgery was also associated with less disease progression.²²

This finding is in line with previous data from the Scandinavian Prostate Cancer Group Study Number 4,²³ as well as the much larger Prostate Testing for Cancer and Treatment (ProtecT) trial,²⁴ both of which reported that metastasis was 1.5 and 2.6 times as common, respectively, in participants in the active surveillance groups. However, in the PIVOT trial, those in the surgery group were significantly more likely than those in the observation group to have erectile dysfunction and urinary incontinence at 10 years.

Therefore, in men with localized disease and in those with low-risk PSA levels, both the PIVOT and ProtecT trials suggest that death from prostate cancer is uncommon and that observation may be more appropriate.

Prostate cancer: Take-home points

• A new look at 2 large trials of PSA screening strengthened evidence that testing in the right patient population can reduce deaths from prostate cancer, but a third recently published trial that found no benefit from 1-time screening may reopen debate on the topic.
• MRI offers a better method than ultrasonography-guided biopsy to triage patients thought to be at high risk of prostate cancer and tends to limit costly overtreatment of disease that likely would not cause death.
• Surgery for prostate cancer may not prolong life but could reduce disease progression, at the risk of more adverse effects.
• Shared decision-making should be practiced when deciding whether to use active surveillance or active treatment of diagnosed prostate cancer.

A 62-year-old man with hypertension, hyperlipidemia, peripheral artery disease, and type 2 diabetes presents for a 6-month follow-up. His medications include aspirin, metformin, lisinopril, and atorvastatin, all of which he takes without problems. Over the past several months, he has noticed that his erections are not adequate for sexual intercourse. He recently heard that a generic version of sildenafil has just become available, and he wonders if it might benefit him.

Erectile dysfunction is common, associated with chronic diseases

Erectile dysfunction, ie, persistent inability to obtain and maintain an erection sufficient to permit satisfactory sexual intercourse,^25,26 is estimated to affect nearly 20% of men over the age of 20 and 75% of men over the age of 75.²⁷

In age-adjusted models, erectile dysfunction has been shown²⁸ to be associated with:

• History of cardiovascular disease (odds ratio [OR] 1.63, 95% CI 1.02–2.63)
• Diabetes (OR 3.90, 95% CI 2.16–7.04)
• Treated hypertension vs no hypertension (OR 2.22, 95% CI 1.30–3.80)
• Current smoking vs never smoking (OR 1.63, 95% CI 1.01–2.62)
• BMI greater than 30 kg/m² vs less than 25 kg/m² (OR 1.80, 95% CI 1.03–3.14).

Because of the strong association between cardiovascular disease and erectile dysfunction, the presence of one often suggests the need to screen for the other.²⁹ While tools such as the International Index of Erectile Function (IIEF-5) have been developed to evaluate erectile dysfunction, it is most often diagnosed on the basis of clinical impression, while validated assessment methods are reserved for clinical trials.²⁸

Multiple causes of erectile dysfunction

Erectile dysfunction arises from inadequate penile tissue response to a sexual signal. The response can be disrupted at several points. For example, damage to vascular smooth muscle cells (eg, from age or obesity) and endothelial cells (from smoking or diabetes) and narrowing of the vascular lumen (from atherosclerosis or hypertension) have all been shown to impair engorgement of the corpus cavernosum.³⁰ In addition, denervation from prostate surgery or spinal trauma and psychogenic causes should be recognized in discussions with patients.

Drugs for erectile dysfunction

Pharmacologic management of erectile dysfunction includes oral, sublingual, intracavernosal, and intraurethral therapies.³¹ Treatment in primary care settings usually includes addressing underlying chronic diseases³² and prescribing phosphodiesterase-5 inhibitors (sildenafil, tadalafil, vardenafil, and avanafil). These drugs work by increasing local concentrations of cyclic guanosine monophosphate in the corpus cavernosum to induce vasodilation.³³

While these 4 drugs are still patent-protected, a manufacturer has been allowed to introduce a generic version of sildenafil into US markets, and a generic version of tadalafil is expected to be available soon.

Sildenafil, tadalafil, and vardenafil have been studied and found to have some degree of effectiveness in erectile dysfunction caused by damage to the penile vasculature, denervation, and spinal cord injury.³⁴ All drugs of this class have adverse effects including headache, facial flushing, and nasal congestion, but the drugs are generally well tolerated.³⁵

Sildenafil and tadalafil improve IIEF-5 scores by a similar margin, raising scores on the erectile domain subsection from approximately 14 of a possible 30 to approximately 24 of 30 in a trial of both drugs.³⁶ However, multiple crossover studies comparing the 2 drugs have shown that nearly 75% of patients prefer tadalafil to sildenafil,^36,37 perhaps because of tadalafil’s longer duration of action.³⁴

There is little evidence to suggest that vardenafil is more effective or more often preferred by patients than tadalafil or sidenafil.^34,38 And though data on the newest drug on the market, avanafil, are limited, a meta-analysis concluded that it may be less effective than tadalafil and without significant differences in terms of safety.³⁹

Other treatments

Lifestyle modifications, especially smoking cessation and exercise, have been shown to reduce the risk of erectile dysfunction with varying effect sizes across studies.^40–42 Moreover, factors such as obesity, alcohol use, and smoking may cause irreversible harm, and thus a healthy lifestyle should be encouraged.⁴¹

While there is only weak evidence for the use of psychological interventions alone for treating most types of erectile dysfunction, one meta-analysis found that the combination of psychological intervention and a phosphodiesterase-5 inhibitor improved sexual satisfaction more than drug therapy alone.⁴³

Erectile dysfunction: Take-home points

• Erectile dysfunction is common, affecting nearly 20% of men over the age of 20 and over 75% of men over the age of 75.
• Erectile dysfunction is often associated with chronic disease and may suggest the need to screen for cardiovascular disease.
• Treating underlying chronic diseases may help, and phosphodiesterase-5 inhibitors are effective; tadalafil may be most often preferred.

A 68-year-old man with a history of hypertension, BPH, and erectile dysfunction presents for a 6-month follow-up. His medication use includes lisinopril, which he takes without problems. He denies any new physical symptoms. His physical examination is unremarkable. He says he has heard about supplements that might help with his sexual performance and hopes to discuss recommendations during the visit.

A burgeoning, unregulated industry

Since the passage of the Dietary Supplement and Health Education Act in 1994, a law that decreased oversight of the supplement industry, spending on supplements has skyrocketed to over $41.1 billion each year.⁴⁴ Advertisements for these products typically claim that they improve general mental and physical health, sexual and romantic performance, leanness, and muscularity.⁴⁵ A national survey of men ages 57 and older reported that the most popular products were aimed at nutrition (such as multivitamins), cardiovascular health (such as omega-3 fatty acids), and chronic conditions (such as saw palmetto for BPH).⁴⁶

Little evidence of efficacy

There is little evidence to support the use of most supplements to improve men’s health. For example, a study in 82,405 men found no association between mortality rates and multivitamin use (hazard ratio [HR] 1.07, 95% CI 0.96–1.19).⁴⁷ Even for specific uses, such as cognitive performance, randomized trials exploring the effects of multivitamins in men have been largely negative.⁴⁸

The positive trials that have been reported are often of low quality and are funded by supplement manufacturers. For example, one of the few trials that reported a positive association between multivitamin supplementation and cognition in men was underpowered (N = 51) and found improvement in only 1 of 19 cognitive domains.⁴⁹ Despite the poor design and results to the contrary, this industry-funded study nevertheless concluded that multivitamins may play a role in improving elements of memory.

Evidence of possible harm from antioxidants

While not always specific to men, many meta-analyses have explored the effects of antioxidant supplements on cardiovascular and mortality risk. Most of them concluded that antioxidant supplements have no benefit and that some may actually be harmful.

For example, multiple meta-analyses of vitamin E supplementation found no cardiovascular benefit but possible increases in all-cause mortality rates in those taking high doses (risk ratio 1.04, 95% CI 1.01–1.07).^50,51

Another meta-analysis of 180,938 partici­pants in high-quality studies found an increased risk of all-cause mortality associated with independent intake of several antioxidant vitamins, including beta-carotene (risk ratio 1.07, 95% CI 1.02–1.11) and vitamin A (risk ratio 1.16, 95% CI 1.10–1.24), while intake of vitamin C and selenium had no impact on mortality.⁵²

Similarly, although nearly 10% of US adults report taking omega-3 fatty acid supplements, a review of 24 randomized controlled trials and meta-analyses published between 2005 and 2012 concluded that only 2 supported the use of these supplements for any health benefit.⁵³

Can supplements improve sexual function, prostate health?

To improve sexual function. A 2015 narrative review of the ingredients in General Nutrition Center’s top 30 best-selling products targeted at improving men’s sexual performance (including improving libido and erectile dysfunction) found only poor evidence for any efficacy.⁵⁴ The few studies that did support the use of select supplements, including B vitamins in people with diabetes, L-arginine, and yohimbine, were deemed to be of poor quality or showed a smaller effect size compared with standard medical therapy.

To prevent prostate cancer. Studies of supplement use to improve prostate health have had mixed results. For example, multiple large case-control studies have suggested that taking vitamin D^55,56 or vitamin C⁵⁷ is not associated with prostate cancer risk, while increased vitamin A^58,59 and E^60,61 intake is associated with inconsistent increases in prostate cancer risk.

In the Selenium and Vitamin E Cancer Prevention Trial,⁶² a randomized controlled trial in 35,533 men, those assigned to receive vitamin E supplementation were 17% more likely to get prostate cancer than were those assigned to placebo (HR 1.17, 99% CI 1.004–1.36, P = .008).

However, there are plausible biologic links between nutraceuticals and prostate cancer. For example, studies have linked genetic polymorphisms in vitamin D receptors⁶³ as well as intake of natural androgen receptor modulators, such as the most active polyphenol in green tea,⁶⁴ to prostate cancer risk and aggressiveness in certain populations. This led a recent review to conclude that there is some biologic plausibility, but at present little epidemiologic evidence, to support any dietary supplement’s ability to broadly affect prostate cancer risk.⁶⁵

Interest continues in exploring the targeted use of nutraceuticals as adjuvant therapy in specific populations at risk of prostate cancer.^66,67

To treat BPH. There is a similar dearth of clinical or population-based evidence that supplements can broadly affect BPH symptoms. For example, in a 2012 Cochrane review of Serenoa repens (saw palmetto) utilizing only high-quality evidence, there was no evidence that supplement use significantly reduced lower urinary tract symptoms, nocturia, or peak urine flow in BPH patients, and this was true even when the supplement was taken at triple-strength doses.⁶⁸

For other diseases. There is also limited evidence that supplements can affect other chronic diseases. For example, a meta-analysis of 3,803 patients found that glucosamine, chondroitin, and their combination had no impact on joint pain or joint space narrowing in patients with osteoarthritis of the knee or hip.⁶⁹

Even when there is some evidence to suggest benefit from supplementation, study heterogeneity and varying evidence quality limit confidence in the conclusions. For example, meta-analyses suggest garlic may improve blood pressure control in those with hypertension⁷⁰ and improve lipid and blood glucose control in type 2 diabetes.⁷¹ However, most of the trials included in those systematic reviews were underpowered, with samples as low as 10 patients, and many suffered from improper design, such as inadequate blinding of researchers. In addition, these meta-analyses often do not report adverse events, suggesting that higher quality studies would be needed to adequately measure event rates. As such, there is need for caution and a case-by-case review before recommending even a seemingly benign supplement like garlic to patients.

In total, there is only limited evidence to support the efficacy of supplements across many diseases and concerns common to men in primary care. This includes improving general health, cardiovascular health, sexual functioning, or other chronic diseases. While a supplement’s placebo effect may at times provide some benefit, supplements are much less strictly regulated since the passing of the 1994 act, and even vitamin supplementation has been shown to be associated with negative health outcomes. As such, a patient’s use of supplements requires careful consideration and shared decision-making.

Supplements: Take-home points

• Supplements are only loosely regulated by the federal government.
• There is some biologic but limited epidemiologic evidence for the use of multivitamins to improve cognition or mortality rates; for the use of antioxidant vitamins or omega-3 fatty acids to improve cardiovascular health; for the use of any of the top-selling sexual enhancement supplements to improve libido or erectile function; and for the use of vitamins or other supplements for improving BPH or reducing prostate cancer risk. Using supplements may in some cases be harmful.
• Given the heterogeneity of studies of supplements to manage chronic diseases and a lack of reporting of adverse events, careful consideration is needed when recommending supplements to patients.

Primary care physicians are tasked with a wide variety of issues affecting men. This article reviews the latest research in 4 areas of men’s health commonly addressed in primary care:

• Medical management of benign prostatic hyperplasia (BPH)
• Prostate cancer screening and treatment
• Medical management of erectile dysfunction
• Use of supplements.

See related commentary

MEDICAL MANAGEMENT OF BPH

An 84-year-old man with a history of hypertension, type 2 diabetes, hyperlipidemia, BPH, mild cognitive impairment, and osteoarthritis presents for a 6-month follow-up, accompanied by his son.

Two years ago he was started on a 5-alpha reductase inhibitor and an alpha-blocker for worsening BPH symptoms. His BPH symptoms are currently under control, with an American Urological Association (AUA) symptom index score of 7 of a possible 35 (higher scores being worse).

However, both the patient and son are concerned about the number of medications he is on and wonder if some could be eliminated.

Assessment tools

BPH is a common cause of lower urinary tract symptoms in older men. Evidence-based tools to help the clinician and patient decide on when to consider treatment for symptoms are:

• The AUA symptom index¹
• The International Prostate Symptom Score (IPSS).²

An AUA symptom index score or IPSS score of 8 through 19 of a possible 35 is consistent with moderate symptoms, while a score of 20 or higher indicates severe symptoms.

Combination therapy or monotherapy?

Monotherapy with an alpha-blocker or a 5-alpha reductase inhibitor is often the first-line treatment for BPH-related lower urinary tract symptoms.³ However, combination therapy with both an alpha-blocker and a 5-alpha reductase inhibitor is another evidence-based option.

The Medical Therapy of Prostatic Symptoms study,⁴ a randomized controlled trial, reported that long-term combination therapy reduced the risk of BPH clinical progression better than monotherapy. The same trial also found that either combination therapy or finasteride alone (a 5-alpha reductase inhibitor) reduced the risk of acute urinary retention and the future need for invasive therapy.

Monotherapy after a period of combination therapy?

There is also evidence to support switching from combination to monotherapy after an initial treatment period.

Matsukawa et al⁵ examined the effects of withdrawing the alpha-blocker from BPH combination therapy in a study in 140 patients. For 12 months, all patients received the alpha-blocker silodosin and the 5-alpha reductase inhibitor dutasteride. At 12 months, the remaining 132 patients (8 patients had been lost to follow-up) were randomized to continue combination therapy or to take dutasteride alone for another 12 months. They were evaluated at 0, 12, and 24 months by questionnaires (the IPSS and Overactive Bladder Symptom Score) and urodynamic testing (uroflowmetry, cystometrography, and pressure-flow studies).

There were no significant differen­ces in subjective symptoms and bladder outlet obstruction between patients who continued combination therapy and those who switched to dutasteride monotherapy. In the monotherapy group, those whose symptoms worsened weighed more (68.8 kg vs 62.6 kg, P =.002) and had a higher body mass index (BMI) (26.2 kg/m² vs 22.8 kg/m², P < .001) than those whose symptoms stayed the same or got better.

These findings of successful alpha-blocker withdrawal were consistent with those of other studies.

The Symptom Management After Reducing Therapy study⁶ showed that 80% of men with an IPSS score less than 20 who changed to dutasteride monotherapy did not have a noticeable worsening of their symptoms.

Baldwin et al⁷ noted similar success after withdrawing the alpha-blocker doxazosin in patients on finasteride.

Review all medications

The National Health and Nutrition Examination Survey noted that the estimated prevalence of polypharmacy increased from 8% in 1999 to 15% in 2011.⁸ Many commonly used medications, such as decongestants, antihistamines, and anticholinergic agents, can worsen BPH symptoms,⁹ so it is reasonable to consistently review the patient’s medications to weigh the risks and benefits and determine which ones align with the patient’s personal care goals.

BPH: Take-home points

• Combination therapy with an alpha-blocker and a 5-alpha reductase inhibitor is an effective regimen for BPH.
• Polypharmacy is a significant problem in the elderly.
• Withdrawing the alpha-blocker component from BPH combination therapy can be considered after 1 year of combination therapy in patients whose symptoms have been well controlled.

A 60-year-old patient calls you after receiving his laboratory testing report from his insurance physical. His prostate-specific antigen (PSA) level is 5.1 ng/mL, and he has several questions:

• Should he have agreed to the screening?
• How effective is the screening?
• What are the next steps?

Is PSA screening useful?

Over the last few years, there has been great debate as to the utility of screening for prostate cancer.

The US Centers for Disease Control and Prevention¹⁰ reported that in 2014, an estimated 172,258 men in the United States were diagnosed with prostate cancer, but only 28,343 men died of it. These statistics support the notion that screening programs may be detecting what might otherwise be a silent disease.

The US Preventive Services Task Force (USPSTF)¹¹ recommends against blanket PSA screening, in view of the low probability that it reduces the risk of death from prostate cancer. For men ages 55 through 69, current guidelines give a grade C recommendation to PSA screening, meaning there is moderate agreement that the benefit is likely small, and screening should be selectively offered based on professional judgment and patient preference. In men ages 70 and older who are not at high risk, the guideline gives screening a grade D recommendation, meaning there is moderate evidence that there is no benefit from the practice. This is a change from the 2012 USPSTF guidelines,¹² which gave a grade D recommendation to PSA screening for all ages.

The American Urological Association¹³ recommends against PSA screening in men under age 40 or ages 70 and older. It does not recommend routine screening in those ages 40 to 54 at average risk, but it says the decision should be individualized in this age group in those at higher risk (eg, with a positive family history, African American). At ages 55 through 69, it recommends shared decision-making, taking into account cancer risk and life expectancy. In those who opt for screening, an interval of 2 years or more may be preferred over annual screening to reduce the risk of overdiagnosis.

The USPSTF recommendations rely heavily on data from 2 trials: the European Randomized Study of Screening for Prostate Cancer (ERSPC)¹⁴ and the Prostate, Lung, Colorectal, and Ovarian Screening (PLCO) trial.¹⁵

The ERSPC¹⁴ demonstrated that screening for prostate cancer reduced deaths from prostate cancer by 20%, with an absolute risk difference of 0.71 deaths per 1,000 men; 1,410 men would need to be screened and 48 additional cases of prostate cancer would need to be treated to prevent 1 death from prostate cancer. Screening also decreased the risk of developing metastatic disease by 30%.¹⁶ On the negative side, screening increased the risk of overdiagnosis and other harms such as bleeding, sepsis, and incontinence.

The PLCO trial,¹⁵ in contrast, found no difference in death rates between men randomly assigned to annual screening and those assigned to usual care. Differences between the trial results were thought to be due to different practice settings as well as study implementation and compliance.

Tsodikov et al¹⁷ reanalyzed data from the ERSPC and the PLCO trial using 3 different mathematical models to estimate the effects of screening in both trials compared with no screening. The analysis found no evidence that the effects of screening vs not screening differed between the 2 trials, ultimately concluding that PSA screening reduced prostate cancer deaths by 25% to 32%, which the authors inferred was primarily a result of earlier detection of cancer.

The Cluster Randomized Trial of PSA Testing for Prostate Cancer,¹⁸ published in March 2018, explored the effect of single PSA screening vs no screening on prostate cancer mortality rates in 419,582 men ages 50 through 69. Although screening detected more cases of low-risk prostate cancer, there was no significant difference in prostate cancer mortality rates after a median follow-up of 10 years. However, 10% to 15% of the control group was estimated to have also been screened, and these results do not directly speak to the efficacy of serial PSA screening.

Extended follow-up of this trial is planned to report on long-term survival benefits and whether screening lowers the risk of metastasis.

Imaging-guided prostate biopsy

Once a patient is found to have an elevated PSA level, standard practice has been to perform transrectal ultrasonography to obtain 12 core biopsy samples. The results indicate whether the prostate contains cancer, how aggressive the cancer is (Gleason score), and whether there is extracapsular extension.

In the past, magnetic resonance imaging (MRI) of the prostate before biopsy was thought to be too costly, and many insurance plans do not currently cover it.

Pahwa et al,¹⁹ however, in a cost-effectiveness study using a decision-analysis model, found that using MRI to detect lesions and then guide biopsy by triaging patients into proper treatment pathways added health benefits in a cost-effective manner in 94.05% of simulations. These benefits were found across all age groups.

This study demonstrated that doctors could use MRI to better evaluate patients for potentially harmful lesions. If a focus of cancer is found, it can be biopsied; if no cancer is seen on MRI, the patient can avoid biopsy completely. Additionally, though MRI tended to miss low-risk cancers, these cancers are thought to disproportionately lead to higher healthcare costs through unnecessary treatment. Therefore, a negative MRI study was believed to be an excellent sign that the patient does not have aggressive prostate cancer. This approach led to a net gain of 0.251 additional quality-adjusted life years compared with the standard biopsy strategy.

The Prostate MRI Imaging Study²⁰ also found MRI to be effective in the prostate cancer workup. In this trial, 576 men who had never undergone biopsy underwent multiparametric MRI, transrectal ultrasonography-guided biopsy, and the reference standard, ie, transperineal template prostate mapping biopsy. Of those who underwent biopsy, 71% received a diagnosis of prostate cancer, and 40% had clinically significant disease. In patients with clinically significant disease, MRI was more sensitive than ultrasonography-guided biopsy (93% vs 48%, P < .0001) but less specific (41% vs 96%, P < .0001).

Based on these findings, if biopsy were performed only in those who had suspicious lesions on MRI, 27% of men with elevated PSA could avoid biopsy and its potential complications such as bleeding and sepsis, which occurred in 5.9% of the biopsy group.

The Prostate Evaluation for Clinically Important Disease: Sampling Using Image Guidance or Not? trial²¹ more recently studied MRI with or without targeted biopsy vs standard transrectal ultrasonography-guided biopsy in 500 men who had not undergone biopsy before, and reported similar results. MRI with or without biopsy led to fewer biopsies and less overdetection of clinically insignificant prostate cancers compared with the standard approach. Furthermore, those in the MRI-targeted biopsy group were 13% less likely to receive a diagnosis of clinically insignificant cancer than those who received the standard biopsy (adjusted difference −13 percentage points, 95% confidence interval [CI] −19 to −7, P < .001).

Together, these data provide another argument for adding multiparametric MRI to the workup of men with an elevated PSA level.

Surveillance vs treatment for prostate cancer

Once prostate cancer is diagnosed, surveillance is becoming an increasingly common management strategy.

The Prostate Cancer Intervention Versus Observation Trial (PIVOT),²² one of the largest and longest trials involving cancer patients, offered further evidence that active surveillance and less intervention for men with prostate cancer is a better approach in many cases. This trial compared prostatectomy and observation alone in a randomized fashion. Inclusion for the study required men to be medically fit for radical prostatectomy, along with having histologically confirmed localized prostate cancer (stage T1-T2NxM0 in the tumor-node-metastasis classification system) of any grade diagnosed within the last 12 months.

During 19.5 years of follow-up, 223 (61.3%) of the 364 men randomly assigned to radical prostatectomy died, compared with 245 (66.8%) of 367 men in the observation group; the difference was not statistically different (P = .06). Only 9.4% of the deaths were due to prostate cancer, 7.4% in the surgery group and 11.4% in the observation group (P = .06).

Surgery was associated with a lower all-cause mortality rate than observation in the subgroup of patients with intermediate-risk prostate cancer (defined as PSA 10–20 ng/mL and a Gleason score of 7). Surgery was also associated with less disease progression.²²

This finding is in line with previous data from the Scandinavian Prostate Cancer Group Study Number 4,²³ as well as the much larger Prostate Testing for Cancer and Treatment (ProtecT) trial,²⁴ both of which reported that metastasis was 1.5 and 2.6 times as common, respectively, in participants in the active surveillance groups. However, in the PIVOT trial, those in the surgery group were significantly more likely than those in the observation group to have erectile dysfunction and urinary incontinence at 10 years.

Therefore, in men with localized disease and in those with low-risk PSA levels, both the PIVOT and ProtecT trials suggest that death from prostate cancer is uncommon and that observation may be more appropriate.

Prostate cancer: Take-home points

• A new look at 2 large trials of PSA screening strengthened evidence that testing in the right patient population can reduce deaths from prostate cancer, but a third recently published trial that found no benefit from 1-time screening may reopen debate on the topic.
• MRI offers a better method than ultrasonography-guided biopsy to triage patients thought to be at high risk of prostate cancer and tends to limit costly overtreatment of disease that likely would not cause death.
• Surgery for prostate cancer may not prolong life but could reduce disease progression, at the risk of more adverse effects.
• Shared decision-making should be practiced when deciding whether to use active surveillance or active treatment of diagnosed prostate cancer.

A 62-year-old man with hypertension, hyperlipidemia, peripheral artery disease, and type 2 diabetes presents for a 6-month follow-up. His medications include aspirin, metformin, lisinopril, and atorvastatin, all of which he takes without problems. Over the past several months, he has noticed that his erections are not adequate for sexual intercourse. He recently heard that a generic version of sildenafil has just become available, and he wonders if it might benefit him.

Erectile dysfunction is common, associated with chronic diseases

Erectile dysfunction, ie, persistent inability to obtain and maintain an erection sufficient to permit satisfactory sexual intercourse,^25,26 is estimated to affect nearly 20% of men over the age of 20 and 75% of men over the age of 75.²⁷

In age-adjusted models, erectile dysfunction has been shown²⁸ to be associated with:

• History of cardiovascular disease (odds ratio [OR] 1.63, 95% CI 1.02–2.63)
• Diabetes (OR 3.90, 95% CI 2.16–7.04)
• Treated hypertension vs no hypertension (OR 2.22, 95% CI 1.30–3.80)
• Current smoking vs never smoking (OR 1.63, 95% CI 1.01–2.62)
• BMI greater than 30 kg/m² vs less than 25 kg/m² (OR 1.80, 95% CI 1.03–3.14).

Because of the strong association between cardiovascular disease and erectile dysfunction, the presence of one often suggests the need to screen for the other.²⁹ While tools such as the International Index of Erectile Function (IIEF-5) have been developed to evaluate erectile dysfunction, it is most often diagnosed on the basis of clinical impression, while validated assessment methods are reserved for clinical trials.²⁸

Multiple causes of erectile dysfunction

Erectile dysfunction arises from inadequate penile tissue response to a sexual signal. The response can be disrupted at several points. For example, damage to vascular smooth muscle cells (eg, from age or obesity) and endothelial cells (from smoking or diabetes) and narrowing of the vascular lumen (from atherosclerosis or hypertension) have all been shown to impair engorgement of the corpus cavernosum.³⁰ In addition, denervation from prostate surgery or spinal trauma and psychogenic causes should be recognized in discussions with patients.

Drugs for erectile dysfunction

Pharmacologic management of erectile dysfunction includes oral, sublingual, intracavernosal, and intraurethral therapies.³¹ Treatment in primary care settings usually includes addressing underlying chronic diseases³² and prescribing phosphodiesterase-5 inhibitors (sildenafil, tadalafil, vardenafil, and avanafil). These drugs work by increasing local concentrations of cyclic guanosine monophosphate in the corpus cavernosum to induce vasodilation.³³

While these 4 drugs are still patent-protected, a manufacturer has been allowed to introduce a generic version of sildenafil into US markets, and a generic version of tadalafil is expected to be available soon.

Sildenafil, tadalafil, and vardenafil have been studied and found to have some degree of effectiveness in erectile dysfunction caused by damage to the penile vasculature, denervation, and spinal cord injury.³⁴ All drugs of this class have adverse effects including headache, facial flushing, and nasal congestion, but the drugs are generally well tolerated.³⁵

Sildenafil and tadalafil improve IIEF-5 scores by a similar margin, raising scores on the erectile domain subsection from approximately 14 of a possible 30 to approximately 24 of 30 in a trial of both drugs.³⁶ However, multiple crossover studies comparing the 2 drugs have shown that nearly 75% of patients prefer tadalafil to sildenafil,^36,37 perhaps because of tadalafil’s longer duration of action.³⁴

There is little evidence to suggest that vardenafil is more effective or more often preferred by patients than tadalafil or sidenafil.^34,38 And though data on the newest drug on the market, avanafil, are limited, a meta-analysis concluded that it may be less effective than tadalafil and without significant differences in terms of safety.³⁹

Other treatments

Lifestyle modifications, especially smoking cessation and exercise, have been shown to reduce the risk of erectile dysfunction with varying effect sizes across studies.^40–42 Moreover, factors such as obesity, alcohol use, and smoking may cause irreversible harm, and thus a healthy lifestyle should be encouraged.⁴¹

While there is only weak evidence for the use of psychological interventions alone for treating most types of erectile dysfunction, one meta-analysis found that the combination of psychological intervention and a phosphodiesterase-5 inhibitor improved sexual satisfaction more than drug therapy alone.⁴³

Erectile dysfunction: Take-home points

• Erectile dysfunction is common, affecting nearly 20% of men over the age of 20 and over 75% of men over the age of 75.
• Erectile dysfunction is often associated with chronic disease and may suggest the need to screen for cardiovascular disease.
• Treating underlying chronic diseases may help, and phosphodiesterase-5 inhibitors are effective; tadalafil may be most often preferred.

A 68-year-old man with a history of hypertension, BPH, and erectile dysfunction presents for a 6-month follow-up. His medication use includes lisinopril, which he takes without problems. He denies any new physical symptoms. His physical examination is unremarkable. He says he has heard about supplements that might help with his sexual performance and hopes to discuss recommendations during the visit.

A burgeoning, unregulated industry

Since the passage of the Dietary Supplement and Health Education Act in 1994, a law that decreased oversight of the supplement industry, spending on supplements has skyrocketed to over $41.1 billion each year.⁴⁴ Advertisements for these products typically claim that they improve general mental and physical health, sexual and romantic performance, leanness, and muscularity.⁴⁵ A national survey of men ages 57 and older reported that the most popular products were aimed at nutrition (such as multivitamins), cardiovascular health (such as omega-3 fatty acids), and chronic conditions (such as saw palmetto for BPH).⁴⁶

Little evidence of efficacy

There is little evidence to support the use of most supplements to improve men’s health. For example, a study in 82,405 men found no association between mortality rates and multivitamin use (hazard ratio [HR] 1.07, 95% CI 0.96–1.19).⁴⁷ Even for specific uses, such as cognitive performance, randomized trials exploring the effects of multivitamins in men have been largely negative.⁴⁸

The positive trials that have been reported are often of low quality and are funded by supplement manufacturers. For example, one of the few trials that reported a positive association between multivitamin supplementation and cognition in men was underpowered (N = 51) and found improvement in only 1 of 19 cognitive domains.⁴⁹ Despite the poor design and results to the contrary, this industry-funded study nevertheless concluded that multivitamins may play a role in improving elements of memory.

Evidence of possible harm from antioxidants

While not always specific to men, many meta-analyses have explored the effects of antioxidant supplements on cardiovascular and mortality risk. Most of them concluded that antioxidant supplements have no benefit and that some may actually be harmful.

For example, multiple meta-analyses of vitamin E supplementation found no cardiovascular benefit but possible increases in all-cause mortality rates in those taking high doses (risk ratio 1.04, 95% CI 1.01–1.07).^50,51

Another meta-analysis of 180,938 partici­pants in high-quality studies found an increased risk of all-cause mortality associated with independent intake of several antioxidant vitamins, including beta-carotene (risk ratio 1.07, 95% CI 1.02–1.11) and vitamin A (risk ratio 1.16, 95% CI 1.10–1.24), while intake of vitamin C and selenium had no impact on mortality.⁵²

Similarly, although nearly 10% of US adults report taking omega-3 fatty acid supplements, a review of 24 randomized controlled trials and meta-analyses published between 2005 and 2012 concluded that only 2 supported the use of these supplements for any health benefit.⁵³

Can supplements improve sexual function, prostate health?

To improve sexual function. A 2015 narrative review of the ingredients in General Nutrition Center’s top 30 best-selling products targeted at improving men’s sexual performance (including improving libido and erectile dysfunction) found only poor evidence for any efficacy.⁵⁴ The few studies that did support the use of select supplements, including B vitamins in people with diabetes, L-arginine, and yohimbine, were deemed to be of poor quality or showed a smaller effect size compared with standard medical therapy.

To prevent prostate cancer. Studies of supplement use to improve prostate health have had mixed results. For example, multiple large case-control studies have suggested that taking vitamin D^55,56 or vitamin C⁵⁷ is not associated with prostate cancer risk, while increased vitamin A^58,59 and E^60,61 intake is associated with inconsistent increases in prostate cancer risk.

In the Selenium and Vitamin E Cancer Prevention Trial,⁶² a randomized controlled trial in 35,533 men, those assigned to receive vitamin E supplementation were 17% more likely to get prostate cancer than were those assigned to placebo (HR 1.17, 99% CI 1.004–1.36, P = .008).

However, there are plausible biologic links between nutraceuticals and prostate cancer. For example, studies have linked genetic polymorphisms in vitamin D receptors⁶³ as well as intake of natural androgen receptor modulators, such as the most active polyphenol in green tea,⁶⁴ to prostate cancer risk and aggressiveness in certain populations. This led a recent review to conclude that there is some biologic plausibility, but at present little epidemiologic evidence, to support any dietary supplement’s ability to broadly affect prostate cancer risk.⁶⁵

Interest continues in exploring the targeted use of nutraceuticals as adjuvant therapy in specific populations at risk of prostate cancer.^66,67

To treat BPH. There is a similar dearth of clinical or population-based evidence that supplements can broadly affect BPH symptoms. For example, in a 2012 Cochrane review of Serenoa repens (saw palmetto) utilizing only high-quality evidence, there was no evidence that supplement use significantly reduced lower urinary tract symptoms, nocturia, or peak urine flow in BPH patients, and this was true even when the supplement was taken at triple-strength doses.⁶⁸

For other diseases. There is also limited evidence that supplements can affect other chronic diseases. For example, a meta-analysis of 3,803 patients found that glucosamine, chondroitin, and their combination had no impact on joint pain or joint space narrowing in patients with osteoarthritis of the knee or hip.⁶⁹

Even when there is some evidence to suggest benefit from supplementation, study heterogeneity and varying evidence quality limit confidence in the conclusions. For example, meta-analyses suggest garlic may improve blood pressure control in those with hypertension⁷⁰ and improve lipid and blood glucose control in type 2 diabetes.⁷¹ However, most of the trials included in those systematic reviews were underpowered, with samples as low as 10 patients, and many suffered from improper design, such as inadequate blinding of researchers. In addition, these meta-analyses often do not report adverse events, suggesting that higher quality studies would be needed to adequately measure event rates. As such, there is need for caution and a case-by-case review before recommending even a seemingly benign supplement like garlic to patients.

In total, there is only limited evidence to support the efficacy of supplements across many diseases and concerns common to men in primary care. This includes improving general health, cardiovascular health, sexual functioning, or other chronic diseases. While a supplement’s placebo effect may at times provide some benefit, supplements are much less strictly regulated since the passing of the 1994 act, and even vitamin supplementation has been shown to be associated with negative health outcomes. As such, a patient’s use of supplements requires careful consideration and shared decision-making.

Supplements: Take-home points

• Supplements are only loosely regulated by the federal government.
• There is some biologic but limited epidemiologic evidence for the use of multivitamins to improve cognition or mortality rates; for the use of antioxidant vitamins or omega-3 fatty acids to improve cardiovascular health; for the use of any of the top-selling sexual enhancement supplements to improve libido or erectile function; and for the use of vitamins or other supplements for improving BPH or reducing prostate cancer risk. Using supplements may in some cases be harmful.
• Given the heterogeneity of studies of supplements to manage chronic diseases and a lack of reporting of adverse events, careful consideration is needed when recommending supplements to patients.

Bisphosphonate therapy minimizes bone loss and reduces fracture risk by up to 50% in patients with osteoporosis,¹ but it is also associated with increased risks of osteonecrosis of the jaw and atypical femoral fracture. Although atypical femoral fractures are rare, they can have a devastating effect. Patient concern about this complication has contributed to a decrease in bisphosphonate use by about half in the last decade or so,^2,3 and we fear this could result in an increase in hip fracture rates.

In this article, we examine the evidence on bisphosphonate-associated atypical femoral fractures, including risks, pathogenesis, treatment, and prevention.

ATYPICAL FRACTURES INVOLVE THE FEMORAL SHAFT, NOT THE HEAD

An atypical femoral fracture is a transverse fracture of the femoral shaft (diaphysis), defined by both clinical criteria and radiographic appearance.

To be defined as atypical, a femoral fracture must meet 4 of the following 5 criteria⁴:

• Occurs with minimal or no trauma
• Has a predominantly transverse fracture line, originating at the lateral cortex and sometimes becoming oblique as it progresses medially across the femur
• Extends through both cortices and may be associated with a medial spike (complete fractures); or involves only the lateral cortex (incomplete fractures)
• Is noncomminuted or minimally comminuted
• Shows localized periosteal or endosteal thickening (termed “beaking” or “flaring”) of the lateral cortex at the fracture site.

Several minor features are also important but are not required, eg:

• Cortical thickening of the femoral shaft
• Unilateral or bilateral prodromal pain preceding the fracture
• Bilateral incomplete or complete femoral diaphysis fractures
• Delayed fracture healing.

Atypical femoral fracture can occur anywhere along the shaft, from just distal to the lesser trochanter to just proximal to the supracondylar flare. However, most occur in 2 areas, with 1 cluster centered at about 41 mm from the lesser trochanter (more common in relatively younger patients) and the other at 187 mm.⁵

ABSOLUTE RISK IS LOW BUT INCREASES WITH LONGER USE

Atypical femoral fractures are rare. Schilcher et al⁶ reviewed radiographs of 1,234 women who had a subtrochanteric or shaft fracture and found 59 (4.6%) of fractures were atypical. In a systematic review of 14 studies,⁷ the incidence ranged from 3.0 to 9.8 cases per 100,000 patient-years.

Furthermore, not all atypical femoral fractures are in bisphosphonate users: 7.4% were in nonusers in 1 series⁸ and 22% in another.⁹

Nevertheless, most studies show that bis­phosphonate use increases the incidence of atypical femoral fracture, and the incidence increases with duration of use, especially after 3 years.⁷

An international task force of the American Society for Bone and Mineral Research listed the absolute risk as between 3.2 and 50 cases per 100,000 patient-years, with longer use (> 5 years) increasing the risk to about 100 per 100,000 patient-years.⁴ After stopping bis­phosphonate therapy, the risk diminished by 70% per year.⁹

In another study, for 0.1 to 1.9 years of therapy, the age-adjusted atypical fracture rates were 1.78 per 100,000 per year (95% confidence interval [CI] 1.5–2.0), increasing to 113.1 per 100,000 per year (95% CI 69.3–156.8) with exposure from 8 to 9.9 years.¹⁰

A case-control study found that more than 5 years of bisphosphonate use increased the fracture risk by an odds ratio of 2.74 (95% CI 1.25–6.02).¹¹

The incidence of typical femoral fracture was higher in those who adhered better to their oral bisphosphonate regimen in some studies,¹² but the opposite was true in others.¹³

The benefits of bisphosphonate therapy in reducing fracture risk, however, outweigh the risk of atypical fracture.⁴

We do not know whether the rate of atypical femoral fracture is increasing. A review of Kaiser Permanente Northwest records found that the rates of atypical femoral shaft fracture had remained stable from 1996 to 2009. However, 61.9% of patients who met the strict radiographic criteria had taken oral bisphosphonates.¹⁴ These data suggest that bisphosphonate use has not increased the overall population-based risk for subtrochanteric and femoral shaft fractures, but that bisphosphonates and other risk factors may have increased the likelihood that such fractures will exhibit atypical radiographic features.

A population-based study in Denmark¹³ found that alendronate use longer than 10 years was associated with an adjusted 30% lower risk of hip fracture and no increase in the risk of subtrochanteric and femoral shaft fracture. In addition, the risk of subtrochanteric and femoral shaft fracture was lower with high adherence to alendronate treatment (based on medication possession ratio > 80%) compared with low adherence (ratio < 50%) (odds ratio 0.88, 95% CI 0.77–0.99). The risk was not increased in current vs past users.

The Danish study¹³ used the coding of the 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10) to identify subtrochanteric and femoral shaft fractures without radiologic review for atypical radiographic features. The lack of specific ICD-10 coding for subtrochanteric and femoral shaft fractures with atypical radiographic features has limited our knowledge of their incidence.

After an atypical femoral fracture, patients have a significant risk of fracture on the contralateral side. In a case-control study, 28% of patients with atypical femoral fracture suffered a contralateral fracture, compared with 0.9% of patients presenting with a typical fracture pattern (odds ratio 42.6, 95% CI 12.8–142.4).¹⁵

Contralateral fracture occurs from 1 month to 4 years after the index atypical femoral fracture.¹⁶

There are reports of bisphosphonate-related low-impact fractures in other sites such as the tibia¹⁷ and forearm.¹⁸ However, they may be too rare to warrant screening.

Mortality rates

A Swedish database study found that patients with atypical femoral fractures, whether bisphosphonate users or nonusers, do not have higher mortality rates than patients with ordinary subtrochanteric or femoral shaft fractures.¹⁹ Furthermore, the mortality rates for those with atypical femoral fracture were similar to rates in the general population. In contrast, patients with an ordinary femoral fracture had a higher mortality risk than the general population.¹⁹

Other studies suggest that atypical femoral fracture may be associated with a less favorable prognosis in older patients,²⁰ but this could be due to differences in demographics, treatment adherence, or postfracture care.²¹

In addition, functional outcomes as measured by independent mobility at discharge and at 3 months were comparable between patients with atypical fracture and those with typical fracture.²²

IMAGING STUDIES

If a long-term bisphosphonate user presents with hip, thigh, or groin pain, imaging studies are recommended.

Plain radiography

Radiography is usually the first step and should include a frontal view of the pelvis (Figure 1) and 2 views of the full length of each femur. If radiography is not conclusive, bone scan or magnetic resonance imaging (MRI) should be considered.

A linear cortex transverse fracture pattern and focal lateral cortical thickening are the most sensitive and specific radiographic features.^23,24 Because of the risk of fracture on the contralateral side, radiographic study of that side is recommended as well.

Computed tomography

Computed tomography (CT) is not sensitive for early stress fractures and, given the radiation burden, is not recommended in the workup of atypical fracture.

Bone scanning

Bone scanning using technetium 99m-labeled methylene diphosphonate with a gamma camera shows active bone turnover. Stress fractures and atypical femoral fractures are most easily identified in the third (delayed) phase of the bone scan. Although bone scanning is highly sensitive, the specificity is limited by lack of spatial resolution. Atypical femoral fracture appears as increased activity in the subtrochanteric region with a predilection for the lateral cortex.

Dual-energy x-ray absorptiometry

Conventional dual-energy x-ray absorptiometry (DXA) extends only to 1 to 2 cm below the lesser trochanter and can therefore miss atypical fractures, which usually occur farther down. The overall detection rate for DXA was 61% in a sample of 33 patients.²⁵

Newer scanners can look at the entire femoral shaft.²⁶ In addition, newer software can quantify focal thickening (beaking) of the lateral cortex and screen patients who have no symptoms. The results of serial measurements can be graphed so that the practitioner can view trends to help assess or rule out potential asymptomatic atypical femoral fracture.

A localized reaction (periosteal thickening of the lateral cortex or beaking) often precedes atypical femoral fracture. A 2017 study reported that patients with high localized reaction (mean height 3.3 mm) that was of the pointed type and was accompanied by prodromal pain had an increased risk of complete or incomplete atypical femoral fracture at that site.²⁷ This finding is used by the newer DXA software. The predictive value of beaking on extended femoral DXA may be as high as 83%.²⁶

Magnetic resonance imaging

The MRI characteristics of atypical femoral fracture are similar to those of other stress fractures except that there is a lateral-to-medial pattern rather than a medial pattern. The earliest findings include periosteal reaction about the lateral cortex with a normal marrow signal.

MRI may be of particular benefit in patients with known atypical femoral fracture to screen the contralateral leg. It should image the entire length of both femurs. Contrast enhancement is not needed.

Regardless of whether initial findings were discovered on conventional radiographs or DXA, MRI confirmation is needed. Radio­nuclide bone scanning is currently not recommended because it lacks specificity. Combination imaging is recommended, with either radiography plus MRI or DXA plus MRI.

DIFFERENTIAL DIAGNOSIS

The differential diagnosis of atypical femoral fracture includes stress fracture, pathologic fracture, hypophosphatasia, and osteogenesis imperfecta.²⁸ Hypophosphatemic osteomalacia can cause Looser zones, which can be confused with atypical femoral fractures but usually occur on the medial side.⁴ Stress fracture of the femur can occur below the lesser trochanter but usually begins in the medial, not the lateral, cortex.

Pathologic fractures from underlying osseous lesions can mimic the cortical beaking of bisphosphonate-related fracture, but they usually show the associated underlying lucent lesion and poorly defined margins. A sinus tract along the region of a chronic osteomyelitis may also appear similar.

Hypophosphatasia is an inborn error of metabolism caused by a loss-of-function mutation in the gene encoding alkaline phosphatase, resulting in pyrophosphate accumulation and causing osteomalacia from impaired mineralization. This can result in femoral pseudofracture that is often bilateral and occurs in the subtrochanteric region.²⁹

Patients with atypical femoral fracture are generally a heterogeneous group, but there are risk factors to note other than bisphosphonate exposure.

Asian women had a risk 8 times higher than white women in 1 study.³⁰

Bone geometry. Mahjoub et al⁸ reported that compared with controls, patients with atypical femoral fracture had greater offset of the femoral shaft from the center of rotation of the femoral head, a more acute angle between the femoral neck and shaft, and greater proximal cortical thickness.

Medications. In addition to bisphosphonates, other drugs associated with atypical femoral fracture include RANK-ligand inhibitors such as denosumab (another drug for osteoporosis),³¹ glucocorticoids,^32,33 and proton pump inhibitors.^32,33

Genetics. Three sisters with atypical femoral fracture were found to have 37 rare mutations in 34 genes, including one in the GGPS1 gene, which codes for geranylgeranyl pyrophosphate synthase—an enzyme that bisphosphonates inhibit.³⁴

Medical conditions other than osteoporosis include collagen diseases, chronic pulmonary disease, asthma, rheumatoid arthritis, and diabetes.³⁵

Clinical recommendations

Current recommendations are to reevaluate bisphosphonate use in patients with osteoporosis after 5 or more years of therapy.³⁶

Given that patients with osteoporosis are at increased risk of typical fracture, those at higher risk should be considered for continued bisphosphonate therapy. Factors for high risk include the following:

• History of fracture on therapy
• Hip T score –2.5 or lower
• Older age (≥ 70)
• Other strong risk factors for fracture such as smoking, alcohol use, corticosteroid use, rheumatoid arthritis, and family history
• World Health Organization FRAX fracture risk score above the country-specific threshold.

Those at lower risk should be considered for a 2- to 3-year bisphosphonate holiday with periodic reevaluation of bone density and, possibly, bone markers.³⁶

WHAT IS THE UNDERLYING PATHOPHYSIOLOGY?

The mechanism by which bisphosphonates increase the risk of atypical femoral fracture is not clear. These drugs work by suppressing bone turnover; however, in theory, prolonged use could suppress it too much and increase bone fragility.

One hypothesis is that bisphosphonates impair the toughening of cortical bone, an important barrier to clinical fracture. This is supported by a study that found bisphosphonate users with atypical femoral fracture had deficits in intrinsic and extrinsic bone toughness, perhaps due to treatment-related increases in matrix mineralization.³⁷ Although this study and others showed an increase in matrix mineralization and reduced mineralization heterogeneity with bisphosphonate use,^38,39 it is unclear whether such changes contributed to reduced toughness or to atypical femoral fracture.

Changes in the skeletal geometry of the lower limb such as femoral neck-shaft angle and femoral curvature alter the stresses and strains experienced by the femoral diaphysis with loading. Because the incidence of incomplete atypical femoral fracture is much greater than that of complete fracture, most incomplete atypical femoral fractures heal before the fracture progresses.

Ultimately, all fractures, including atypical femoral fractures, occur when mechanical stress and strain exceed bone strength.

Antiresorptive drugs such as bisphosphonates, estrogen, calcitonin, and RANK ligand inhibitors prevent hip fracture by increasing the strength of the proximal femur—perhaps at the expense of the strength (or toughness) of the subtrochanteric shaft. It is also possible that treatment-related increases in hip strength (and reduced hip fracture rates) promote or sustain the transfer of stress and strain to femoral regions that experience lesser or no increases in strength from treatment, which likely includes the shaft.^40,41

CT studies in Japanese women with osteoporosis have shown that 2 years of zoledronate therapy had greater effects in the hip than in the femoral shaft, with significant increases in cortical thickness and volumetric bone mineral density at the femoral neck and intertrochanteric region compared with baseline.⁴² But zoledronate did not increase femoral shaft cortical thickness and caused only a minor increase in femoral shaft volumetric bone mineral density. Fracture patterns may have depended on damage and effects of bone turnover on mass and structure.

This hypothetical scenario portrays a possible “hip survival bias” mechanism for atypical femoral fracture, with the association with antiresorptive drugs arising from greater stress and strain in cortical regions where these fractures occur rather than from treatment-related reductions in cortical bone strength or toughness.

PRODROMAL PAIN IS COMMON

From 32% to 76% of patients who have incomplete or developing atypical femoral fracture present with a prodrome of groin or hip pain.^4,43 Prodromal pain occurs any time from 2 weeks to several years before the fracture, presenting as pain in the anterior or lateral thigh or in the groin.

Prodromal pain in a patient on antiresorptive therapy should be a signal for the clinician to obtain a radiograph of the hip and to look for contralateral symptoms and fractures. The most common mechanism of injury appears to be a ground-level fall or even a nontraumatic activity such as walking or stepping off a curb.

MEDICAL MANAGEMENT

In bisphosphonate users with radiographic evidence of atypical femoral fracture, the bis­phosphonate should be discontinued and the patient assessed for calcium and vitamin D deficiency, with supplements prescribed if needed.⁴

For patients with incomplete fracture and persistent pain after 3 months of medical management, prophylactic surgical nail fixation is recommended to prevent complete fracture.

Teriparatide, which has been associated with enhanced bone fracture healing, is a possible treatment to promote healing of atypical femoral fracture, either alone or as an adjunct to surgical fixation. A systematic review published in 2015 supported the use of teriparatide for enhancing fracture healing in atypical femoral fracture.⁴⁴ In addition, a 10-patient series⁴⁵ showed that incomplete fractures without radiolucent lines responded to teriparatide alone, whereas those with radiolucent lines needed intramedullary nailing.

These results suggest that teriparatide works best when the fracture site is stable, either inherently or with surgical fixation.

ORTHOPEDIC CARE

Orthopedic care for atypical femoral fracture differs depending on whether the patient experiences pain and whether the fracture is incomplete or complete. Figure 2 shows a treatment algorithm for atypical femoral fracture.

These are difficult fractures to manage, complicated by delayed healing in the elderly, complex displacement patterns, altered bone geometry, and risk of fracture in the opposite limb, all of which raise questions about recommending protected weight-bearing exercise.

Furthermore, atypical femoral fracture is often associated with increased anterolateral bowing of the femur, making it difficult to insert an intramedullary nail: the radius of curvature of the bone is shorter than that of a standard femoral nail. This mismatch can lead to intraoperative complications such as iatrogenic fracture during prophylactic nailing, malunion from excess straightening of the femur (which can itself lead to leg length discrepancy), and gapping of the fracture site, particularly on the medial side.

Intramedullary nailing is the first-line treatment for complete atypical femoral fracture, although the risk of delayed healing and revision surgery may be somewhat higher than with typical femoral fracture.⁴⁶ Prophylactic intramedullary nailing should be considered for a patient with intractable pain.²

A radiograph of the opposite leg should be obtained routinely, looking for an asymptomatic fracture. Bisphosphonates should be discontinued and calcium and vitamin D continued. Teriparatide therapy can be considered as an alternative treatment.

Conservative management for incomplete fracture without pain

Incomplete atypical femoral fracture unaccompanied by pain can be followed conservatively.⁴⁷ In addition to stopping antiresorptive therapy, patients need to avoid high-impact and repetitive-impact activities such as jumping or running. If pain occurs, patients should begin protected weight-bearing exercise.

Treatment is uncertain for incomplete fracture with pain

For patients with incomplete atypical femoral fracture and pain, treatment is controversial. Regimens that include 2 to 3 months of protected weight-bearing exercise, a full metabolic bone workup, calcium and vitamin D supplementation, and anabolic bone agents have produced some success. Some authors have reported poor results from conservative care, with few patients achieving pain relief or signs of complete healing.^48,49 Additionally, if an incomplete fracture is found in the opposite femur, protected weight-bearing of both legs may not be possible.

Patients with incomplete fracture should be monitored regularly with radiography and physical examination. If there is progression of the fracture, escalation of pain, or failure to heal within 2 to 3 months, then surgical treatment is necessary.

Prophylactic placement of an intramedullary nail to prevent completion of the fracture and allow a return to full weight-bearing is generally advised.⁵⁰ A long locking plate can be used if bone deformities make it difficult to place an intramedullary nail; however, nails are preferred because they allow formation of endochondral callus, which can be helpful in these difficult-to-heal fractures.

Results from retrospective reviews have shown that surgically treated patients with bis­phosphonate-associated incomplete atypical femoral fracture were more likely than those treated nonsurgically to be pain-free (81% vs 64%) and have radiographic healing (100% vs 18% at final follow-up).⁴⁶ Results have also been positive for those with complete atypical femoral fracture. At 6 months, 64% of surgically treated patients were pain-free and 98% were radiographically healed.⁵¹

The unusual geometry of the femur in patients with atypical femoral fracture and the presence of intramedullary cortical callus makes the placement of an intramedullary femoral rod more complex than in typical femoral fracture.⁸

Intramedullary nailing of atypical femoral fracture is a challenge for even the most experienced surgeon, and vigilance is imperative to avoid iatrogenic fracture and malunion.

MANY QUESTIONS REMAIN

We need more studies on the pathophysiology of bisphosphonate-associated atypical femoral fracture, the value of periodic screening with DXA, and which factors predict high risk (eg, Asian ethnicity, use of certain medications, femoral geometry). In addition, we need more data on the success of conservative management of incomplete fracture, including use of teriparatide.