Over the last decade, several studies have demonstrated that deep brain stimulation (DBS) is among the most effective approaches for the treatment of patients with advanced movement disorders, including chorea, levodopa-induced dyskinesia, tremor, and dystonia.¹ The goal of DBS is to restore function or relieve pain by stimulating neuronal activity through surgically implanted electrodes. DBS produces marked and persistent reductions in abnormal movements in patients with common hyperkinetic disorders, with a generally low incidence of serious adverse events in pediatric patients and adults.

DEEP BRAIN STIMULATION FOR ESSENTIAL TREMOR

Improvement with thalamic DBS

The ventral intermediate nucleus (VIM) of the thalamus is the most common target for DBS treatment of essential tremor. Several studies have demonstrated significant long-term improvement in tremor following thalamic DBS.³ Most studies enrolled 20 to 30 patients, who were followed for 1 to 5 years after device implantation. On average, these studies reported an improvement in overall tremor of approximately 50% from baseline with thalamic DBS.

Patient selection and stimulation parameters

Symptoms targeted for DBS treatment include unilateral and sometimes bilateral limb tremor. Some evidence exists for effectiveness in axial and vocal tremor as well. Factors to consider in patient selection for DBS surgery include tremor severity, degree of refractoriness to medication, and type of tremor. In addition, individual patient characteristics should be considered, including age, comorbid conditions, surgical risk, patient preference, social and employment factors, and social support.

Reprinted with permission from Journal of Clinical Neurophysiology (Cooper SE, et al. A model predicting optimal parameters for deep brain stimulation in essential tremor. J Clin Neurophysiol 2008; 2:26–273). Copyright © 2008 by the ACNS.

DEEP BRAIN STIMULATION FOR DYSTONIA

Dystonia is characterized by involuntary twisting muscle contractions causing abnormal postures sometimes accompanied by jerky or repetitive involuntary movements. It may be classified according to the body part affected as generalized, segmental, or focal; in some cases it may be classified as multifocal dystonia or hemidystonia. Dystonia is also classified as primary or secondary, according to etiology. Primary dystonias are those not caused by any other identifiable condition and not associated with other neurologic abnormalities. These include idiopathic and some genetic dystonias, such as the DYT1 torsinA gene mutation. DBS of the globus pallidus internus (GPi) or subthalamic nucleus (STN) was approved by the US Food and Drug Administration under a humanitarian device exemption in 2003 for the treatment of primary generalized dystonia (PGD) in patients aged 7 years and older; GPi is the more common target).¹

Evidence of efficacy

Several clinical studies have demonstrated the efficacy of DBS for patients with disabling PGD that is unresponsive to pharmacotherapy.

Long-term efficacy. Isaias and colleagues examined long-term safety and efficacy of DBS in 30 consecutive patients with PGD who were followed for at least 3 years after pallidal DBS surgery.⁶ DBS was delivered bilaterally in 28 patients and unilaterally in 2 patients. Clinical rating scales of motor function improved by a mean of 82.5% after 2 years, and dystonia-related disability improved by a mean of 75.2%. Improvement in motor function from baseline was noted for all 30 subjects. In five patients who were followed for 7 years, improvement in motor function remained greater than 80% at the last follow-up visit. Transient regressions were noted for patients with hardware failures or whose batteries had reached the end of life. Stimulation-related adverse events were reported for three patients and included speech difficulties and, in one patient, transient blepharospasm.

Vidailhet and colleagues examined the efficacy of bilateral pallidal stimulation in 22 patients with PGD who were followed prospectively for 3 years.⁷ Mean improvement from baseline in motor function on a dystonia rating scale was 51% after 1 year and 58% after 3 years (P = .03). Significant improvement was noted for individual ratings of upper and lower limb function scores. Health-related quality of life was also significantly improved at 3-year follow-up (P = .05). Serious adverse events were reported for three patients, including two lead fractures and one infection.

Results from double-blind trial. Kupsch and colleagues performed a randomized, double-blind clinical trial comparing pallidal DBS versus device implantation and sham stimulation in 40 patients with primary segmental or generalized dystonia.⁸ After 3 months, the mean change from baseline in severity of dystonia was 15.8% for patients who received DBS versus 1.4% with sham stimulation (P < .001). At the conclusion of the double-blind treatment phase, patients entered an open-label extension phase in which all patients received DBS for another 3 months. The initial benefit of treatment was sustained across the entire 6-month study period for patients initially randomized to DBS, whereas patients who were initially randomized to sham stimulation exhibited improved motor function during the open-label extension phase. Ratings of disability and quality of life also improved for patients receiving DBS at the end of the 6-month study. Adverse events included dysarthria (five patients), serious infections (four patients), and lead dislodgement (one patient).

Response with DYT1mutation. Coubes and colleagues examined the long-term efficacy and safety of bilateral DBS in 31 children and adults with PGD.⁹ PGD is associated with autosomal DYT1 mutations in approximately 30% of cases, and these authors examined the effects of treatment in patients with and without the DYT1 mutation. After 2 years of treatment, mean scores on a dystonia clinical rating scale decreased by 79% from baseline, and mean disability ratings decreased by 65%. The improvement in clinical dystonia rating scale scores was significantly greater for children than adults after 2 years (84.7% vs 70.1%; P = .04). In children, functional improvement was greater after 2 years in the subset of patients with DYT1 mutations than in the subset of patients without (76.1% vs 44.5%; P = .03), whereas in adults, DYT1 mutation status did not significantly influence response to treatment. One case of unilateral infection was noted, which required removal of the implant with successful reimplantation 6 months later. No other adverse events were reported.

Appropriate patients for DBS include those with an unequivocal diagnosis of dystonia and significant disability. Etiology and type of dystonia should also be considered. Patients with secondary dystonia (eg, due to structural brain lesions or heredodegenerative disorders) generally do not respond to DBS as well as patients with primary dystonias. A possible exception is tardive dystonia, which is caused by past exposure to dopamine receptor–blocking drugs. Although it is a secondary dystonia, tardive dystonia may respond well to DBS. Data on this point remain limited. Moreover, with tardive dystonia (as well as Sydenham chorea and poststroke hemiballismus), there may be spontaneous remission. DBS in these conditions should therefore be considered when enough time has elapsed that the likelihood of spontaneous remission is low.¹

Not all dystonic symptoms have been shown to respond equally to DBS. Evidence of effectiveness is stronger and more consistent for limb and axial dystonia than for dystonic impairment of speech and swallowing. Phasic dystonia (jerky or rhythmic movements) appears to respond better than fixed postures. A critical point is that fixed postures not caused by electrically active muscle contraction will not respond to DBS. For example, bony deformity of the spine, joint disease, or tendon shortening cannot be expected to improve with DBS. The situation is complicated, since such conditions may develop as secondary consequences of dystonia. The potential for their development may warrant earlier rather than later DBS surgery in childhood-onset PGD.¹⁰

UNRESOLVED ISSUES IN DBS FOR DYSTONIA

How aggressively should other therapies be tried before starting DBS?

Pharmacologic options include a range of oral, intramuscular, and intrathecal agents. Injection of botulinum toxin to denervate affected muscles is a mainstay of treatment for focal or segmental dystonia, but often fails to improve symptoms because of the involvement of a large number of muscles, complexity of the movement pattern, or the development of neutralizing antibodies.⁸ With the exception of levodopa-responsive PGD, other pharmacologic therapy for PGD is generally of limited effectiveness for controlling symptoms of dystonia.⁹ Oral or intrathecal baclofen may improve symptoms, but often produces unacceptable sedation.

How important is intraoperative microelectrophysiology?

Although contemporary imaging techniques are important in the correct placement of stimulating electrodes, the available techniques do not always provide sufficient resolution to delineate the STN or GPi. The accuracy of electrode placement may also be influenced by distortions caused by lack of homogeneity among magnetic resonance images, brain shift, and signal deflections from cannulae or electrodes.¹⁴ These errors may result in significant deviation of electrode placement from the intended target. Microelectrode recording and micro-stimulation may be used to map the target region and refine the surgical target. It is widely, but not universally, held that this strategy contributes to superior accuracy and outcomes; it ordinarily requires an awake procedure, which is not always feasible in patients with severe dystonia or in pediatric patients.¹¹

How should be programming (stimulator adjustment) be performed?

Research continues to refine our understanding of how electrical parameters such as voltage, frequency, and pulse width affect clinical outcomes in patients undergoing DBS for dystonia. Some programming approaches, such as long pulse width and high frequency, that were once generally accepted are now widely questioned. Another major unresolved question is: “How long should it take to see the results of stimulation?” In the clinical studies described above, continued improvement was generally observed over months or even years, and, in most patients, stimulators are incrementally adjusted over an extended period. However, some patients may experience much more rapid onset of benefit.

Long-term DBS management of dystonia

Unlike DBS for Parkinson disease or even essential tremor, DBS for dystonia is performed in young patients. This creates special challenges in pediatric patients, who can be expected to grow and develop after device implantation. As a result, children may require additional surgeries to reposition devices.

In addition, the most widely used devices require repeated battery replacement surgeries, although newer rechargeable devices are becoming available.

Finally, there is a nontrivial incidence of hardware-related complications when devices are used continuously for many years. Although individual dystonia patients vary in the acuity of their response to the cessation of stimulation,⁶ deterioration can be acute and dramatic. In long-term studies of bilateral pallidal stimulation described above, hardware failures were the most commonly reported adverse events, including unilateral or bilateral lead fracture.^7,9 These appear to be more frequent in patients with dystonia than in other movement disorders.

SUMMARY AND CONCLUSIONS

Deep brain stimulation produces marked and long-lasting improvement in motor function and disability in patients with hyperkinetic disorders. In patients with essential tremor, stimulation usually targets the VIM of the thalamus. Reduction in tremor is most closely related to stimulation frequency and voltage, whereas pulse width has little effect on treatment outcome. In patients with dystonia, stimulation typically targets the GPi or STN. Long-term prospective clinical trials demonstrated significant reductions in motor severity rating scale scores. Selecting patients for DBS requires careful consideration of a range of factors, including the specific clinical presentation, treatment history, and social support. Areas of current investigation include optimal stimulation programming, intraoperative mapping, and the long-term efficacy and safety of stimulation.

Over the last decade, several studies have demonstrated that deep brain stimulation (DBS) is among the most effective approaches for the treatment of patients with advanced movement disorders, including chorea, levodopa-induced dyskinesia, tremor, and dystonia.¹ The goal of DBS is to restore function or relieve pain by stimulating neuronal activity through surgically implanted electrodes. DBS produces marked and persistent reductions in abnormal movements in patients with common hyperkinetic disorders, with a generally low incidence of serious adverse events in pediatric patients and adults.

DEEP BRAIN STIMULATION FOR ESSENTIAL TREMOR

Improvement with thalamic DBS

The ventral intermediate nucleus (VIM) of the thalamus is the most common target for DBS treatment of essential tremor. Several studies have demonstrated significant long-term improvement in tremor following thalamic DBS.³ Most studies enrolled 20 to 30 patients, who were followed for 1 to 5 years after device implantation. On average, these studies reported an improvement in overall tremor of approximately 50% from baseline with thalamic DBS.

Patient selection and stimulation parameters

Symptoms targeted for DBS treatment include unilateral and sometimes bilateral limb tremor. Some evidence exists for effectiveness in axial and vocal tremor as well. Factors to consider in patient selection for DBS surgery include tremor severity, degree of refractoriness to medication, and type of tremor. In addition, individual patient characteristics should be considered, including age, comorbid conditions, surgical risk, patient preference, social and employment factors, and social support.

Reprinted with permission from Journal of Clinical Neurophysiology (Cooper SE, et al. A model predicting optimal parameters for deep brain stimulation in essential tremor. J Clin Neurophysiol 2008; 2:26–273). Copyright © 2008 by the ACNS.

DEEP BRAIN STIMULATION FOR DYSTONIA

Dystonia is characterized by involuntary twisting muscle contractions causing abnormal postures sometimes accompanied by jerky or repetitive involuntary movements. It may be classified according to the body part affected as generalized, segmental, or focal; in some cases it may be classified as multifocal dystonia or hemidystonia. Dystonia is also classified as primary or secondary, according to etiology. Primary dystonias are those not caused by any other identifiable condition and not associated with other neurologic abnormalities. These include idiopathic and some genetic dystonias, such as the DYT1 torsinA gene mutation. DBS of the globus pallidus internus (GPi) or subthalamic nucleus (STN) was approved by the US Food and Drug Administration under a humanitarian device exemption in 2003 for the treatment of primary generalized dystonia (PGD) in patients aged 7 years and older; GPi is the more common target).¹

Evidence of efficacy

Several clinical studies have demonstrated the efficacy of DBS for patients with disabling PGD that is unresponsive to pharmacotherapy.

Long-term efficacy. Isaias and colleagues examined long-term safety and efficacy of DBS in 30 consecutive patients with PGD who were followed for at least 3 years after pallidal DBS surgery.⁶ DBS was delivered bilaterally in 28 patients and unilaterally in 2 patients. Clinical rating scales of motor function improved by a mean of 82.5% after 2 years, and dystonia-related disability improved by a mean of 75.2%. Improvement in motor function from baseline was noted for all 30 subjects. In five patients who were followed for 7 years, improvement in motor function remained greater than 80% at the last follow-up visit. Transient regressions were noted for patients with hardware failures or whose batteries had reached the end of life. Stimulation-related adverse events were reported for three patients and included speech difficulties and, in one patient, transient blepharospasm.

Vidailhet and colleagues examined the efficacy of bilateral pallidal stimulation in 22 patients with PGD who were followed prospectively for 3 years.⁷ Mean improvement from baseline in motor function on a dystonia rating scale was 51% after 1 year and 58% after 3 years (P = .03). Significant improvement was noted for individual ratings of upper and lower limb function scores. Health-related quality of life was also significantly improved at 3-year follow-up (P = .05). Serious adverse events were reported for three patients, including two lead fractures and one infection.

Results from double-blind trial. Kupsch and colleagues performed a randomized, double-blind clinical trial comparing pallidal DBS versus device implantation and sham stimulation in 40 patients with primary segmental or generalized dystonia.⁸ After 3 months, the mean change from baseline in severity of dystonia was 15.8% for patients who received DBS versus 1.4% with sham stimulation (P < .001). At the conclusion of the double-blind treatment phase, patients entered an open-label extension phase in which all patients received DBS for another 3 months. The initial benefit of treatment was sustained across the entire 6-month study period for patients initially randomized to DBS, whereas patients who were initially randomized to sham stimulation exhibited improved motor function during the open-label extension phase. Ratings of disability and quality of life also improved for patients receiving DBS at the end of the 6-month study. Adverse events included dysarthria (five patients), serious infections (four patients), and lead dislodgement (one patient).

Response with DYT1mutation. Coubes and colleagues examined the long-term efficacy and safety of bilateral DBS in 31 children and adults with PGD.⁹ PGD is associated with autosomal DYT1 mutations in approximately 30% of cases, and these authors examined the effects of treatment in patients with and without the DYT1 mutation. After 2 years of treatment, mean scores on a dystonia clinical rating scale decreased by 79% from baseline, and mean disability ratings decreased by 65%. The improvement in clinical dystonia rating scale scores was significantly greater for children than adults after 2 years (84.7% vs 70.1%; P = .04). In children, functional improvement was greater after 2 years in the subset of patients with DYT1 mutations than in the subset of patients without (76.1% vs 44.5%; P = .03), whereas in adults, DYT1 mutation status did not significantly influence response to treatment. One case of unilateral infection was noted, which required removal of the implant with successful reimplantation 6 months later. No other adverse events were reported.

Appropriate patients for DBS include those with an unequivocal diagnosis of dystonia and significant disability. Etiology and type of dystonia should also be considered. Patients with secondary dystonia (eg, due to structural brain lesions or heredodegenerative disorders) generally do not respond to DBS as well as patients with primary dystonias. A possible exception is tardive dystonia, which is caused by past exposure to dopamine receptor–blocking drugs. Although it is a secondary dystonia, tardive dystonia may respond well to DBS. Data on this point remain limited. Moreover, with tardive dystonia (as well as Sydenham chorea and poststroke hemiballismus), there may be spontaneous remission. DBS in these conditions should therefore be considered when enough time has elapsed that the likelihood of spontaneous remission is low.¹

Not all dystonic symptoms have been shown to respond equally to DBS. Evidence of effectiveness is stronger and more consistent for limb and axial dystonia than for dystonic impairment of speech and swallowing. Phasic dystonia (jerky or rhythmic movements) appears to respond better than fixed postures. A critical point is that fixed postures not caused by electrically active muscle contraction will not respond to DBS. For example, bony deformity of the spine, joint disease, or tendon shortening cannot be expected to improve with DBS. The situation is complicated, since such conditions may develop as secondary consequences of dystonia. The potential for their development may warrant earlier rather than later DBS surgery in childhood-onset PGD.¹⁰

UNRESOLVED ISSUES IN DBS FOR DYSTONIA

How aggressively should other therapies be tried before starting DBS?

Pharmacologic options include a range of oral, intramuscular, and intrathecal agents. Injection of botulinum toxin to denervate affected muscles is a mainstay of treatment for focal or segmental dystonia, but often fails to improve symptoms because of the involvement of a large number of muscles, complexity of the movement pattern, or the development of neutralizing antibodies.⁸ With the exception of levodopa-responsive PGD, other pharmacologic therapy for PGD is generally of limited effectiveness for controlling symptoms of dystonia.⁹ Oral or intrathecal baclofen may improve symptoms, but often produces unacceptable sedation.

How important is intraoperative microelectrophysiology?

Although contemporary imaging techniques are important in the correct placement of stimulating electrodes, the available techniques do not always provide sufficient resolution to delineate the STN or GPi. The accuracy of electrode placement may also be influenced by distortions caused by lack of homogeneity among magnetic resonance images, brain shift, and signal deflections from cannulae or electrodes.¹⁴ These errors may result in significant deviation of electrode placement from the intended target. Microelectrode recording and micro-stimulation may be used to map the target region and refine the surgical target. It is widely, but not universally, held that this strategy contributes to superior accuracy and outcomes; it ordinarily requires an awake procedure, which is not always feasible in patients with severe dystonia or in pediatric patients.¹¹

How should be programming (stimulator adjustment) be performed?

Research continues to refine our understanding of how electrical parameters such as voltage, frequency, and pulse width affect clinical outcomes in patients undergoing DBS for dystonia. Some programming approaches, such as long pulse width and high frequency, that were once generally accepted are now widely questioned. Another major unresolved question is: “How long should it take to see the results of stimulation?” In the clinical studies described above, continued improvement was generally observed over months or even years, and, in most patients, stimulators are incrementally adjusted over an extended period. However, some patients may experience much more rapid onset of benefit.

Long-term DBS management of dystonia

Unlike DBS for Parkinson disease or even essential tremor, DBS for dystonia is performed in young patients. This creates special challenges in pediatric patients, who can be expected to grow and develop after device implantation. As a result, children may require additional surgeries to reposition devices.

In addition, the most widely used devices require repeated battery replacement surgeries, although newer rechargeable devices are becoming available.

Finally, there is a nontrivial incidence of hardware-related complications when devices are used continuously for many years. Although individual dystonia patients vary in the acuity of their response to the cessation of stimulation,⁶ deterioration can be acute and dramatic. In long-term studies of bilateral pallidal stimulation described above, hardware failures were the most commonly reported adverse events, including unilateral or bilateral lead fracture.^7,9 These appear to be more frequent in patients with dystonia than in other movement disorders.

SUMMARY AND CONCLUSIONS

Deep brain stimulation produces marked and long-lasting improvement in motor function and disability in patients with hyperkinetic disorders. In patients with essential tremor, stimulation usually targets the VIM of the thalamus. Reduction in tremor is most closely related to stimulation frequency and voltage, whereas pulse width has little effect on treatment outcome. In patients with dystonia, stimulation typically targets the GPi or STN. Long-term prospective clinical trials demonstrated significant reductions in motor severity rating scale scores. Selecting patients for DBS requires careful consideration of a range of factors, including the specific clinical presentation, treatment history, and social support. Areas of current investigation include optimal stimulation programming, intraoperative mapping, and the long-term efficacy and safety of stimulation.

Over the last decade, several studies have demonstrated that deep brain stimulation (DBS) is among the most effective approaches for the treatment of patients with advanced movement disorders, including chorea, levodopa-induced dyskinesia, tremor, and dystonia.¹ The goal of DBS is to restore function or relieve pain by stimulating neuronal activity through surgically implanted electrodes. DBS produces marked and persistent reductions in abnormal movements in patients with common hyperkinetic disorders, with a generally low incidence of serious adverse events in pediatric patients and adults.

DEEP BRAIN STIMULATION FOR ESSENTIAL TREMOR

Improvement with thalamic DBS

The ventral intermediate nucleus (VIM) of the thalamus is the most common target for DBS treatment of essential tremor. Several studies have demonstrated significant long-term improvement in tremor following thalamic DBS.³ Most studies enrolled 20 to 30 patients, who were followed for 1 to 5 years after device implantation. On average, these studies reported an improvement in overall tremor of approximately 50% from baseline with thalamic DBS.

Patient selection and stimulation parameters

Symptoms targeted for DBS treatment include unilateral and sometimes bilateral limb tremor. Some evidence exists for effectiveness in axial and vocal tremor as well. Factors to consider in patient selection for DBS surgery include tremor severity, degree of refractoriness to medication, and type of tremor. In addition, individual patient characteristics should be considered, including age, comorbid conditions, surgical risk, patient preference, social and employment factors, and social support.

Reprinted with permission from Journal of Clinical Neurophysiology (Cooper SE, et al. A model predicting optimal parameters for deep brain stimulation in essential tremor. J Clin Neurophysiol 2008; 2:26–273). Copyright © 2008 by the ACNS.

DEEP BRAIN STIMULATION FOR DYSTONIA

Dystonia is characterized by involuntary twisting muscle contractions causing abnormal postures sometimes accompanied by jerky or repetitive involuntary movements. It may be classified according to the body part affected as generalized, segmental, or focal; in some cases it may be classified as multifocal dystonia or hemidystonia. Dystonia is also classified as primary or secondary, according to etiology. Primary dystonias are those not caused by any other identifiable condition and not associated with other neurologic abnormalities. These include idiopathic and some genetic dystonias, such as the DYT1 torsinA gene mutation. DBS of the globus pallidus internus (GPi) or subthalamic nucleus (STN) was approved by the US Food and Drug Administration under a humanitarian device exemption in 2003 for the treatment of primary generalized dystonia (PGD) in patients aged 7 years and older; GPi is the more common target).¹

Evidence of efficacy

Several clinical studies have demonstrated the efficacy of DBS for patients with disabling PGD that is unresponsive to pharmacotherapy.

Long-term efficacy. Isaias and colleagues examined long-term safety and efficacy of DBS in 30 consecutive patients with PGD who were followed for at least 3 years after pallidal DBS surgery.⁶ DBS was delivered bilaterally in 28 patients and unilaterally in 2 patients. Clinical rating scales of motor function improved by a mean of 82.5% after 2 years, and dystonia-related disability improved by a mean of 75.2%. Improvement in motor function from baseline was noted for all 30 subjects. In five patients who were followed for 7 years, improvement in motor function remained greater than 80% at the last follow-up visit. Transient regressions were noted for patients with hardware failures or whose batteries had reached the end of life. Stimulation-related adverse events were reported for three patients and included speech difficulties and, in one patient, transient blepharospasm.

Vidailhet and colleagues examined the efficacy of bilateral pallidal stimulation in 22 patients with PGD who were followed prospectively for 3 years.⁷ Mean improvement from baseline in motor function on a dystonia rating scale was 51% after 1 year and 58% after 3 years (P = .03). Significant improvement was noted for individual ratings of upper and lower limb function scores. Health-related quality of life was also significantly improved at 3-year follow-up (P = .05). Serious adverse events were reported for three patients, including two lead fractures and one infection.

Results from double-blind trial. Kupsch and colleagues performed a randomized, double-blind clinical trial comparing pallidal DBS versus device implantation and sham stimulation in 40 patients with primary segmental or generalized dystonia.⁸ After 3 months, the mean change from baseline in severity of dystonia was 15.8% for patients who received DBS versus 1.4% with sham stimulation (P < .001). At the conclusion of the double-blind treatment phase, patients entered an open-label extension phase in which all patients received DBS for another 3 months. The initial benefit of treatment was sustained across the entire 6-month study period for patients initially randomized to DBS, whereas patients who were initially randomized to sham stimulation exhibited improved motor function during the open-label extension phase. Ratings of disability and quality of life also improved for patients receiving DBS at the end of the 6-month study. Adverse events included dysarthria (five patients), serious infections (four patients), and lead dislodgement (one patient).

Response with DYT1mutation. Coubes and colleagues examined the long-term efficacy and safety of bilateral DBS in 31 children and adults with PGD.⁹ PGD is associated with autosomal DYT1 mutations in approximately 30% of cases, and these authors examined the effects of treatment in patients with and without the DYT1 mutation. After 2 years of treatment, mean scores on a dystonia clinical rating scale decreased by 79% from baseline, and mean disability ratings decreased by 65%. The improvement in clinical dystonia rating scale scores was significantly greater for children than adults after 2 years (84.7% vs 70.1%; P = .04). In children, functional improvement was greater after 2 years in the subset of patients with DYT1 mutations than in the subset of patients without (76.1% vs 44.5%; P = .03), whereas in adults, DYT1 mutation status did not significantly influence response to treatment. One case of unilateral infection was noted, which required removal of the implant with successful reimplantation 6 months later. No other adverse events were reported.

Appropriate patients for DBS include those with an unequivocal diagnosis of dystonia and significant disability. Etiology and type of dystonia should also be considered. Patients with secondary dystonia (eg, due to structural brain lesions or heredodegenerative disorders) generally do not respond to DBS as well as patients with primary dystonias. A possible exception is tardive dystonia, which is caused by past exposure to dopamine receptor–blocking drugs. Although it is a secondary dystonia, tardive dystonia may respond well to DBS. Data on this point remain limited. Moreover, with tardive dystonia (as well as Sydenham chorea and poststroke hemiballismus), there may be spontaneous remission. DBS in these conditions should therefore be considered when enough time has elapsed that the likelihood of spontaneous remission is low.¹

Not all dystonic symptoms have been shown to respond equally to DBS. Evidence of effectiveness is stronger and more consistent for limb and axial dystonia than for dystonic impairment of speech and swallowing. Phasic dystonia (jerky or rhythmic movements) appears to respond better than fixed postures. A critical point is that fixed postures not caused by electrically active muscle contraction will not respond to DBS. For example, bony deformity of the spine, joint disease, or tendon shortening cannot be expected to improve with DBS. The situation is complicated, since such conditions may develop as secondary consequences of dystonia. The potential for their development may warrant earlier rather than later DBS surgery in childhood-onset PGD.¹⁰

UNRESOLVED ISSUES IN DBS FOR DYSTONIA

How aggressively should other therapies be tried before starting DBS?

Pharmacologic options include a range of oral, intramuscular, and intrathecal agents. Injection of botulinum toxin to denervate affected muscles is a mainstay of treatment for focal or segmental dystonia, but often fails to improve symptoms because of the involvement of a large number of muscles, complexity of the movement pattern, or the development of neutralizing antibodies.⁸ With the exception of levodopa-responsive PGD, other pharmacologic therapy for PGD is generally of limited effectiveness for controlling symptoms of dystonia.⁹ Oral or intrathecal baclofen may improve symptoms, but often produces unacceptable sedation.

How important is intraoperative microelectrophysiology?

Although contemporary imaging techniques are important in the correct placement of stimulating electrodes, the available techniques do not always provide sufficient resolution to delineate the STN or GPi. The accuracy of electrode placement may also be influenced by distortions caused by lack of homogeneity among magnetic resonance images, brain shift, and signal deflections from cannulae or electrodes.¹⁴ These errors may result in significant deviation of electrode placement from the intended target. Microelectrode recording and micro-stimulation may be used to map the target region and refine the surgical target. It is widely, but not universally, held that this strategy contributes to superior accuracy and outcomes; it ordinarily requires an awake procedure, which is not always feasible in patients with severe dystonia or in pediatric patients.¹¹

How should be programming (stimulator adjustment) be performed?

Research continues to refine our understanding of how electrical parameters such as voltage, frequency, and pulse width affect clinical outcomes in patients undergoing DBS for dystonia. Some programming approaches, such as long pulse width and high frequency, that were once generally accepted are now widely questioned. Another major unresolved question is: “How long should it take to see the results of stimulation?” In the clinical studies described above, continued improvement was generally observed over months or even years, and, in most patients, stimulators are incrementally adjusted over an extended period. However, some patients may experience much more rapid onset of benefit.

Long-term DBS management of dystonia

Unlike DBS for Parkinson disease or even essential tremor, DBS for dystonia is performed in young patients. This creates special challenges in pediatric patients, who can be expected to grow and develop after device implantation. As a result, children may require additional surgeries to reposition devices.

In addition, the most widely used devices require repeated battery replacement surgeries, although newer rechargeable devices are becoming available.

Finally, there is a nontrivial incidence of hardware-related complications when devices are used continuously for many years. Although individual dystonia patients vary in the acuity of their response to the cessation of stimulation,⁶ deterioration can be acute and dramatic. In long-term studies of bilateral pallidal stimulation described above, hardware failures were the most commonly reported adverse events, including unilateral or bilateral lead fracture.^7,9 These appear to be more frequent in patients with dystonia than in other movement disorders.

SUMMARY AND CONCLUSIONS

Deep brain stimulation produces marked and long-lasting improvement in motor function and disability in patients with hyperkinetic disorders. In patients with essential tremor, stimulation usually targets the VIM of the thalamus. Reduction in tremor is most closely related to stimulation frequency and voltage, whereas pulse width has little effect on treatment outcome. In patients with dystonia, stimulation typically targets the GPi or STN. Long-term prospective clinical trials demonstrated significant reductions in motor severity rating scale scores. Selecting patients for DBS requires careful consideration of a range of factors, including the specific clinical presentation, treatment history, and social support. Areas of current investigation include optimal stimulation programming, intraoperative mapping, and the long-term efficacy and safety of stimulation.

The classic case of aortic stenosis is in an otherwise healthy middle-aged patient with symptomatic severe disease who is referred to a cardiac surgeon for surgical aortic valve replacement. Unfortunately, physicians who manage valvular heart disease do not encounter this straightforward scenario on a regular basis. Rather, patients come with comorbidities such as advanced age, pulmonary disease, renal dysfunction, coronary artery disease, and significant left ventricular dysfunction. They also come with severe aortic stenosis without symptoms.

See related article

In this issue of the Cleveland Clinic Journal of Medicine, Sawaya and colleagues¹ review the management of aortic stenosis, focusing on clinically challenging scenarios such as low-flow, low-gradient aortic stenosis, low-gradient severe aortic stenosis with a normal ejection fraction, aortic stenosis in elderly patients, moderate aortic stenosis in patients undergoing other cardiac surgery, and transcatheter aortic valve replacement, according to the guidelines from the American College of Cardiology and American Heart Association.²

In addition to the situations covered in their review, a few other complicated situations in patients with severe aortic stenosis also merit discussion. We discuss these below.

ASYMPTOMATIC SEVERE AORTIC STENOSIS AND A NORMAL EJECTION FRACTION

Patients with aortic stenosis may be unaware of their decline in functional capacity, since the illness is gradually progressive. In these patients, exercise testing is often done, as it can uncover limitations and determine the need for aortic valve replacement. Another group of patients with asymptomatic severe aortic stenosis who need aortic valve replacement are those whose ejection fraction is less than 50%.

However, many patients with asymptomatic aortic stenosis pass the stress test with flying colors—no symptoms, no blood pressure changes, no arrhythmias—and have a normal ejection fraction. Managing these patients can be more complicated.

Lancellotti et al³ described a group of patients with asymptomatic severe aortic stenosis, a normal ejection fraction, an aortic valve area smaller than 1 cm², and normal results on exercise testing. Rates of the primary end point (cardiovascular death or need for aortic valve replacement due to symptoms or left ventricular dysfunction) were assessed in subsets of patients grouped on the basis of two variables:

• Left ventricular stroke volume index (flow)—either normal or low (< 35 mL/m²) and
• Mean gradient—either high or low (< 40 mm Hg).

The prevalence rates and 2-year event rates (which were substantial) were as follows:

• Normal flow, high gradient—51% of patients; event rate 56%
• Normal flow, low gradient—31% of patients; event rate 17%
• Low flow, high gradient—10% of patients; event rate 70%
• Low flow, low gradient—7% of patients; event rate 73%.

Mihaljevic et al⁴ at our institution found that left ventricular hypertrophy at the time of surgery for aortic stenosis may have lasting negative consequences. In an observational study of 3,049 patients who underwent aortic valve replacement, severe left ventricular hypertrophy preceded symptoms in 17%. Additionally, the survival rate at 10 years in the group whose left ventricular mass was greater than 185 g/m² was 45% at 10 years, compared with 65% in patients whose left ventricular mass was less than 100 g/m². Left ventricular hypertrophy may, therefore, eventually become another factor that we use in defining the appropriateness of surgery in patients with severe but asymptomatic aortic stenosis.

Comment. Not all patients who have severe aortic stenosis, no symptoms, and a “normal” ejection fraction are the same. Our view of what constitutes appropriate left ventricular function in aortic stenosis has changed and now encompasses diastolic filling values, myocardial velocity, and patterns of hypertrophy in addition to ejection fraction. Surgery is already considered reasonable for patients with asymptomatic but “extremely severe” aortic stenosis (aortic valve area < 0.6 cm², jet velocity > 5 m/sec, mean gradient > 60 mm Hg), and it may improve long-term survival.^2,5

However, closer inspection of left ventricular mechanics may also identify another group of patients whose prognosis is worse than in the rest. It is possible that a more thorough evaluation of left ventricular mechanics, including strain imaging, will provide a more elegant way to risk-stratify patients and help clinicians decide when to intervene in this difficult group of patients.⁶

While these factors are not yet a part of the diagnostic algorithm, the work by Lancellotti et al³ and Mihaljevic et al⁴ sheds light on the idea that evaluation of advanced echocardiographic variables may provide clinical insights into whether patients should undergo aortic valve replacement.

PCI FOR CONCOMITANT SEVERE CORONARY ARTERY DISEASE

The risk factors for aortic stenosis are similar to those for coronary artery disease, and many patients with moderate or severe aortic stenosis also have significant coronary disease. These patients are traditionally referred for combined surgical aortic valve replacement and coronary artery bypass grafting.

Patients who have the combination of both diseases have a worse prognosis, and adding coronary artery bypass grafting to surgical aortic valve replacement increases the perioperative mortality rate.⁷

With advances in transcatheter aortic valve replacement, attention has turned to managing concomitant coronary artery disease percutaneously as well. Until recently, however, there were few data on the safety of percutaneous coronary intervention (PCI) in patients with severe aortic stenosis.

Goel et al⁸ analyzed the outcomes of 254 patients with severe aortic stenosis who underwent PCI at our institution, compared with a propensity-matched group of 508 patients without aortic stenosis undergoing PCI. Overall, the 30-day mortality rate did not differ significantly between the two groups (4.3% vs 4.7%, P = .20), nor did the rate of complications such as contrast nephropathy, periprocedural myocardial infarction, and hemodynamic compromise during the procedure. In subgroup analysis, patients who had severe aortic stenosis and ejection fractions of 30% or less had a significantly higher risk of death than those with ejection fractions greater than 30% (15.4% vs 1.2%, P < .001).

Comment. This information is important, since many patients with severe aortic stenosis also have coronary artery disease. Certainly, for patients with significant coronary artery disease and severe aortic stenosis who cannot undergo surgery, the findings are especially encouraging with respect to the safety of PCI.

The findings also suggest that in patients for whom transcatheter aortic valve replacement can be performed in a timely fashion, a completely percutaneous approach to treating aortic stenosis and coronary artery disease may be reasonable. This hypothesis must be further investigated, but the preliminary data are encouraging.

TRANSCATHETER AORTIC VALVE REPLACEMENT IN LOWER-RISK PATIENTS

The PARTNER (Placement of Aortic Transcatheter Valves) trial showed that transcatheter aortic valve replacement was superior to medical therapy alone for patients who cannot undergo surgery, and not inferior to surgical aortic valve replacement for patients at high surgical risk, ie, a Society of Thoracic Surgeons (STS) mortality risk score greater than 10%.⁹

Given these encouraging results, the PARTNER II trial is now randomizing patients who are at moderate surgical risk (STS score > 4%) to surgical vs transcatheter aortic valve replacement.

Since transcatheter aortic valve replacement has been performed in Europe under the Conformité Européenne (CE) marking since 2007, diffusion of the procedure there has occurred in a more rapid fashion than in the United States. As a result, a number of patients with low or moderate surgical risk have undergone this procedure.

Lange et al¹⁰ summarized their experience at a single center in Munich, Germany, with an eye toward patient selection and surgical risk. Between 2007 and 2010, 420 patients underwent transcatheter aortic valve replacement. When the authors divided the cases into quartiles according to the sequence in which they were seen, they found a statistically significant decline in the STS score over time, from 7.1% in the earliest quartile to 4.8% in the latest quartile (P < .001), indicating the procedure was diffusing into lower-risk groups. With respect to outcome, the 6-month mortality rate declined from 23.5% to 12.4%; this was likely due to a combination of patient-related factors (more patients at lower risk over time), device advances, and greater operator experience. Also of note, only 70% of patients in the latest quartile were intubated for the procedure.

Comment. Diffusion of transcatheter aortic valve replacement in the United States is following a thoughtful path, with patients being assessed by “heart teams” of clinical cardiologists, interventional cardiologists, imaging cardiologists, and cardiac surgeons, and with strict criteria for site approval to perform commercial placement of the Edwards Sapien valve. In keeping with this controlled process, future randomized studies (such as PARTNER II) of transcatheter aortic valve replacement in lower-risk patients will be necessary before this procedure can be widely applied to this patient group. The results are, therefore, eagerly anticipated, but preliminary experience from Europe is encouraging.

BALLOON AORTIC VALVULOPLASTY IS SEEING A RESURGENCE

In large part due to rising interest in managing aortic stenosis and to the anticipated diffusion of transcatheter aortic valve replacement, balloon aortic valvuloplasty has seen a resurgence in recent years.

This procedure can be considered in a number of situations. In patients with severe aortic stenosis who are hemodynamically unstable and for whom urgent aortic valve replacement is not feasible, balloon valvuloplasty may serve as a “bridge” to valve replacement. Similarly, we have seen significant functional improvement in patients after balloon aortic valvuloplasty, so that some who initially were unable to undergo aortic valve replacement have improved to a point that either transcatheter or surgical replacement could be performed safely. In patients who need urgent noncardiac surgery, balloon valvuloplasty may be considered as a temporizing measure in the hope of reducing the risks of perioperative hemodynamic changes associated with anesthesia.

Many patients with severe aortic stenosis have comorbidities such as chronic obstructive pulmonary disease or liver or kidney disease that make it difficult to discern the degree to which aortic stenosis contributes to their symptoms. In such cases, the balloon procedure may provide a therapeutic answer; improvement of symptoms points to aortic stenosis as the driver of symptoms and may push for a more definitive valve replacement option.

Finally, in patients with no option for either transcatheter or surgical aortic valve replacement, balloon aortic valvuloplasty may be considered as a palliative measure.

The benefit of this procedure is only temporary, and restenosis generally occurs within 6 months. Therefore, its value as a stand-alone procedure is limited, and the overall survival rate is significantly improved only when it is used as a bridge to valve replacement.

It should be noted that balloon aortic valvuloplasty carries significant risk. The 30-day mortality rate may be as high as 10%, usually due to either aortic regurgitation (as a complication of the procedure) or persistent heart failure. Other complications occur in up to 15% of cases and include stroke, peripheral vascular complications (due to the size of the devices used and concomitant incidence of peripheral arterial disease), coronary occlusion, need for permanent pacemaker implantation, cardiac tamponade, and cardiac arrest. In patients who require a repeat procedure, it entails similar risks and outcomes as the first procedure.

Comment. Balloon aortic valvuloplasty holds an important place in the treatment of patients with severe aortic stenosis. In our experience, it is most often performed to bridge severely symptomatic patients to transcatheter or surgical aortic valve replacement, or to better understand the contribution of aortic stenosis to functional limitation in patients with multiple comorbidities. It has tremendous potential to alleviate symptoms and provide an opportunity for functional improvement, in turn allowing definitive treatment with aortic valve replacement and improved quality and quantity of life in patients with severe aortic stenosis.

MANAGING SEVERE STENOSIS IS FULFILLING, BUT CHALLENGING

Managing patients with severe aortic stenosis is very fulfilling but at the same time can be extraordinarily challenging. It requires a patient-by-patient analysis of clinical, echocardiographic, and hemodynamic data. In some cases, the relationship between aortic stenosis and current symptoms or future outcomes may be in doubt, and provocative testing or balloon aortic valvuloplasty may be necessary to provide further direction. A meticulous assessment, requiring the expertise of clinicians, imagers, interventionalists, and surgeons is often necessary to deliver optimal care to this group of patients.

The classic case of aortic stenosis is in an otherwise healthy middle-aged patient with symptomatic severe disease who is referred to a cardiac surgeon for surgical aortic valve replacement. Unfortunately, physicians who manage valvular heart disease do not encounter this straightforward scenario on a regular basis. Rather, patients come with comorbidities such as advanced age, pulmonary disease, renal dysfunction, coronary artery disease, and significant left ventricular dysfunction. They also come with severe aortic stenosis without symptoms.

See related article

In this issue of the Cleveland Clinic Journal of Medicine, Sawaya and colleagues¹ review the management of aortic stenosis, focusing on clinically challenging scenarios such as low-flow, low-gradient aortic stenosis, low-gradient severe aortic stenosis with a normal ejection fraction, aortic stenosis in elderly patients, moderate aortic stenosis in patients undergoing other cardiac surgery, and transcatheter aortic valve replacement, according to the guidelines from the American College of Cardiology and American Heart Association.²

In addition to the situations covered in their review, a few other complicated situations in patients with severe aortic stenosis also merit discussion. We discuss these below.

ASYMPTOMATIC SEVERE AORTIC STENOSIS AND A NORMAL EJECTION FRACTION

Patients with aortic stenosis may be unaware of their decline in functional capacity, since the illness is gradually progressive. In these patients, exercise testing is often done, as it can uncover limitations and determine the need for aortic valve replacement. Another group of patients with asymptomatic severe aortic stenosis who need aortic valve replacement are those whose ejection fraction is less than 50%.

However, many patients with asymptomatic aortic stenosis pass the stress test with flying colors—no symptoms, no blood pressure changes, no arrhythmias—and have a normal ejection fraction. Managing these patients can be more complicated.

Lancellotti et al³ described a group of patients with asymptomatic severe aortic stenosis, a normal ejection fraction, an aortic valve area smaller than 1 cm², and normal results on exercise testing. Rates of the primary end point (cardiovascular death or need for aortic valve replacement due to symptoms or left ventricular dysfunction) were assessed in subsets of patients grouped on the basis of two variables:

• Left ventricular stroke volume index (flow)—either normal or low (< 35 mL/m²) and
• Mean gradient—either high or low (< 40 mm Hg).

The prevalence rates and 2-year event rates (which were substantial) were as follows:

• Normal flow, high gradient—51% of patients; event rate 56%
• Normal flow, low gradient—31% of patients; event rate 17%
• Low flow, high gradient—10% of patients; event rate 70%
• Low flow, low gradient—7% of patients; event rate 73%.

Mihaljevic et al⁴ at our institution found that left ventricular hypertrophy at the time of surgery for aortic stenosis may have lasting negative consequences. In an observational study of 3,049 patients who underwent aortic valve replacement, severe left ventricular hypertrophy preceded symptoms in 17%. Additionally, the survival rate at 10 years in the group whose left ventricular mass was greater than 185 g/m² was 45% at 10 years, compared with 65% in patients whose left ventricular mass was less than 100 g/m². Left ventricular hypertrophy may, therefore, eventually become another factor that we use in defining the appropriateness of surgery in patients with severe but asymptomatic aortic stenosis.

Comment. Not all patients who have severe aortic stenosis, no symptoms, and a “normal” ejection fraction are the same. Our view of what constitutes appropriate left ventricular function in aortic stenosis has changed and now encompasses diastolic filling values, myocardial velocity, and patterns of hypertrophy in addition to ejection fraction. Surgery is already considered reasonable for patients with asymptomatic but “extremely severe” aortic stenosis (aortic valve area < 0.6 cm², jet velocity > 5 m/sec, mean gradient > 60 mm Hg), and it may improve long-term survival.^2,5

However, closer inspection of left ventricular mechanics may also identify another group of patients whose prognosis is worse than in the rest. It is possible that a more thorough evaluation of left ventricular mechanics, including strain imaging, will provide a more elegant way to risk-stratify patients and help clinicians decide when to intervene in this difficult group of patients.⁶

While these factors are not yet a part of the diagnostic algorithm, the work by Lancellotti et al³ and Mihaljevic et al⁴ sheds light on the idea that evaluation of advanced echocardiographic variables may provide clinical insights into whether patients should undergo aortic valve replacement.

PCI FOR CONCOMITANT SEVERE CORONARY ARTERY DISEASE

The risk factors for aortic stenosis are similar to those for coronary artery disease, and many patients with moderate or severe aortic stenosis also have significant coronary disease. These patients are traditionally referred for combined surgical aortic valve replacement and coronary artery bypass grafting.

Patients who have the combination of both diseases have a worse prognosis, and adding coronary artery bypass grafting to surgical aortic valve replacement increases the perioperative mortality rate.⁷

With advances in transcatheter aortic valve replacement, attention has turned to managing concomitant coronary artery disease percutaneously as well. Until recently, however, there were few data on the safety of percutaneous coronary intervention (PCI) in patients with severe aortic stenosis.

Goel et al⁸ analyzed the outcomes of 254 patients with severe aortic stenosis who underwent PCI at our institution, compared with a propensity-matched group of 508 patients without aortic stenosis undergoing PCI. Overall, the 30-day mortality rate did not differ significantly between the two groups (4.3% vs 4.7%, P = .20), nor did the rate of complications such as contrast nephropathy, periprocedural myocardial infarction, and hemodynamic compromise during the procedure. In subgroup analysis, patients who had severe aortic stenosis and ejection fractions of 30% or less had a significantly higher risk of death than those with ejection fractions greater than 30% (15.4% vs 1.2%, P < .001).

Comment. This information is important, since many patients with severe aortic stenosis also have coronary artery disease. Certainly, for patients with significant coronary artery disease and severe aortic stenosis who cannot undergo surgery, the findings are especially encouraging with respect to the safety of PCI.

The findings also suggest that in patients for whom transcatheter aortic valve replacement can be performed in a timely fashion, a completely percutaneous approach to treating aortic stenosis and coronary artery disease may be reasonable. This hypothesis must be further investigated, but the preliminary data are encouraging.

TRANSCATHETER AORTIC VALVE REPLACEMENT IN LOWER-RISK PATIENTS

The PARTNER (Placement of Aortic Transcatheter Valves) trial showed that transcatheter aortic valve replacement was superior to medical therapy alone for patients who cannot undergo surgery, and not inferior to surgical aortic valve replacement for patients at high surgical risk, ie, a Society of Thoracic Surgeons (STS) mortality risk score greater than 10%.⁹

Given these encouraging results, the PARTNER II trial is now randomizing patients who are at moderate surgical risk (STS score > 4%) to surgical vs transcatheter aortic valve replacement.

Since transcatheter aortic valve replacement has been performed in Europe under the Conformité Européenne (CE) marking since 2007, diffusion of the procedure there has occurred in a more rapid fashion than in the United States. As a result, a number of patients with low or moderate surgical risk have undergone this procedure.

Lange et al¹⁰ summarized their experience at a single center in Munich, Germany, with an eye toward patient selection and surgical risk. Between 2007 and 2010, 420 patients underwent transcatheter aortic valve replacement. When the authors divided the cases into quartiles according to the sequence in which they were seen, they found a statistically significant decline in the STS score over time, from 7.1% in the earliest quartile to 4.8% in the latest quartile (P < .001), indicating the procedure was diffusing into lower-risk groups. With respect to outcome, the 6-month mortality rate declined from 23.5% to 12.4%; this was likely due to a combination of patient-related factors (more patients at lower risk over time), device advances, and greater operator experience. Also of note, only 70% of patients in the latest quartile were intubated for the procedure.

Comment. Diffusion of transcatheter aortic valve replacement in the United States is following a thoughtful path, with patients being assessed by “heart teams” of clinical cardiologists, interventional cardiologists, imaging cardiologists, and cardiac surgeons, and with strict criteria for site approval to perform commercial placement of the Edwards Sapien valve. In keeping with this controlled process, future randomized studies (such as PARTNER II) of transcatheter aortic valve replacement in lower-risk patients will be necessary before this procedure can be widely applied to this patient group. The results are, therefore, eagerly anticipated, but preliminary experience from Europe is encouraging.

BALLOON AORTIC VALVULOPLASTY IS SEEING A RESURGENCE

In large part due to rising interest in managing aortic stenosis and to the anticipated diffusion of transcatheter aortic valve replacement, balloon aortic valvuloplasty has seen a resurgence in recent years.

This procedure can be considered in a number of situations. In patients with severe aortic stenosis who are hemodynamically unstable and for whom urgent aortic valve replacement is not feasible, balloon valvuloplasty may serve as a “bridge” to valve replacement. Similarly, we have seen significant functional improvement in patients after balloon aortic valvuloplasty, so that some who initially were unable to undergo aortic valve replacement have improved to a point that either transcatheter or surgical replacement could be performed safely. In patients who need urgent noncardiac surgery, balloon valvuloplasty may be considered as a temporizing measure in the hope of reducing the risks of perioperative hemodynamic changes associated with anesthesia.

Many patients with severe aortic stenosis have comorbidities such as chronic obstructive pulmonary disease or liver or kidney disease that make it difficult to discern the degree to which aortic stenosis contributes to their symptoms. In such cases, the balloon procedure may provide a therapeutic answer; improvement of symptoms points to aortic stenosis as the driver of symptoms and may push for a more definitive valve replacement option.

Finally, in patients with no option for either transcatheter or surgical aortic valve replacement, balloon aortic valvuloplasty may be considered as a palliative measure.

The benefit of this procedure is only temporary, and restenosis generally occurs within 6 months. Therefore, its value as a stand-alone procedure is limited, and the overall survival rate is significantly improved only when it is used as a bridge to valve replacement.

It should be noted that balloon aortic valvuloplasty carries significant risk. The 30-day mortality rate may be as high as 10%, usually due to either aortic regurgitation (as a complication of the procedure) or persistent heart failure. Other complications occur in up to 15% of cases and include stroke, peripheral vascular complications (due to the size of the devices used and concomitant incidence of peripheral arterial disease), coronary occlusion, need for permanent pacemaker implantation, cardiac tamponade, and cardiac arrest. In patients who require a repeat procedure, it entails similar risks and outcomes as the first procedure.

Comment. Balloon aortic valvuloplasty holds an important place in the treatment of patients with severe aortic stenosis. In our experience, it is most often performed to bridge severely symptomatic patients to transcatheter or surgical aortic valve replacement, or to better understand the contribution of aortic stenosis to functional limitation in patients with multiple comorbidities. It has tremendous potential to alleviate symptoms and provide an opportunity for functional improvement, in turn allowing definitive treatment with aortic valve replacement and improved quality and quantity of life in patients with severe aortic stenosis.

MANAGING SEVERE STENOSIS IS FULFILLING, BUT CHALLENGING

Managing patients with severe aortic stenosis is very fulfilling but at the same time can be extraordinarily challenging. It requires a patient-by-patient analysis of clinical, echocardiographic, and hemodynamic data. In some cases, the relationship between aortic stenosis and current symptoms or future outcomes may be in doubt, and provocative testing or balloon aortic valvuloplasty may be necessary to provide further direction. A meticulous assessment, requiring the expertise of clinicians, imagers, interventionalists, and surgeons is often necessary to deliver optimal care to this group of patients.

The classic case of aortic stenosis is in an otherwise healthy middle-aged patient with symptomatic severe disease who is referred to a cardiac surgeon for surgical aortic valve replacement. Unfortunately, physicians who manage valvular heart disease do not encounter this straightforward scenario on a regular basis. Rather, patients come with comorbidities such as advanced age, pulmonary disease, renal dysfunction, coronary artery disease, and significant left ventricular dysfunction. They also come with severe aortic stenosis without symptoms.

See related article

In this issue of the Cleveland Clinic Journal of Medicine, Sawaya and colleagues¹ review the management of aortic stenosis, focusing on clinically challenging scenarios such as low-flow, low-gradient aortic stenosis, low-gradient severe aortic stenosis with a normal ejection fraction, aortic stenosis in elderly patients, moderate aortic stenosis in patients undergoing other cardiac surgery, and transcatheter aortic valve replacement, according to the guidelines from the American College of Cardiology and American Heart Association.²

In addition to the situations covered in their review, a few other complicated situations in patients with severe aortic stenosis also merit discussion. We discuss these below.

ASYMPTOMATIC SEVERE AORTIC STENOSIS AND A NORMAL EJECTION FRACTION

Patients with aortic stenosis may be unaware of their decline in functional capacity, since the illness is gradually progressive. In these patients, exercise testing is often done, as it can uncover limitations and determine the need for aortic valve replacement. Another group of patients with asymptomatic severe aortic stenosis who need aortic valve replacement are those whose ejection fraction is less than 50%.

However, many patients with asymptomatic aortic stenosis pass the stress test with flying colors—no symptoms, no blood pressure changes, no arrhythmias—and have a normal ejection fraction. Managing these patients can be more complicated.

Lancellotti et al³ described a group of patients with asymptomatic severe aortic stenosis, a normal ejection fraction, an aortic valve area smaller than 1 cm², and normal results on exercise testing. Rates of the primary end point (cardiovascular death or need for aortic valve replacement due to symptoms or left ventricular dysfunction) were assessed in subsets of patients grouped on the basis of two variables:

• Left ventricular stroke volume index (flow)—either normal or low (< 35 mL/m²) and
• Mean gradient—either high or low (< 40 mm Hg).

The prevalence rates and 2-year event rates (which were substantial) were as follows:

• Normal flow, high gradient—51% of patients; event rate 56%
• Normal flow, low gradient—31% of patients; event rate 17%
• Low flow, high gradient—10% of patients; event rate 70%
• Low flow, low gradient—7% of patients; event rate 73%.

Mihaljevic et al⁴ at our institution found that left ventricular hypertrophy at the time of surgery for aortic stenosis may have lasting negative consequences. In an observational study of 3,049 patients who underwent aortic valve replacement, severe left ventricular hypertrophy preceded symptoms in 17%. Additionally, the survival rate at 10 years in the group whose left ventricular mass was greater than 185 g/m² was 45% at 10 years, compared with 65% in patients whose left ventricular mass was less than 100 g/m². Left ventricular hypertrophy may, therefore, eventually become another factor that we use in defining the appropriateness of surgery in patients with severe but asymptomatic aortic stenosis.

Comment. Not all patients who have severe aortic stenosis, no symptoms, and a “normal” ejection fraction are the same. Our view of what constitutes appropriate left ventricular function in aortic stenosis has changed and now encompasses diastolic filling values, myocardial velocity, and patterns of hypertrophy in addition to ejection fraction. Surgery is already considered reasonable for patients with asymptomatic but “extremely severe” aortic stenosis (aortic valve area < 0.6 cm², jet velocity > 5 m/sec, mean gradient > 60 mm Hg), and it may improve long-term survival.^2,5

However, closer inspection of left ventricular mechanics may also identify another group of patients whose prognosis is worse than in the rest. It is possible that a more thorough evaluation of left ventricular mechanics, including strain imaging, will provide a more elegant way to risk-stratify patients and help clinicians decide when to intervene in this difficult group of patients.⁶

While these factors are not yet a part of the diagnostic algorithm, the work by Lancellotti et al³ and Mihaljevic et al⁴ sheds light on the idea that evaluation of advanced echocardiographic variables may provide clinical insights into whether patients should undergo aortic valve replacement.

PCI FOR CONCOMITANT SEVERE CORONARY ARTERY DISEASE

The risk factors for aortic stenosis are similar to those for coronary artery disease, and many patients with moderate or severe aortic stenosis also have significant coronary disease. These patients are traditionally referred for combined surgical aortic valve replacement and coronary artery bypass grafting.

Patients who have the combination of both diseases have a worse prognosis, and adding coronary artery bypass grafting to surgical aortic valve replacement increases the perioperative mortality rate.⁷

With advances in transcatheter aortic valve replacement, attention has turned to managing concomitant coronary artery disease percutaneously as well. Until recently, however, there were few data on the safety of percutaneous coronary intervention (PCI) in patients with severe aortic stenosis.

Goel et al⁸ analyzed the outcomes of 254 patients with severe aortic stenosis who underwent PCI at our institution, compared with a propensity-matched group of 508 patients without aortic stenosis undergoing PCI. Overall, the 30-day mortality rate did not differ significantly between the two groups (4.3% vs 4.7%, P = .20), nor did the rate of complications such as contrast nephropathy, periprocedural myocardial infarction, and hemodynamic compromise during the procedure. In subgroup analysis, patients who had severe aortic stenosis and ejection fractions of 30% or less had a significantly higher risk of death than those with ejection fractions greater than 30% (15.4% vs 1.2%, P < .001).

Comment. This information is important, since many patients with severe aortic stenosis also have coronary artery disease. Certainly, for patients with significant coronary artery disease and severe aortic stenosis who cannot undergo surgery, the findings are especially encouraging with respect to the safety of PCI.

The findings also suggest that in patients for whom transcatheter aortic valve replacement can be performed in a timely fashion, a completely percutaneous approach to treating aortic stenosis and coronary artery disease may be reasonable. This hypothesis must be further investigated, but the preliminary data are encouraging.

TRANSCATHETER AORTIC VALVE REPLACEMENT IN LOWER-RISK PATIENTS

The PARTNER (Placement of Aortic Transcatheter Valves) trial showed that transcatheter aortic valve replacement was superior to medical therapy alone for patients who cannot undergo surgery, and not inferior to surgical aortic valve replacement for patients at high surgical risk, ie, a Society of Thoracic Surgeons (STS) mortality risk score greater than 10%.⁹

Given these encouraging results, the PARTNER II trial is now randomizing patients who are at moderate surgical risk (STS score > 4%) to surgical vs transcatheter aortic valve replacement.

Since transcatheter aortic valve replacement has been performed in Europe under the Conformité Européenne (CE) marking since 2007, diffusion of the procedure there has occurred in a more rapid fashion than in the United States. As a result, a number of patients with low or moderate surgical risk have undergone this procedure.

Lange et al¹⁰ summarized their experience at a single center in Munich, Germany, with an eye toward patient selection and surgical risk. Between 2007 and 2010, 420 patients underwent transcatheter aortic valve replacement. When the authors divided the cases into quartiles according to the sequence in which they were seen, they found a statistically significant decline in the STS score over time, from 7.1% in the earliest quartile to 4.8% in the latest quartile (P < .001), indicating the procedure was diffusing into lower-risk groups. With respect to outcome, the 6-month mortality rate declined from 23.5% to 12.4%; this was likely due to a combination of patient-related factors (more patients at lower risk over time), device advances, and greater operator experience. Also of note, only 70% of patients in the latest quartile were intubated for the procedure.

Comment. Diffusion of transcatheter aortic valve replacement in the United States is following a thoughtful path, with patients being assessed by “heart teams” of clinical cardiologists, interventional cardiologists, imaging cardiologists, and cardiac surgeons, and with strict criteria for site approval to perform commercial placement of the Edwards Sapien valve. In keeping with this controlled process, future randomized studies (such as PARTNER II) of transcatheter aortic valve replacement in lower-risk patients will be necessary before this procedure can be widely applied to this patient group. The results are, therefore, eagerly anticipated, but preliminary experience from Europe is encouraging.

BALLOON AORTIC VALVULOPLASTY IS SEEING A RESURGENCE

In large part due to rising interest in managing aortic stenosis and to the anticipated diffusion of transcatheter aortic valve replacement, balloon aortic valvuloplasty has seen a resurgence in recent years.

This procedure can be considered in a number of situations. In patients with severe aortic stenosis who are hemodynamically unstable and for whom urgent aortic valve replacement is not feasible, balloon valvuloplasty may serve as a “bridge” to valve replacement. Similarly, we have seen significant functional improvement in patients after balloon aortic valvuloplasty, so that some who initially were unable to undergo aortic valve replacement have improved to a point that either transcatheter or surgical replacement could be performed safely. In patients who need urgent noncardiac surgery, balloon valvuloplasty may be considered as a temporizing measure in the hope of reducing the risks of perioperative hemodynamic changes associated with anesthesia.

Many patients with severe aortic stenosis have comorbidities such as chronic obstructive pulmonary disease or liver or kidney disease that make it difficult to discern the degree to which aortic stenosis contributes to their symptoms. In such cases, the balloon procedure may provide a therapeutic answer; improvement of symptoms points to aortic stenosis as the driver of symptoms and may push for a more definitive valve replacement option.

Finally, in patients with no option for either transcatheter or surgical aortic valve replacement, balloon aortic valvuloplasty may be considered as a palliative measure.

The benefit of this procedure is only temporary, and restenosis generally occurs within 6 months. Therefore, its value as a stand-alone procedure is limited, and the overall survival rate is significantly improved only when it is used as a bridge to valve replacement.

It should be noted that balloon aortic valvuloplasty carries significant risk. The 30-day mortality rate may be as high as 10%, usually due to either aortic regurgitation (as a complication of the procedure) or persistent heart failure. Other complications occur in up to 15% of cases and include stroke, peripheral vascular complications (due to the size of the devices used and concomitant incidence of peripheral arterial disease), coronary occlusion, need for permanent pacemaker implantation, cardiac tamponade, and cardiac arrest. In patients who require a repeat procedure, it entails similar risks and outcomes as the first procedure.

Comment. Balloon aortic valvuloplasty holds an important place in the treatment of patients with severe aortic stenosis. In our experience, it is most often performed to bridge severely symptomatic patients to transcatheter or surgical aortic valve replacement, or to better understand the contribution of aortic stenosis to functional limitation in patients with multiple comorbidities. It has tremendous potential to alleviate symptoms and provide an opportunity for functional improvement, in turn allowing definitive treatment with aortic valve replacement and improved quality and quantity of life in patients with severe aortic stenosis.

MANAGING SEVERE STENOSIS IS FULFILLING, BUT CHALLENGING

Managing patients with severe aortic stenosis is very fulfilling but at the same time can be extraordinarily challenging. It requires a patient-by-patient analysis of clinical, echocardiographic, and hemodynamic data. In some cases, the relationship between aortic stenosis and current symptoms or future outcomes may be in doubt, and provocative testing or balloon aortic valvuloplasty may be necessary to provide further direction. A meticulous assessment, requiring the expertise of clinicians, imagers, interventionalists, and surgeons is often necessary to deliver optimal care to this group of patients.

For some patients with aortic stenosis, the choice of management is simple; for others it is less so. Patients who have severe, symptomatic stenosis and who have low surgical risk should undergo aortic valve replacement. But if the stenosis is severe but asymptomatic, or if the patient is at higher surgical risk, or if there seems to be a mismatch in the hemodynamic variables, the situation is more complicated.

See related editorial

Fortunately, we have evidence and guidelines to go on. In this paper we review the indications for surgical and transcatheter aortic valve replacement, focusing on the areas of less certainty.

AN INDOLENT DISEASE, UNTIL IT ISN’T

Aortic stenosis is the most common valvular disease and the third most prevalent form of cardiovascular disease in the Western world, after hypertension and coronary artery disease. It is largely a disease of the elderly; its prevalence increases with age, and it is present in 2% to 7% of patients over age 65.^1,2

At first, its course is indolent, as it progresses slowly over years to decades. However, this is followed by rapid clinical deterioration and a high death rate after symptoms develop.

SURGICAL AORTIC VALVE REPLACEMENT FOR SEVERE SYMPTOMATIC STENOSIS

Classic symptoms of aortic stenosis include angina, heart failure, and syncope. Once symptoms appear, patients with severe aortic stenosis should be promptly referred for surgical aortic valve replacement, as survival is poor unless outflow obstruction is relieved (Figure 1). The onset of symptoms confers a poor prognosis: patients die within an average of 5 years after the onset of angina, 3 years after the onset of syncope, and 2 years after the onset of heart failure symptoms. The overall mortality rate is 75% at 3 years without surgery.^3,4 Furthermore, 8% to 34% of patients with symptoms die suddenly.

Advances in prosthetic-valve design, cardiopulmonary bypass, surgical technique, and anesthesia have steadily improved the outcomes of aortic valve surgery. An analysis of the Society of Thoracic Surgeons (STS) database in 2006 showed that during the previous decade the death rate during isolated aortic valve replacement decreased from 3.4% to 2.6%. For patients under age 70 at the time of surgery, the rate of death was 1.3%, and in those ages 80 to 85, the 30-day mortality rate was less than 5%.⁵

Patients who survive surgery enjoy a near-normal life expectancy: 99% survive at least 5 years, 85% at least 10 years, and 82% at least 15 years.^6,7 Nearly all have improvement in their ejection fraction and heart failure symptoms, and those who had more advanced symptoms before surgery enjoy the most benefit afterward.^8,9

Recommendation. Surgical valve replacement for symptomatic severe aortic stenosis receives a class I recommendation, level of evidence B, in the current guidelines from the American College of Cardiology (ACC) and the American Heart Association (AHA).^10,11 (See Table 1 for an explanation of the classes of recommendations and levels of evidence.)

TWO RISK-ASSESSMENT SCORES

There are two widely used scores for assessing the risk of aortic valve replacement: the European System for Cardiac Operative Risk Evaluation (EuroSCORE) and the STS score. Each has limitations.

The EuroSCORE was developed to predict the risk of dying in the hospital after adult cardiac surgery. It has been shown to predict the short-term and the long-term risk of death after heart valve surgery.¹² Unfortunately, it overestimates the dangers of isolated aortic valve replacement in the patients at highest risk.^13,14

The STS score, a logistic model, reflects more closely the operative and 30-day mortality rates for the patients at highest risk undergoing surgical aortic valve replacement.^15,16 It was used to assess patients for surgical or transcatheter aortic valve replacement in the Placement of Aortic Transcatheter Valves (PARTNER) trial.¹⁷

These risk scores, though not perfect, are helpful as part of an overall estimation of risk that includes functional status, cardiac function, and comorbidities.

OTHER INDICATIONS FOR SURGICAL AORTIC VALVE REPLACEMENT

For patients with severe but asymptomatic aortic stenosis, surgical referral is standard practice in several circumstances.

Asymptomatic severe aortic stenosis with a low ejection fraction

Early studies found significant differences in survival beginning as early as 3 years after valve replacement between those whose preoperative ejection fraction was greater than 50% and those with a lower ejection fraction.⁴ Delaying surgery in these patients may lead to irreversible left ventricular dysfunction and worse survival.

Recommendation. The AHA and the ACC recommend surgical aortic valve replacement for patients who have no symptoms and whose left ventricular ejection fraction is less than 50% (class I indication, level of evidence C).^10,11

Asymptomatic severe aortic stenosis in patients undergoing other cardiac surgery

Recommendation. Even if it is causing no symptoms, a severely stenotic aortic valve ought to be replaced if the ejection fraction is greater than 50% and the patient is undergoing another type of heart surgery, such as coronary artery bypass grafting, aortic surgery, or surgery on other heart valves (class I indication, level of evidence B).^10,11

Asymptomatic moderate aortic stenosis in patients undergoing other cardiac surgery

When patients with a mildly or moderately stenotic aortic valve undergo other types of cardiac surgery, the decision to replace the valve is more difficult. Clinicians have to consider the increase in risk caused by adding aortic valve replacement to the planned surgery compared with the future likelihood of aortic stenosis progressing to a severe symptomatic state and eventually requiring a second cardiac surgery.

We have no evidence from a large prospective randomized controlled trial regarding prophylactic valve replacement at the time of coronary bypass surgery. However, a review of outcomes from the STS database between 1995 and 2000 found that patients under age 70 with a peak aortic gradient greater than “about 28 mm Hg” (correlating with a moderate degree of stenosis) benefited from prophylactic valve replacement at the time of coronary artery bypass surgery.¹⁸

These conclusions were supported by a subsequent retrospective analysis that found a significant survival advantage at 8 years in favor of prophylactic valve replacement at the time of bypass surgery for those with moderate (but not mild) aortic stenosis.¹⁹

Recommendation. The AHA and ACC give a class IIb endorsement, level of evidence B, for aortic valve replacement in patients with asymptomatic moderate aortic stenosis undergoing coronary bypass, valve, or aortic surgery.^10,11

SEVERE ASYMPTOMATIC STENOSIS: WHICH TESTS HELP IN DECIDING?

A patient without symptoms presents a greater challenge than one with symptoms.

If surgery is deferred, the prognosis is usually excellent in such patients. Pellikka et al²⁰ found that patients with severe asymptomatic aortic stenosis who did not undergo surgery had a rate of sudden cardiac death of about 1% per year of follow-up. However, physicians worry about missing the rapid development of symptoms of aortic stenosis in patients who previously had none. Pallikka et al also found that, at 5 years, only 20% of patients had not undergone aortic valve replacement or had not died of cardiovascular causes.²⁰

Many researchers advocate surgical aortic valve replacement for severe asymptomatic aortic stenosis. However, the operative risk is 3% overall and has to be weighed against the 1%-per-year risk of death in patients who do not undergo surgery. Therefore, we need a way to identify a subgroup of patients without symptoms who are at higher risk.

Exercise stress testing

Some patients might subconsciously adapt to aortic stenosis by reducing their physical activity. In these “asymptomatic” patients, exercise stress testing can uncover symptoms in around 40%.²¹

In a group of people with severe asymptomatic aortic stenosis, a positive treadmill test (defined as an abnormal blood pressure response, ST segment changes, symptoms such as limiting dyspnea, chest discomfort, or dizziness on a modified Bruce protocol, or complex ventricular arrhythmias) strongly predicted the onset of symptoms or the need for surgery. At 24 months, only 19% of those who had had a positive exercise test result remained alive, symptom-free, and without valve replacement, compared with 85% of those who had had a negative test result.²²

Subsequent study found that symptoms with exercise were the strongest predictor of the onset of symptoms of aortic stenosis, especially among patients under age 70, in whom the symptoms of fatigue and breathlessness are more specific than in the elderly.²³

Recommendation. Exercise testing is recommended in patients with severe asymptomatic aortic stenosis (class IIa indication, level of evidence B) as a means of identifying those who are likely to develop symptoms or who might benefit from surgery. Surgery for those who have an abnormal exercise stress response receives a class IIb, level of evidence C recommendation from the ACC/AHA and a class IC from the European Society of Cardiology.^24,25

Exercise stress echocardiography to measure change in transvalvular gradient

Emerging data suggest that exercise stress echocardiography may provide incremental prognostic information in patients with severe asymptomatic aortic stenosis. In fact, two studies showed that an exercise-induced increase in the transvalvular gradient of more than 20 mm Hg²⁶ or 18 mm Hg²⁷ predicts future cardiac events. This increase reflects fixed valve stenosis with limited valve compliance.

Other echocardiographic variables

Additional data have shown that severe aortic stenosis (valve area < 0.6 cm²), aortic velocity greater than 4.0 m/s, and severe calcification confer a higher risk of developing symptoms.^28,29

Recommendation. The ACC and AHA say that surgical aortic valve replacement may be considered in patients without symptoms who have a high likelihood of rapid progression of aortic stenosis (ie, who are older or have severe calcification or coronary artery disease) or if surgery might be delayed at the time of symptom onset (class IIb, level of evidence C).

Aortic valve replacement can also be considered for extremely severe aortic stenosis (valve area < 0.6 cm²), mean gradient > 60 mm Hg, and velocity > 5.0 m/s if the operative mortality rate is 1.0% or less (class IIb, level of evidence C).

Brain natriuretic peptide levels

Measuring the brain natriuretic peptide (BNP) level may help if symptoms are unclear; higher levels suggest cardiac decompensation.²⁸

One study showed that BNP levels are higher in patients with symptomatic aortic stenosis than in those with asymptomatic severe disease, and correlate with symptom severity.³⁰ In addition, in two other studies, higher BNP and N-terminal BNP levels were shown to predict disease progression, symptom onset, and poorer event-free survival.^31,32

In severe asymptomatic aortic stenosis, natriuretic peptides may provide important prognostic information beyond clinical and echocardiographic evaluation. Furthermore, in a recent study, Monin et al³³ proposed a risk score that integrates peak aortic jet velocity, BNP level, and sex (women being at higher risk) in predicting who would benefit from early surgery in patients with severe asymptomatic aortic stenosis.³³

SPECIAL CONSIDERATIONS

Low-output, low-gradient aortic stenosis: True severe stenosis vs pseudostenosis

Patients with a low ejection fraction (< 50%) and a high mean transvalvular gradient (> 30 or 40 mm Hg) pose no therapeutic dilemma. They have true afterload mismatch and improve markedly with surgery.³⁴ However, patients with an even lower ejection fraction (< 35% or 40%) and a low mean transvalvular gradient (< 30 or 40 mm Hg) pose more of a problem.

It is hard to tell if these patients have true severe aortic stenosis or pseudostenosis due to primary myocardial dysfunction. In pseudostenosis, the aortic valves are moderately diseased, and leaflet opening is reduced by a failing ventricle. When cardiac output is low, the formulae used to calculate the aortic valve area become less accurate, so that patients with cardiomyopathy who have only mild or moderate aortic stenosis may appear to have severe stenosis.

Patients with pseudostenosis have a high risk of dying during surgical aortic valve replacement, approaching 50%, and benefit more from evidence-based heart failure management.^35,36 In patients with true stenosis, ventricular dysfunction is mainly a result of severe stenosis and should improve after aortic valve replacement.

Dobutamine stress echocardiography can be used in patients with low-flow, low-gradient aortic stenosis to distinguish true severe stenosis from pseudostenosis. Dobutamine, an inotropic drug, increases the stroke volume so that patients with true severe aortic stenosis increase their transvalvular gradient and velocity with no or minimal change in the valve area. Conversely, in patients with pseudostenosis, the increase in stroke volume will open the aortic valve further and cause no or minimal increase in transvalvular gradient and velocity, but will increase the calculated valve area, confirming that aortic stenosis only is mild to moderate.³⁷

Patients with low-flow, low-gradient aortic stenosis are at higher risk during surgical aortic valve replacement. Many studies have reported a 30-day mortality rate between 9% and 18%, although risks vary considerably within this population.^38,39

Contractile reserve. Dobutamine stress echocardiography has also been used to identify patients with severe aortic stenosis who can increase their ejection fraction and stroke volume (Figure 2).^40,41 These patients are said to have “contractile reserve” and do better with surgery than those who lack adequate contractile reserve. Contractile reserve is defined as an increase of more than 20% in stroke volume during low-dose dobutamine infusion.^42,43 In one small nonrandomized study, patients with contractile reserve had a 5% mortality rate at 30 days, compared with 32% in patients with no contractile reserve.^44,45

In fact, patients with no contractile reserve have a high operative mortality rate during aortic valve replacement, but those who survive the operation have improvements in symptoms, functional class, and ejection fraction similar to those in patients who do have contractile reserve.⁴⁶

On the other hand, if patients with no contractile reserve are treated conservatively, they have a much worse prognosis than those managed surgically.⁴⁷ While it is true that patients without contractile reserve did not have a statistically significant difference in mortality rates with aortic valve replacement (P = .07) in a study by Monin et al,⁴⁴ the difference was staggering between the group who underwent aortic valve replacement and the group who received medical treatment alone (hazard ratio = 0.47, 95% confidence interval 0.31–1.05, P = .07). The difference in the mortality rates may not have reached statistical significance because of the study’s small sample size.

A few years later, the same group published a similar paper with a larger study sample, focusing on patients with no contractile reserve. Using 42 propensity-matched patients, they found a statistically significantly higher 5-year survival rate in patients with no contractile reserve who underwent aortic valve replacement than in similar patients who received medical management (65% ± 11% vs 11 ± 7%, P = .019).⁴⁷

Hence, surgery may be a better option than medical treatment for this select high-risk group despite the higher operative mortality risk. Transcatheter aortic valve implantation may also offer an interesting alternative to surgical aortic valve replacement in this particular subset of patients.⁴⁸

Low-gradient ‘severe’ aortic stenosis with preserved ejection fraction or ‘paradoxically low-flow aortic stenosis’

Low-gradient “severe” aortic stenosis with a preserved left ventricular ejection fraction is a recently recognized clinical entity in patients with severe aortic stenosis who present with a lower-than-expected transvalvular gradient on the basis of generally accepted values.⁴⁹ (A patient with severe aortic stenosis and preserved ejection fraction is expected to generate a mean transaortic gradient greater than 40 mm Hg.²⁴) This situation remains incompletely understood but has been shown in retrospective studies to foretell a poor prognosis.^50–52

This subgroup of patients has pronounced left ventricular concentric remodeling with a small left ventricular cavity, impaired left ventricular filling, and reduced systolic longitudinal myocardial shortening.⁴⁴

Herrmann et al⁵³ provided more insight into the pathophysiology by showing that patients with this condition exhibit more pronounced myocardial fibrosis on myocardial biopsy and more pronounced late subendocardial enhancement on magnetic resonance imaging. These patients also displayed a significant decrease in mitral ring displacement and systolic strain. These abnormalities result in a low stroke volume despite a preserved ejection fraction and consequently a lower transvalvular gradient (< 40 mm Hg).

This disease pattern, in which the low gradient is interpreted as mild to moderate aortic stenosis, may lead to underestimation of stenosis severity and, thus, to inappropriate delay of aortic valve replacement.

However, other conditions can cause this hemodynamic situation with a lower-than-expected gradient. It can arise from a small left ventricle that correlates with a small body size, yielding a lower-than-normal stroke volume, measurement errors in determining stroke volume and valve area by Doppler echocardiography, systemic hypertension (which can influence estimation of the gradient by Doppler echocardiography), and inconsistency in the definition of severe aortic stenosis in the current guidelines relating to cutoffs of valve area in relation to those of jet velocity and gradient.⁵⁴

This subgroup of patients seems to be at a more advanced stage and has a poorer prognosis if treated medically rather than surgically. When symptomatic, low-gradient severe aortic stenosis should be treated surgically, with one study showing excellent outcomes with aortic valve replacement.⁵⁰

However, a recent study by Jander et al⁵⁵ showed that patients with low-gradient severe aortic stenosis and normal ejection fraction have outcomes similar to those in patients with moderate aortic stenosis, suggesting a strategy of medical therapy and close monitoring.⁵⁵ Of note, the subset of patients reported in this substudy of the Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) trial did not really fit the pattern of low-gradient severe aortic stenosis described by Hachicha et al⁵⁰ and other groups.^51,56 These patients had aortic valve areas in the severe range but mean transaortic gradients in the moderate range, and in light of the other echocardiographic findings in these patients, the area-gradient discordances were predominantly due to small body surface area and measurement errors. These patients indeed had near-normal left ventricular size, no left ventricular hypertrophy, and no evidence of concentric remodeling.

Finally, the findings of the study by Jander et al⁵⁵ are discordant with those of another substudy of the SEAS trial,⁵⁷ which reported that paradoxical low-flow aortic stenosis occurred in about 7% of the cohort (compared with 52% in the study by Jander et al⁵⁵) and was associated with more pronounced concentric remodeling and more severe impairment of myocardial function.

Whether intervention in patients with low-gradient severe aortic stenosis and valve area less than 1.0 cm² improves outcomes remains to be confirmed and reproduced in future prospective studies.

The risks of cardiac surgery increase with age. Older patients may be more deconditioned and have more comorbidities than younger patients, placing them at greater risk of a poor outcome.

Several retrospective studies of valve replacement in octogenarians have found that operative mortality rates range from 5.7% to 9% during isolated aortic valve replacement.^58–60 Note that, using the STS score, the operative mortality risk increases only from 1.2% in a 70-year-old man with no comorbidities to 1.8% in an 80-year-old man undergoing aortic valve replacement plus coronary artery bypass grafting.⁶¹

As in younger patients, valve replacement results in a significant survival benefit and symptomatic improvement. Yet up to 30% of patients with severe aortic stenosis are not referred for surgery because surgery is believed to be too risky.⁶² The conditions most frequently cited by physicians when declining to refer patients for surgery include a low ejection fraction, advanced age, and advanced comorbidities. None of these is an absolute contraindication to surgery.

A recent retrospective study of 443 elderly patients (mean age 79.5) showed that those with left ventricular concentric remodeling, lower stroke volume, elevated left ventricular filling pressures, and mildly elevated pulmonary artery pressures have a very bad prognosis, with a mortality rate of 50.5% at 3.3 ± 2.7 years.⁶³

Despite the higher operative mortality risk, these patients face a dismal prognosis when treated medically and should be referred to a cardiologist or cardiothoracic surgeon for an assessment of their operative risk and, potentially, for referral for catheter-based valve replacement.

Acutely ill patients

In critically ill patients with aortic stenosis and cardiogenic shock, the use of intravenous sodium nitroprusside increases cardiac output and decreases pulmonary artery wedge pressure, allowing patients to transition to surgery or vasodilator therapy. The mechanism seems to be an increase in myocardial contractility rather than a decrease in peripheral resistance. The reduction in filling pressure and concurrent increase in coronary blood flow relieves ischemia and subsequently enhances contractility.⁶⁴

TRANSCATHETER AORTIC VALVE REPLACEMENT

Until recently, patients with severe aortic stenosis who were deemed to be at high surgical risk were referred for balloon valvuloplasty as a palliative option. The procedure consists of balloon inflation across the aortic valve to relieve the stenosis.

Most patients have improved symptoms and a decrease in pressure gradient immediately after the procedure, but the results are not durable, with a high restenosis rate within 6 to 12 months and no decrease in the mortality rate.⁶⁵ (There is some evidence that serial balloon dilation improves survival.⁶⁶)

The procedure has several limitations, including a risk of embolic stroke, myocardial infarction, and, sometimes, perforation of the left ventricle. It is only used in people who do not wish to have surgery or as a bridge to surgical aortic valve replacement in hemodynamically unstable patients.

Advances in transcatheter technologies have made nonsurgical valve replacement a reality that is increasingly available to a broader population of patients. The first percutaneous valve replacement in a human was performed in 2002.⁶⁷ Since then, multiple registries from centers around the world, especially in Europe, have shown that it can be performed in high-risk patients with outcomes very comparable to those of surgical aortic valve replacement as predicted by the STS score and EuroSCORE.^68,69 Procedural success rates have increased from around 80% in the initial experience to over 95% in the most current series.⁷⁰

Results from randomized trials

The long-awaited PARTNER A and B trials have been published.

The PARTNER B trial¹⁷ randomized patients with severe aortic stenosis who were not considered by the STS score to be suitable candidates for surgery to standard therapy (which included balloon valvoplasty in 84%) or transcatheter aortic valve replacement. There was a dramatic 20% absolute improvement in survival at 1 year with transcatheter replacement, with the survival curve continuing to diverge at 1 year. The rate of death from any cause was 30.7% with transcatheter aortic valve replacement vs 50.7% with standard therapy (hazard ratio with transcatheter replacement 0.55; P < .001).

The major concerns about transcatheter aortic valve replacement borne out in the study are procedural complications, namely stroke and vascular events. At 30 days, transcatheter replacement, as compared with standard therapy, was associated with a higher incidence of major stroke (5.0% vs 1.1%, P = .06) and major vascular complications (16.2% vs 1.1%, P < .001).¹⁷

On the other hand, the PARTNER A trial randomized high-risk patients deemed operable by the STS score to either transcatheter or surgical aortic valve replacement. The rate of death at 1 year from any cause was similar in both groups (24.2% vs 26.8%; P = .44), but again at the expense of higher rates of vascular complications (11.0% vs 3.2%, P < .001 at 30 days) and stroke (5.1% vs 2.4%; P = .07 at 1 year) in the transcatheter group. However, the surgical group had higher rates of major bleeding (19.5% vs 9.3%; P < .001) and new-onset atrial fibrillation (16.0% vs 8.6%, P = .06).⁷¹

Transcatheter aortic valve replacement has modernized the way we treat aortic stenosis and without a shred of doubt will become the standard of therapy for severe symptomatic aortic stenosis in patients who are not candidates for surgery. For the high-risk operable patient, the benefit of avoiding a sternotomy should be weighed against the higher risk of stroke and vascular complications with the transcatheter procedure. The availability of smaller delivery systems, better expertise, and better vascular access selection should decrease the rate of complications in the future.

For some patients with aortic stenosis, the choice of management is simple; for others it is less so. Patients who have severe, symptomatic stenosis and who have low surgical risk should undergo aortic valve replacement. But if the stenosis is severe but asymptomatic, or if the patient is at higher surgical risk, or if there seems to be a mismatch in the hemodynamic variables, the situation is more complicated.

See related editorial

Fortunately, we have evidence and guidelines to go on. In this paper we review the indications for surgical and transcatheter aortic valve replacement, focusing on the areas of less certainty.

AN INDOLENT DISEASE, UNTIL IT ISN’T

Aortic stenosis is the most common valvular disease and the third most prevalent form of cardiovascular disease in the Western world, after hypertension and coronary artery disease. It is largely a disease of the elderly; its prevalence increases with age, and it is present in 2% to 7% of patients over age 65.^1,2

At first, its course is indolent, as it progresses slowly over years to decades. However, this is followed by rapid clinical deterioration and a high death rate after symptoms develop.

SURGICAL AORTIC VALVE REPLACEMENT FOR SEVERE SYMPTOMATIC STENOSIS

Classic symptoms of aortic stenosis include angina, heart failure, and syncope. Once symptoms appear, patients with severe aortic stenosis should be promptly referred for surgical aortic valve replacement, as survival is poor unless outflow obstruction is relieved (Figure 1). The onset of symptoms confers a poor prognosis: patients die within an average of 5 years after the onset of angina, 3 years after the onset of syncope, and 2 years after the onset of heart failure symptoms. The overall mortality rate is 75% at 3 years without surgery.^3,4 Furthermore, 8% to 34% of patients with symptoms die suddenly.

Advances in prosthetic-valve design, cardiopulmonary bypass, surgical technique, and anesthesia have steadily improved the outcomes of aortic valve surgery. An analysis of the Society of Thoracic Surgeons (STS) database in 2006 showed that during the previous decade the death rate during isolated aortic valve replacement decreased from 3.4% to 2.6%. For patients under age 70 at the time of surgery, the rate of death was 1.3%, and in those ages 80 to 85, the 30-day mortality rate was less than 5%.⁵

Patients who survive surgery enjoy a near-normal life expectancy: 99% survive at least 5 years, 85% at least 10 years, and 82% at least 15 years.^6,7 Nearly all have improvement in their ejection fraction and heart failure symptoms, and those who had more advanced symptoms before surgery enjoy the most benefit afterward.^8,9

Recommendation. Surgical valve replacement for symptomatic severe aortic stenosis receives a class I recommendation, level of evidence B, in the current guidelines from the American College of Cardiology (ACC) and the American Heart Association (AHA).^10,11 (See Table 1 for an explanation of the classes of recommendations and levels of evidence.)

TWO RISK-ASSESSMENT SCORES

There are two widely used scores for assessing the risk of aortic valve replacement: the European System for Cardiac Operative Risk Evaluation (EuroSCORE) and the STS score. Each has limitations.

The EuroSCORE was developed to predict the risk of dying in the hospital after adult cardiac surgery. It has been shown to predict the short-term and the long-term risk of death after heart valve surgery.¹² Unfortunately, it overestimates the dangers of isolated aortic valve replacement in the patients at highest risk.^13,14

The STS score, a logistic model, reflects more closely the operative and 30-day mortality rates for the patients at highest risk undergoing surgical aortic valve replacement.^15,16 It was used to assess patients for surgical or transcatheter aortic valve replacement in the Placement of Aortic Transcatheter Valves (PARTNER) trial.¹⁷

These risk scores, though not perfect, are helpful as part of an overall estimation of risk that includes functional status, cardiac function, and comorbidities.

OTHER INDICATIONS FOR SURGICAL AORTIC VALVE REPLACEMENT

For patients with severe but asymptomatic aortic stenosis, surgical referral is standard practice in several circumstances.

Asymptomatic severe aortic stenosis with a low ejection fraction

Early studies found significant differences in survival beginning as early as 3 years after valve replacement between those whose preoperative ejection fraction was greater than 50% and those with a lower ejection fraction.⁴ Delaying surgery in these patients may lead to irreversible left ventricular dysfunction and worse survival.

Recommendation. The AHA and the ACC recommend surgical aortic valve replacement for patients who have no symptoms and whose left ventricular ejection fraction is less than 50% (class I indication, level of evidence C).^10,11

Asymptomatic severe aortic stenosis in patients undergoing other cardiac surgery

Recommendation. Even if it is causing no symptoms, a severely stenotic aortic valve ought to be replaced if the ejection fraction is greater than 50% and the patient is undergoing another type of heart surgery, such as coronary artery bypass grafting, aortic surgery, or surgery on other heart valves (class I indication, level of evidence B).^10,11

Asymptomatic moderate aortic stenosis in patients undergoing other cardiac surgery

When patients with a mildly or moderately stenotic aortic valve undergo other types of cardiac surgery, the decision to replace the valve is more difficult. Clinicians have to consider the increase in risk caused by adding aortic valve replacement to the planned surgery compared with the future likelihood of aortic stenosis progressing to a severe symptomatic state and eventually requiring a second cardiac surgery.

We have no evidence from a large prospective randomized controlled trial regarding prophylactic valve replacement at the time of coronary bypass surgery. However, a review of outcomes from the STS database between 1995 and 2000 found that patients under age 70 with a peak aortic gradient greater than “about 28 mm Hg” (correlating with a moderate degree of stenosis) benefited from prophylactic valve replacement at the time of coronary artery bypass surgery.¹⁸

These conclusions were supported by a subsequent retrospective analysis that found a significant survival advantage at 8 years in favor of prophylactic valve replacement at the time of bypass surgery for those with moderate (but not mild) aortic stenosis.¹⁹

Recommendation. The AHA and ACC give a class IIb endorsement, level of evidence B, for aortic valve replacement in patients with asymptomatic moderate aortic stenosis undergoing coronary bypass, valve, or aortic surgery.^10,11

SEVERE ASYMPTOMATIC STENOSIS: WHICH TESTS HELP IN DECIDING?

A patient without symptoms presents a greater challenge than one with symptoms.

If surgery is deferred, the prognosis is usually excellent in such patients. Pellikka et al²⁰ found that patients with severe asymptomatic aortic stenosis who did not undergo surgery had a rate of sudden cardiac death of about 1% per year of follow-up. However, physicians worry about missing the rapid development of symptoms of aortic stenosis in patients who previously had none. Pallikka et al also found that, at 5 years, only 20% of patients had not undergone aortic valve replacement or had not died of cardiovascular causes.²⁰

Many researchers advocate surgical aortic valve replacement for severe asymptomatic aortic stenosis. However, the operative risk is 3% overall and has to be weighed against the 1%-per-year risk of death in patients who do not undergo surgery. Therefore, we need a way to identify a subgroup of patients without symptoms who are at higher risk.

Exercise stress testing

Some patients might subconsciously adapt to aortic stenosis by reducing their physical activity. In these “asymptomatic” patients, exercise stress testing can uncover symptoms in around 40%.²¹

In a group of people with severe asymptomatic aortic stenosis, a positive treadmill test (defined as an abnormal blood pressure response, ST segment changes, symptoms such as limiting dyspnea, chest discomfort, or dizziness on a modified Bruce protocol, or complex ventricular arrhythmias) strongly predicted the onset of symptoms or the need for surgery. At 24 months, only 19% of those who had had a positive exercise test result remained alive, symptom-free, and without valve replacement, compared with 85% of those who had had a negative test result.²²

Subsequent study found that symptoms with exercise were the strongest predictor of the onset of symptoms of aortic stenosis, especially among patients under age 70, in whom the symptoms of fatigue and breathlessness are more specific than in the elderly.²³

Recommendation. Exercise testing is recommended in patients with severe asymptomatic aortic stenosis (class IIa indication, level of evidence B) as a means of identifying those who are likely to develop symptoms or who might benefit from surgery. Surgery for those who have an abnormal exercise stress response receives a class IIb, level of evidence C recommendation from the ACC/AHA and a class IC from the European Society of Cardiology.^24,25

Exercise stress echocardiography to measure change in transvalvular gradient

Emerging data suggest that exercise stress echocardiography may provide incremental prognostic information in patients with severe asymptomatic aortic stenosis. In fact, two studies showed that an exercise-induced increase in the transvalvular gradient of more than 20 mm Hg²⁶ or 18 mm Hg²⁷ predicts future cardiac events. This increase reflects fixed valve stenosis with limited valve compliance.

Other echocardiographic variables

Additional data have shown that severe aortic stenosis (valve area < 0.6 cm²), aortic velocity greater than 4.0 m/s, and severe calcification confer a higher risk of developing symptoms.^28,29

Recommendation. The ACC and AHA say that surgical aortic valve replacement may be considered in patients without symptoms who have a high likelihood of rapid progression of aortic stenosis (ie, who are older or have severe calcification or coronary artery disease) or if surgery might be delayed at the time of symptom onset (class IIb, level of evidence C).

Aortic valve replacement can also be considered for extremely severe aortic stenosis (valve area < 0.6 cm²), mean gradient > 60 mm Hg, and velocity > 5.0 m/s if the operative mortality rate is 1.0% or less (class IIb, level of evidence C).

Brain natriuretic peptide levels

Measuring the brain natriuretic peptide (BNP) level may help if symptoms are unclear; higher levels suggest cardiac decompensation.²⁸

One study showed that BNP levels are higher in patients with symptomatic aortic stenosis than in those with asymptomatic severe disease, and correlate with symptom severity.³⁰ In addition, in two other studies, higher BNP and N-terminal BNP levels were shown to predict disease progression, symptom onset, and poorer event-free survival.^31,32

In severe asymptomatic aortic stenosis, natriuretic peptides may provide important prognostic information beyond clinical and echocardiographic evaluation. Furthermore, in a recent study, Monin et al³³ proposed a risk score that integrates peak aortic jet velocity, BNP level, and sex (women being at higher risk) in predicting who would benefit from early surgery in patients with severe asymptomatic aortic stenosis.³³

SPECIAL CONSIDERATIONS

Low-output, low-gradient aortic stenosis: True severe stenosis vs pseudostenosis

Patients with a low ejection fraction (< 50%) and a high mean transvalvular gradient (> 30 or 40 mm Hg) pose no therapeutic dilemma. They have true afterload mismatch and improve markedly with surgery.³⁴ However, patients with an even lower ejection fraction (< 35% or 40%) and a low mean transvalvular gradient (< 30 or 40 mm Hg) pose more of a problem.

It is hard to tell if these patients have true severe aortic stenosis or pseudostenosis due to primary myocardial dysfunction. In pseudostenosis, the aortic valves are moderately diseased, and leaflet opening is reduced by a failing ventricle. When cardiac output is low, the formulae used to calculate the aortic valve area become less accurate, so that patients with cardiomyopathy who have only mild or moderate aortic stenosis may appear to have severe stenosis.

Patients with pseudostenosis have a high risk of dying during surgical aortic valve replacement, approaching 50%, and benefit more from evidence-based heart failure management.^35,36 In patients with true stenosis, ventricular dysfunction is mainly a result of severe stenosis and should improve after aortic valve replacement.

Dobutamine stress echocardiography can be used in patients with low-flow, low-gradient aortic stenosis to distinguish true severe stenosis from pseudostenosis. Dobutamine, an inotropic drug, increases the stroke volume so that patients with true severe aortic stenosis increase their transvalvular gradient and velocity with no or minimal change in the valve area. Conversely, in patients with pseudostenosis, the increase in stroke volume will open the aortic valve further and cause no or minimal increase in transvalvular gradient and velocity, but will increase the calculated valve area, confirming that aortic stenosis only is mild to moderate.³⁷

Patients with low-flow, low-gradient aortic stenosis are at higher risk during surgical aortic valve replacement. Many studies have reported a 30-day mortality rate between 9% and 18%, although risks vary considerably within this population.^38,39

Contractile reserve. Dobutamine stress echocardiography has also been used to identify patients with severe aortic stenosis who can increase their ejection fraction and stroke volume (Figure 2).^40,41 These patients are said to have “contractile reserve” and do better with surgery than those who lack adequate contractile reserve. Contractile reserve is defined as an increase of more than 20% in stroke volume during low-dose dobutamine infusion.^42,43 In one small nonrandomized study, patients with contractile reserve had a 5% mortality rate at 30 days, compared with 32% in patients with no contractile reserve.^44,45

In fact, patients with no contractile reserve have a high operative mortality rate during aortic valve replacement, but those who survive the operation have improvements in symptoms, functional class, and ejection fraction similar to those in patients who do have contractile reserve.⁴⁶

On the other hand, if patients with no contractile reserve are treated conservatively, they have a much worse prognosis than those managed surgically.⁴⁷ While it is true that patients without contractile reserve did not have a statistically significant difference in mortality rates with aortic valve replacement (P = .07) in a study by Monin et al,⁴⁴ the difference was staggering between the group who underwent aortic valve replacement and the group who received medical treatment alone (hazard ratio = 0.47, 95% confidence interval 0.31–1.05, P = .07). The difference in the mortality rates may not have reached statistical significance because of the study’s small sample size.

A few years later, the same group published a similar paper with a larger study sample, focusing on patients with no contractile reserve. Using 42 propensity-matched patients, they found a statistically significantly higher 5-year survival rate in patients with no contractile reserve who underwent aortic valve replacement than in similar patients who received medical management (65% ± 11% vs 11 ± 7%, P = .019).⁴⁷

Hence, surgery may be a better option than medical treatment for this select high-risk group despite the higher operative mortality risk. Transcatheter aortic valve implantation may also offer an interesting alternative to surgical aortic valve replacement in this particular subset of patients.⁴⁸

Low-gradient ‘severe’ aortic stenosis with preserved ejection fraction or ‘paradoxically low-flow aortic stenosis’

Low-gradient “severe” aortic stenosis with a preserved left ventricular ejection fraction is a recently recognized clinical entity in patients with severe aortic stenosis who present with a lower-than-expected transvalvular gradient on the basis of generally accepted values.⁴⁹ (A patient with severe aortic stenosis and preserved ejection fraction is expected to generate a mean transaortic gradient greater than 40 mm Hg.²⁴) This situation remains incompletely understood but has been shown in retrospective studies to foretell a poor prognosis.^50–52

This subgroup of patients has pronounced left ventricular concentric remodeling with a small left ventricular cavity, impaired left ventricular filling, and reduced systolic longitudinal myocardial shortening.⁴⁴

Herrmann et al⁵³ provided more insight into the pathophysiology by showing that patients with this condition exhibit more pronounced myocardial fibrosis on myocardial biopsy and more pronounced late subendocardial enhancement on magnetic resonance imaging. These patients also displayed a significant decrease in mitral ring displacement and systolic strain. These abnormalities result in a low stroke volume despite a preserved ejection fraction and consequently a lower transvalvular gradient (< 40 mm Hg).

This disease pattern, in which the low gradient is interpreted as mild to moderate aortic stenosis, may lead to underestimation of stenosis severity and, thus, to inappropriate delay of aortic valve replacement.

However, other conditions can cause this hemodynamic situation with a lower-than-expected gradient. It can arise from a small left ventricle that correlates with a small body size, yielding a lower-than-normal stroke volume, measurement errors in determining stroke volume and valve area by Doppler echocardiography, systemic hypertension (which can influence estimation of the gradient by Doppler echocardiography), and inconsistency in the definition of severe aortic stenosis in the current guidelines relating to cutoffs of valve area in relation to those of jet velocity and gradient.⁵⁴

This subgroup of patients seems to be at a more advanced stage and has a poorer prognosis if treated medically rather than surgically. When symptomatic, low-gradient severe aortic stenosis should be treated surgically, with one study showing excellent outcomes with aortic valve replacement.⁵⁰

However, a recent study by Jander et al⁵⁵ showed that patients with low-gradient severe aortic stenosis and normal ejection fraction have outcomes similar to those in patients with moderate aortic stenosis, suggesting a strategy of medical therapy and close monitoring.⁵⁵ Of note, the subset of patients reported in this substudy of the Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) trial did not really fit the pattern of low-gradient severe aortic stenosis described by Hachicha et al⁵⁰ and other groups.^51,56 These patients had aortic valve areas in the severe range but mean transaortic gradients in the moderate range, and in light of the other echocardiographic findings in these patients, the area-gradient discordances were predominantly due to small body surface area and measurement errors. These patients indeed had near-normal left ventricular size, no left ventricular hypertrophy, and no evidence of concentric remodeling.

Finally, the findings of the study by Jander et al⁵⁵ are discordant with those of another substudy of the SEAS trial,⁵⁷ which reported that paradoxical low-flow aortic stenosis occurred in about 7% of the cohort (compared with 52% in the study by Jander et al⁵⁵) and was associated with more pronounced concentric remodeling and more severe impairment of myocardial function.

Whether intervention in patients with low-gradient severe aortic stenosis and valve area less than 1.0 cm² improves outcomes remains to be confirmed and reproduced in future prospective studies.

The risks of cardiac surgery increase with age. Older patients may be more deconditioned and have more comorbidities than younger patients, placing them at greater risk of a poor outcome.

Several retrospective studies of valve replacement in octogenarians have found that operative mortality rates range from 5.7% to 9% during isolated aortic valve replacement.^58–60 Note that, using the STS score, the operative mortality risk increases only from 1.2% in a 70-year-old man with no comorbidities to 1.8% in an 80-year-old man undergoing aortic valve replacement plus coronary artery bypass grafting.⁶¹

As in younger patients, valve replacement results in a significant survival benefit and symptomatic improvement. Yet up to 30% of patients with severe aortic stenosis are not referred for surgery because surgery is believed to be too risky.⁶² The conditions most frequently cited by physicians when declining to refer patients for surgery include a low ejection fraction, advanced age, and advanced comorbidities. None of these is an absolute contraindication to surgery.

A recent retrospective study of 443 elderly patients (mean age 79.5) showed that those with left ventricular concentric remodeling, lower stroke volume, elevated left ventricular filling pressures, and mildly elevated pulmonary artery pressures have a very bad prognosis, with a mortality rate of 50.5% at 3.3 ± 2.7 years.⁶³

Despite the higher operative mortality risk, these patients face a dismal prognosis when treated medically and should be referred to a cardiologist or cardiothoracic surgeon for an assessment of their operative risk and, potentially, for referral for catheter-based valve replacement.

Acutely ill patients

In critically ill patients with aortic stenosis and cardiogenic shock, the use of intravenous sodium nitroprusside increases cardiac output and decreases pulmonary artery wedge pressure, allowing patients to transition to surgery or vasodilator therapy. The mechanism seems to be an increase in myocardial contractility rather than a decrease in peripheral resistance. The reduction in filling pressure and concurrent increase in coronary blood flow relieves ischemia and subsequently enhances contractility.⁶⁴

TRANSCATHETER AORTIC VALVE REPLACEMENT

Until recently, patients with severe aortic stenosis who were deemed to be at high surgical risk were referred for balloon valvuloplasty as a palliative option. The procedure consists of balloon inflation across the aortic valve to relieve the stenosis.

Most patients have improved symptoms and a decrease in pressure gradient immediately after the procedure, but the results are not durable, with a high restenosis rate within 6 to 12 months and no decrease in the mortality rate.⁶⁵ (There is some evidence that serial balloon dilation improves survival.⁶⁶)

The procedure has several limitations, including a risk of embolic stroke, myocardial infarction, and, sometimes, perforation of the left ventricle. It is only used in people who do not wish to have surgery or as a bridge to surgical aortic valve replacement in hemodynamically unstable patients.

Advances in transcatheter technologies have made nonsurgical valve replacement a reality that is increasingly available to a broader population of patients. The first percutaneous valve replacement in a human was performed in 2002.⁶⁷ Since then, multiple registries from centers around the world, especially in Europe, have shown that it can be performed in high-risk patients with outcomes very comparable to those of surgical aortic valve replacement as predicted by the STS score and EuroSCORE.^68,69 Procedural success rates have increased from around 80% in the initial experience to over 95% in the most current series.⁷⁰

Results from randomized trials

The long-awaited PARTNER A and B trials have been published.

The PARTNER B trial¹⁷ randomized patients with severe aortic stenosis who were not considered by the STS score to be suitable candidates for surgery to standard therapy (which included balloon valvoplasty in 84%) or transcatheter aortic valve replacement. There was a dramatic 20% absolute improvement in survival at 1 year with transcatheter replacement, with the survival curve continuing to diverge at 1 year. The rate of death from any cause was 30.7% with transcatheter aortic valve replacement vs 50.7% with standard therapy (hazard ratio with transcatheter replacement 0.55; P < .001).

The major concerns about transcatheter aortic valve replacement borne out in the study are procedural complications, namely stroke and vascular events. At 30 days, transcatheter replacement, as compared with standard therapy, was associated with a higher incidence of major stroke (5.0% vs 1.1%, P = .06) and major vascular complications (16.2% vs 1.1%, P < .001).¹⁷

On the other hand, the PARTNER A trial randomized high-risk patients deemed operable by the STS score to either transcatheter or surgical aortic valve replacement. The rate of death at 1 year from any cause was similar in both groups (24.2% vs 26.8%; P = .44), but again at the expense of higher rates of vascular complications (11.0% vs 3.2%, P < .001 at 30 days) and stroke (5.1% vs 2.4%; P = .07 at 1 year) in the transcatheter group. However, the surgical group had higher rates of major bleeding (19.5% vs 9.3%; P < .001) and new-onset atrial fibrillation (16.0% vs 8.6%, P = .06).⁷¹

Transcatheter aortic valve replacement has modernized the way we treat aortic stenosis and without a shred of doubt will become the standard of therapy for severe symptomatic aortic stenosis in patients who are not candidates for surgery. For the high-risk operable patient, the benefit of avoiding a sternotomy should be weighed against the higher risk of stroke and vascular complications with the transcatheter procedure. The availability of smaller delivery systems, better expertise, and better vascular access selection should decrease the rate of complications in the future.

For some patients with aortic stenosis, the choice of management is simple; for others it is less so. Patients who have severe, symptomatic stenosis and who have low surgical risk should undergo aortic valve replacement. But if the stenosis is severe but asymptomatic, or if the patient is at higher surgical risk, or if there seems to be a mismatch in the hemodynamic variables, the situation is more complicated.

See related editorial

Fortunately, we have evidence and guidelines to go on. In this paper we review the indications for surgical and transcatheter aortic valve replacement, focusing on the areas of less certainty.

AN INDOLENT DISEASE, UNTIL IT ISN’T

Aortic stenosis is the most common valvular disease and the third most prevalent form of cardiovascular disease in the Western world, after hypertension and coronary artery disease. It is largely a disease of the elderly; its prevalence increases with age, and it is present in 2% to 7% of patients over age 65.^1,2

At first, its course is indolent, as it progresses slowly over years to decades. However, this is followed by rapid clinical deterioration and a high death rate after symptoms develop.

SURGICAL AORTIC VALVE REPLACEMENT FOR SEVERE SYMPTOMATIC STENOSIS

Classic symptoms of aortic stenosis include angina, heart failure, and syncope. Once symptoms appear, patients with severe aortic stenosis should be promptly referred for surgical aortic valve replacement, as survival is poor unless outflow obstruction is relieved (Figure 1). The onset of symptoms confers a poor prognosis: patients die within an average of 5 years after the onset of angina, 3 years after the onset of syncope, and 2 years after the onset of heart failure symptoms. The overall mortality rate is 75% at 3 years without surgery.^3,4 Furthermore, 8% to 34% of patients with symptoms die suddenly.

Advances in prosthetic-valve design, cardiopulmonary bypass, surgical technique, and anesthesia have steadily improved the outcomes of aortic valve surgery. An analysis of the Society of Thoracic Surgeons (STS) database in 2006 showed that during the previous decade the death rate during isolated aortic valve replacement decreased from 3.4% to 2.6%. For patients under age 70 at the time of surgery, the rate of death was 1.3%, and in those ages 80 to 85, the 30-day mortality rate was less than 5%.⁵

Patients who survive surgery enjoy a near-normal life expectancy: 99% survive at least 5 years, 85% at least 10 years, and 82% at least 15 years.^6,7 Nearly all have improvement in their ejection fraction and heart failure symptoms, and those who had more advanced symptoms before surgery enjoy the most benefit afterward.^8,9

Recommendation. Surgical valve replacement for symptomatic severe aortic stenosis receives a class I recommendation, level of evidence B, in the current guidelines from the American College of Cardiology (ACC) and the American Heart Association (AHA).^10,11 (See Table 1 for an explanation of the classes of recommendations and levels of evidence.)

TWO RISK-ASSESSMENT SCORES

There are two widely used scores for assessing the risk of aortic valve replacement: the European System for Cardiac Operative Risk Evaluation (EuroSCORE) and the STS score. Each has limitations.

The EuroSCORE was developed to predict the risk of dying in the hospital after adult cardiac surgery. It has been shown to predict the short-term and the long-term risk of death after heart valve surgery.¹² Unfortunately, it overestimates the dangers of isolated aortic valve replacement in the patients at highest risk.^13,14

The STS score, a logistic model, reflects more closely the operative and 30-day mortality rates for the patients at highest risk undergoing surgical aortic valve replacement.^15,16 It was used to assess patients for surgical or transcatheter aortic valve replacement in the Placement of Aortic Transcatheter Valves (PARTNER) trial.¹⁷

These risk scores, though not perfect, are helpful as part of an overall estimation of risk that includes functional status, cardiac function, and comorbidities.

OTHER INDICATIONS FOR SURGICAL AORTIC VALVE REPLACEMENT

For patients with severe but asymptomatic aortic stenosis, surgical referral is standard practice in several circumstances.

Asymptomatic severe aortic stenosis with a low ejection fraction

Early studies found significant differences in survival beginning as early as 3 years after valve replacement between those whose preoperative ejection fraction was greater than 50% and those with a lower ejection fraction.⁴ Delaying surgery in these patients may lead to irreversible left ventricular dysfunction and worse survival.

Recommendation. The AHA and the ACC recommend surgical aortic valve replacement for patients who have no symptoms and whose left ventricular ejection fraction is less than 50% (class I indication, level of evidence C).^10,11

Asymptomatic severe aortic stenosis in patients undergoing other cardiac surgery

Recommendation. Even if it is causing no symptoms, a severely stenotic aortic valve ought to be replaced if the ejection fraction is greater than 50% and the patient is undergoing another type of heart surgery, such as coronary artery bypass grafting, aortic surgery, or surgery on other heart valves (class I indication, level of evidence B).^10,11

Asymptomatic moderate aortic stenosis in patients undergoing other cardiac surgery

When patients with a mildly or moderately stenotic aortic valve undergo other types of cardiac surgery, the decision to replace the valve is more difficult. Clinicians have to consider the increase in risk caused by adding aortic valve replacement to the planned surgery compared with the future likelihood of aortic stenosis progressing to a severe symptomatic state and eventually requiring a second cardiac surgery.

We have no evidence from a large prospective randomized controlled trial regarding prophylactic valve replacement at the time of coronary bypass surgery. However, a review of outcomes from the STS database between 1995 and 2000 found that patients under age 70 with a peak aortic gradient greater than “about 28 mm Hg” (correlating with a moderate degree of stenosis) benefited from prophylactic valve replacement at the time of coronary artery bypass surgery.¹⁸

These conclusions were supported by a subsequent retrospective analysis that found a significant survival advantage at 8 years in favor of prophylactic valve replacement at the time of bypass surgery for those with moderate (but not mild) aortic stenosis.¹⁹

Recommendation. The AHA and ACC give a class IIb endorsement, level of evidence B, for aortic valve replacement in patients with asymptomatic moderate aortic stenosis undergoing coronary bypass, valve, or aortic surgery.^10,11

SEVERE ASYMPTOMATIC STENOSIS: WHICH TESTS HELP IN DECIDING?

A patient without symptoms presents a greater challenge than one with symptoms.

If surgery is deferred, the prognosis is usually excellent in such patients. Pellikka et al²⁰ found that patients with severe asymptomatic aortic stenosis who did not undergo surgery had a rate of sudden cardiac death of about 1% per year of follow-up. However, physicians worry about missing the rapid development of symptoms of aortic stenosis in patients who previously had none. Pallikka et al also found that, at 5 years, only 20% of patients had not undergone aortic valve replacement or had not died of cardiovascular causes.²⁰

Many researchers advocate surgical aortic valve replacement for severe asymptomatic aortic stenosis. However, the operative risk is 3% overall and has to be weighed against the 1%-per-year risk of death in patients who do not undergo surgery. Therefore, we need a way to identify a subgroup of patients without symptoms who are at higher risk.

Exercise stress testing

Some patients might subconsciously adapt to aortic stenosis by reducing their physical activity. In these “asymptomatic” patients, exercise stress testing can uncover symptoms in around 40%.²¹

In a group of people with severe asymptomatic aortic stenosis, a positive treadmill test (defined as an abnormal blood pressure response, ST segment changes, symptoms such as limiting dyspnea, chest discomfort, or dizziness on a modified Bruce protocol, or complex ventricular arrhythmias) strongly predicted the onset of symptoms or the need for surgery. At 24 months, only 19% of those who had had a positive exercise test result remained alive, symptom-free, and without valve replacement, compared with 85% of those who had had a negative test result.²²

Subsequent study found that symptoms with exercise were the strongest predictor of the onset of symptoms of aortic stenosis, especially among patients under age 70, in whom the symptoms of fatigue and breathlessness are more specific than in the elderly.²³

Recommendation. Exercise testing is recommended in patients with severe asymptomatic aortic stenosis (class IIa indication, level of evidence B) as a means of identifying those who are likely to develop symptoms or who might benefit from surgery. Surgery for those who have an abnormal exercise stress response receives a class IIb, level of evidence C recommendation from the ACC/AHA and a class IC from the European Society of Cardiology.^24,25

Exercise stress echocardiography to measure change in transvalvular gradient

Emerging data suggest that exercise stress echocardiography may provide incremental prognostic information in patients with severe asymptomatic aortic stenosis. In fact, two studies showed that an exercise-induced increase in the transvalvular gradient of more than 20 mm Hg²⁶ or 18 mm Hg²⁷ predicts future cardiac events. This increase reflects fixed valve stenosis with limited valve compliance.

Other echocardiographic variables

Additional data have shown that severe aortic stenosis (valve area < 0.6 cm²), aortic velocity greater than 4.0 m/s, and severe calcification confer a higher risk of developing symptoms.^28,29

Recommendation. The ACC and AHA say that surgical aortic valve replacement may be considered in patients without symptoms who have a high likelihood of rapid progression of aortic stenosis (ie, who are older or have severe calcification or coronary artery disease) or if surgery might be delayed at the time of symptom onset (class IIb, level of evidence C).

Aortic valve replacement can also be considered for extremely severe aortic stenosis (valve area < 0.6 cm²), mean gradient > 60 mm Hg, and velocity > 5.0 m/s if the operative mortality rate is 1.0% or less (class IIb, level of evidence C).

Brain natriuretic peptide levels

Measuring the brain natriuretic peptide (BNP) level may help if symptoms are unclear; higher levels suggest cardiac decompensation.²⁸

One study showed that BNP levels are higher in patients with symptomatic aortic stenosis than in those with asymptomatic severe disease, and correlate with symptom severity.³⁰ In addition, in two other studies, higher BNP and N-terminal BNP levels were shown to predict disease progression, symptom onset, and poorer event-free survival.^31,32

In severe asymptomatic aortic stenosis, natriuretic peptides may provide important prognostic information beyond clinical and echocardiographic evaluation. Furthermore, in a recent study, Monin et al³³ proposed a risk score that integrates peak aortic jet velocity, BNP level, and sex (women being at higher risk) in predicting who would benefit from early surgery in patients with severe asymptomatic aortic stenosis.³³

SPECIAL CONSIDERATIONS

Low-output, low-gradient aortic stenosis: True severe stenosis vs pseudostenosis

Patients with a low ejection fraction (< 50%) and a high mean transvalvular gradient (> 30 or 40 mm Hg) pose no therapeutic dilemma. They have true afterload mismatch and improve markedly with surgery.³⁴ However, patients with an even lower ejection fraction (< 35% or 40%) and a low mean transvalvular gradient (< 30 or 40 mm Hg) pose more of a problem.

It is hard to tell if these patients have true severe aortic stenosis or pseudostenosis due to primary myocardial dysfunction. In pseudostenosis, the aortic valves are moderately diseased, and leaflet opening is reduced by a failing ventricle. When cardiac output is low, the formulae used to calculate the aortic valve area become less accurate, so that patients with cardiomyopathy who have only mild or moderate aortic stenosis may appear to have severe stenosis.

Patients with pseudostenosis have a high risk of dying during surgical aortic valve replacement, approaching 50%, and benefit more from evidence-based heart failure management.^35,36 In patients with true stenosis, ventricular dysfunction is mainly a result of severe stenosis and should improve after aortic valve replacement.

Dobutamine stress echocardiography can be used in patients with low-flow, low-gradient aortic stenosis to distinguish true severe stenosis from pseudostenosis. Dobutamine, an inotropic drug, increases the stroke volume so that patients with true severe aortic stenosis increase their transvalvular gradient and velocity with no or minimal change in the valve area. Conversely, in patients with pseudostenosis, the increase in stroke volume will open the aortic valve further and cause no or minimal increase in transvalvular gradient and velocity, but will increase the calculated valve area, confirming that aortic stenosis only is mild to moderate.³⁷

Patients with low-flow, low-gradient aortic stenosis are at higher risk during surgical aortic valve replacement. Many studies have reported a 30-day mortality rate between 9% and 18%, although risks vary considerably within this population.^38,39

Contractile reserve. Dobutamine stress echocardiography has also been used to identify patients with severe aortic stenosis who can increase their ejection fraction and stroke volume (Figure 2).^40,41 These patients are said to have “contractile reserve” and do better with surgery than those who lack adequate contractile reserve. Contractile reserve is defined as an increase of more than 20% in stroke volume during low-dose dobutamine infusion.^42,43 In one small nonrandomized study, patients with contractile reserve had a 5% mortality rate at 30 days, compared with 32% in patients with no contractile reserve.^44,45

In fact, patients with no contractile reserve have a high operative mortality rate during aortic valve replacement, but those who survive the operation have improvements in symptoms, functional class, and ejection fraction similar to those in patients who do have contractile reserve.⁴⁶

On the other hand, if patients with no contractile reserve are treated conservatively, they have a much worse prognosis than those managed surgically.⁴⁷ While it is true that patients without contractile reserve did not have a statistically significant difference in mortality rates with aortic valve replacement (P = .07) in a study by Monin et al,⁴⁴ the difference was staggering between the group who underwent aortic valve replacement and the group who received medical treatment alone (hazard ratio = 0.47, 95% confidence interval 0.31–1.05, P = .07). The difference in the mortality rates may not have reached statistical significance because of the study’s small sample size.

A few years later, the same group published a similar paper with a larger study sample, focusing on patients with no contractile reserve. Using 42 propensity-matched patients, they found a statistically significantly higher 5-year survival rate in patients with no contractile reserve who underwent aortic valve replacement than in similar patients who received medical management (65% ± 11% vs 11 ± 7%, P = .019).⁴⁷

Hence, surgery may be a better option than medical treatment for this select high-risk group despite the higher operative mortality risk. Transcatheter aortic valve implantation may also offer an interesting alternative to surgical aortic valve replacement in this particular subset of patients.⁴⁸

Low-gradient ‘severe’ aortic stenosis with preserved ejection fraction or ‘paradoxically low-flow aortic stenosis’

Low-gradient “severe” aortic stenosis with a preserved left ventricular ejection fraction is a recently recognized clinical entity in patients with severe aortic stenosis who present with a lower-than-expected transvalvular gradient on the basis of generally accepted values.⁴⁹ (A patient with severe aortic stenosis and preserved ejection fraction is expected to generate a mean transaortic gradient greater than 40 mm Hg.²⁴) This situation remains incompletely understood but has been shown in retrospective studies to foretell a poor prognosis.^50–52

This subgroup of patients has pronounced left ventricular concentric remodeling with a small left ventricular cavity, impaired left ventricular filling, and reduced systolic longitudinal myocardial shortening.⁴⁴

Herrmann et al⁵³ provided more insight into the pathophysiology by showing that patients with this condition exhibit more pronounced myocardial fibrosis on myocardial biopsy and more pronounced late subendocardial enhancement on magnetic resonance imaging. These patients also displayed a significant decrease in mitral ring displacement and systolic strain. These abnormalities result in a low stroke volume despite a preserved ejection fraction and consequently a lower transvalvular gradient (< 40 mm Hg).

This disease pattern, in which the low gradient is interpreted as mild to moderate aortic stenosis, may lead to underestimation of stenosis severity and, thus, to inappropriate delay of aortic valve replacement.

However, other conditions can cause this hemodynamic situation with a lower-than-expected gradient. It can arise from a small left ventricle that correlates with a small body size, yielding a lower-than-normal stroke volume, measurement errors in determining stroke volume and valve area by Doppler echocardiography, systemic hypertension (which can influence estimation of the gradient by Doppler echocardiography), and inconsistency in the definition of severe aortic stenosis in the current guidelines relating to cutoffs of valve area in relation to those of jet velocity and gradient.⁵⁴

This subgroup of patients seems to be at a more advanced stage and has a poorer prognosis if treated medically rather than surgically. When symptomatic, low-gradient severe aortic stenosis should be treated surgically, with one study showing excellent outcomes with aortic valve replacement.⁵⁰

However, a recent study by Jander et al⁵⁵ showed that patients with low-gradient severe aortic stenosis and normal ejection fraction have outcomes similar to those in patients with moderate aortic stenosis, suggesting a strategy of medical therapy and close monitoring.⁵⁵ Of note, the subset of patients reported in this substudy of the Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) trial did not really fit the pattern of low-gradient severe aortic stenosis described by Hachicha et al⁵⁰ and other groups.^51,56 These patients had aortic valve areas in the severe range but mean transaortic gradients in the moderate range, and in light of the other echocardiographic findings in these patients, the area-gradient discordances were predominantly due to small body surface area and measurement errors. These patients indeed had near-normal left ventricular size, no left ventricular hypertrophy, and no evidence of concentric remodeling.

Finally, the findings of the study by Jander et al⁵⁵ are discordant with those of another substudy of the SEAS trial,⁵⁷ which reported that paradoxical low-flow aortic stenosis occurred in about 7% of the cohort (compared with 52% in the study by Jander et al⁵⁵) and was associated with more pronounced concentric remodeling and more severe impairment of myocardial function.

Whether intervention in patients with low-gradient severe aortic stenosis and valve area less than 1.0 cm² improves outcomes remains to be confirmed and reproduced in future prospective studies.

The risks of cardiac surgery increase with age. Older patients may be more deconditioned and have more comorbidities than younger patients, placing them at greater risk of a poor outcome.

Several retrospective studies of valve replacement in octogenarians have found that operative mortality rates range from 5.7% to 9% during isolated aortic valve replacement.^58–60 Note that, using the STS score, the operative mortality risk increases only from 1.2% in a 70-year-old man with no comorbidities to 1.8% in an 80-year-old man undergoing aortic valve replacement plus coronary artery bypass grafting.⁶¹

As in younger patients, valve replacement results in a significant survival benefit and symptomatic improvement. Yet up to 30% of patients with severe aortic stenosis are not referred for surgery because surgery is believed to be too risky.⁶² The conditions most frequently cited by physicians when declining to refer patients for surgery include a low ejection fraction, advanced age, and advanced comorbidities. None of these is an absolute contraindication to surgery.

A recent retrospective study of 443 elderly patients (mean age 79.5) showed that those with left ventricular concentric remodeling, lower stroke volume, elevated left ventricular filling pressures, and mildly elevated pulmonary artery pressures have a very bad prognosis, with a mortality rate of 50.5% at 3.3 ± 2.7 years.⁶³

Despite the higher operative mortality risk, these patients face a dismal prognosis when treated medically and should be referred to a cardiologist or cardiothoracic surgeon for an assessment of their operative risk and, potentially, for referral for catheter-based valve replacement.

Acutely ill patients

In critically ill patients with aortic stenosis and cardiogenic shock, the use of intravenous sodium nitroprusside increases cardiac output and decreases pulmonary artery wedge pressure, allowing patients to transition to surgery or vasodilator therapy. The mechanism seems to be an increase in myocardial contractility rather than a decrease in peripheral resistance. The reduction in filling pressure and concurrent increase in coronary blood flow relieves ischemia and subsequently enhances contractility.⁶⁴

TRANSCATHETER AORTIC VALVE REPLACEMENT

Until recently, patients with severe aortic stenosis who were deemed to be at high surgical risk were referred for balloon valvuloplasty as a palliative option. The procedure consists of balloon inflation across the aortic valve to relieve the stenosis.

Most patients have improved symptoms and a decrease in pressure gradient immediately after the procedure, but the results are not durable, with a high restenosis rate within 6 to 12 months and no decrease in the mortality rate.⁶⁵ (There is some evidence that serial balloon dilation improves survival.⁶⁶)

The procedure has several limitations, including a risk of embolic stroke, myocardial infarction, and, sometimes, perforation of the left ventricle. It is only used in people who do not wish to have surgery or as a bridge to surgical aortic valve replacement in hemodynamically unstable patients.

Advances in transcatheter technologies have made nonsurgical valve replacement a reality that is increasingly available to a broader population of patients. The first percutaneous valve replacement in a human was performed in 2002.⁶⁷ Since then, multiple registries from centers around the world, especially in Europe, have shown that it can be performed in high-risk patients with outcomes very comparable to those of surgical aortic valve replacement as predicted by the STS score and EuroSCORE.^68,69 Procedural success rates have increased from around 80% in the initial experience to over 95% in the most current series.⁷⁰

Results from randomized trials

The long-awaited PARTNER A and B trials have been published.

The PARTNER B trial¹⁷ randomized patients with severe aortic stenosis who were not considered by the STS score to be suitable candidates for surgery to standard therapy (which included balloon valvoplasty in 84%) or transcatheter aortic valve replacement. There was a dramatic 20% absolute improvement in survival at 1 year with transcatheter replacement, with the survival curve continuing to diverge at 1 year. The rate of death from any cause was 30.7% with transcatheter aortic valve replacement vs 50.7% with standard therapy (hazard ratio with transcatheter replacement 0.55; P < .001).

The major concerns about transcatheter aortic valve replacement borne out in the study are procedural complications, namely stroke and vascular events. At 30 days, transcatheter replacement, as compared with standard therapy, was associated with a higher incidence of major stroke (5.0% vs 1.1%, P = .06) and major vascular complications (16.2% vs 1.1%, P < .001).¹⁷

On the other hand, the PARTNER A trial randomized high-risk patients deemed operable by the STS score to either transcatheter or surgical aortic valve replacement. The rate of death at 1 year from any cause was similar in both groups (24.2% vs 26.8%; P = .44), but again at the expense of higher rates of vascular complications (11.0% vs 3.2%, P < .001 at 30 days) and stroke (5.1% vs 2.4%; P = .07 at 1 year) in the transcatheter group. However, the surgical group had higher rates of major bleeding (19.5% vs 9.3%; P < .001) and new-onset atrial fibrillation (16.0% vs 8.6%, P = .06).⁷¹

Transcatheter aortic valve replacement has modernized the way we treat aortic stenosis and without a shred of doubt will become the standard of therapy for severe symptomatic aortic stenosis in patients who are not candidates for surgery. For the high-risk operable patient, the benefit of avoiding a sternotomy should be weighed against the higher risk of stroke and vascular complications with the transcatheter procedure. The availability of smaller delivery systems, better expertise, and better vascular access selection should decrease the rate of complications in the future.

Why do primary care physicians in nontropical parts of the world need to be on the lookout for tropical diseases such as dengue?

First, more people are traveling than ever before, and second, more people are traveling to parts of the world where dengue and other tropical diseases are endemic. Thus, dengue should now be included in the differential diagnosis of fever in anyone returning from travel to a part of the world where dengue is endemic (Table 1).¹

The number of cases of dengue in returning US travelers has been increasing steadily over the past 25 years.^2,3 Dengue is a more common cause of febrile illness than malaria in US travelers returning from all tropical and subtropical regions except Africa.⁴

Moreover, dengue may be gaining a permanent foothold in the United States, and in areas of the country where mosquitoes that can transmit the virus are found, primary care physicians are the first line of defense in public health. Specifically, to prevent the virus from becoming established locally, primary care physicians need to quickly identify and report cases to public health authorities, who can promptly follow up and initiate prevention measures.⁵

Underscoring the importance of dengue, this infection was added in 2009 to the list of nationally notifiable infectious diseases in the United States.⁶

A COMMON INFECTION WORLDWIDE

Dengue causes up to 100 million new infections, 500,000 hospitalizations, and 25,000 deaths every year in the 2.5 billion people who live in subtropical and tropical areas of the world.⁷ It is transmitted by mosquitoes of the genus Aedes (which, unlike most other mosquitoes, often bite in the daytime), and it is the most common arboviral infection (ie, transmitted mainly by arthropods) worldwide.

The dengue virus belongs to the family of flaviviruses, which includes West Nile virus, St. Louis encephalitis, and yellow fever. It has four closely related but antigenically distinct serotypes, designated DENV-1, DENV-2, DENV-3, and DENV-4. Infection with one serotype induces lifetime homotypic immunity but only short-lived heterotypic immunity to the other dengue serotypes.⁸ Hence, a person can be infected over time with each of the four serotypes. The first infection is termed the primary infection; subsequent infection with any of the remaining three serotypes is termed a secondary infection.

Source: US Centers for Disease Control and Prevention.

Dengue virus transmission has been expanding since the end of World War II in Asia and since the 1980s in the Americas following the end of many regional vector-control programs.⁹ Although dengue is known to occur in tropical Africa, its epidemiology is less well defined on that continent (Figure 1).¹⁰

Explosive epidemics of dengue occur when there are enough mosquitoes and a susceptible population across a broad age range, ie, both children and adults.^11,12 Transmission can be halted with vigorous vector-control programs, or it slows and stops when the pool of susceptible people is exhausted.^13–15

On the other hand, hyperendemic transmission occurs in areas in which multiple virus serotypes continuously circulate in a large pool of susceptible people. In these areas, dengue seroprevalence increases with age, and most adults are immune.

Anyone of any age who travels from a nonendemic area to an epidemic or hyperendemic area is at risk of infection.¹⁶

DENGUE IN THE UNITED STATES

Reported cases of dengue in South and Central America, the Caribbean, and Mexico, common destinations for US travelers, have increased more than fourfold since the 1980s. There were a total of approximately 1 million cases in the 10-year period ending in 1989, compared with more than 4.5 million in the 8-year period from 2000 to 2007.¹¹

The geographic proximity of these areas to the continental United States and the large numbers of US residents travelling to these areas have raised concern that dengue could emerge in the continental United States in areas where potential vectors exist (see below).¹⁷ Furthermore, several US territories and former territories where tourism is an economic mainstay, including the Commonwealth of Puerto Rico,¹⁸ the US Virgin Islands,¹⁹ American Samoa, and other smaller Pacific island jurisdictions such as Palau,²⁰ have reported dengue virus circulation.

Arbonet database, maintained by the US Centers for Disease Control and Prevention.

Adding to the concern that dengue could gain a persistent foothold in the United States, competent dengue vectors are found here. Two mosquito vectors, Aedes aegypti21 and Aedes albopictus,^22,23 live in some areas of the southwestern and southeastern United States (Figure 2), with Aedes aegypti being the more competent transmitter. Both vectors may be abundant in warmer months. This raises a concern that a returning dengue-infected traveler could initiate an outbreak of autochthonous transmission.²⁴

Notably, endemic dengue transmission has occurred in the past in the United States, and the virus is circulating here again at low levels. From 1946 to 1980, no cases of dengue were acquired in the continental United States. However, since 1980 there have been seven outbreaks of laboratory-confirmed, locally acquired dengue along the Texas-Mexico border.^25–27 More recently, local transmission emerged in Key West, Florida,^28,29 the first outbreak since 1945 of dengue in the continental United States not to occur near the Texas-Mexico border. And in early 2011, nonsustained but locally acquired transmission was confirmed in Hawaii after a transmission-free decade.³⁰

THE CLINICAL SPECTRUM OF DENGUE VIRUS INFECTION

Most primary and secondary dengue infections are asymptomatic.⁸ The common forms of clinically apparent disease include self-limited, undifferentiated fever and classic dengue fever (Figure 3). Severe disease, manifesting as either dengue hemorrhagic fever or dengue shock syndrome, is a rare outcome of dengue virus infection, estimated to occur in 1% of cases worldwide.³¹ However, the true proportion of severe infection among all dengue cases seen in travelers is difficult to assess reliably.

Asymptomatic infection

Clinical dengue disease is relatively uncommon, as between 60% and 80% of infections are asymptomatic, particularly in children and adults who never have been infected before.³²

In the recent outbreak in Key West, 28 symptomatic cases of locally acquired dengue were detected. The US Centers for Disease Control and Prevention (CDC) conducted a serologic survey of 240 healthy, randomly selected residents of Key West and found evidence of recent infection in 5.4% of those tested.²⁸ Based on this finding, the CDC estimated that 1,000 people had been infected, of whom more than 90% had no symptoms. However, no attempt was made to differentiate primary from secondary infection in this serosurvey. (Previous infection with one dengue serotype places some individuals at risk for severe dengue if infected with a different serotype in the future.)

Uncomplicated dengue infection

Undifferentiated fever³³ and classic dengue fever are the most common manifestations of clinical dengue infection. Also known as breakbone fever, classic dengue fever is a fever-arthralgia-rash syndrome.¹

The onset is acute, with a high fever (though rarely greater than 40.5°C [104.9°F]) 3 to 14 days (usually 5 to 9 days) after the patient was bitten by an Aedes mosquito. Therefore, a febrile illness beginning more than 2 weeks after returning from travel to an endemic area is unlikely to be dengue, and another diagnosis should be sought.

A prodrome of headache, backache, fatigue, chills, anorexia, and occasionally a rash may precede the onset of fever by about 12 hours.

With fever comes a severe frontal headache, associated retro-orbital pain with eye movement, and conjunctival injection.

Some patients develop a bright, erythematous, maculopapular eruption 2 to 6 days into the illness that appears first on the trunk and then spreads to the face and extremities, with characteristic islands of unaffected skin throughout the involved area.³⁴

Severe back or groin pain occurs in about 60% of adult patients.³⁵

Anorexia, nausea, and vomiting are common.

Patients remain febrile for about 5 days, although some experience a biphasic (saddleback) fever that declines after 2 to 3 days, only to recur in about 24 hours.

In some patients, relative bradycardia is seen 2 to 3 days after fever onset.

Lymphadenopathy, sore throat, diarrhea or constipation, cutaneous hypesthesia, dysuria, dysgeusia, hepatitis, aseptic meningitis, and encephalopathy with delirium have been reported. Splenomegaly is rare.

In classic dengue fever, initial neutropenia and lymphopenia with subsequent lymphocytosis and monocytosis are often noted.

Mild hepatitis can be seen; aspartate aminotransferase and alanine aminotransferase levels can be two to three times the upper limit of normal.

Hemorrhagic manifestations, eg, petechiae, gingival bleeding, and epistaxis, may be seen in patients with mild thrombocytopenia even if they have no evidence of hemoconcentration or evidence of vascular instability.¹⁸

Although clinical dengue infection is usually self-limiting, acute symptoms can be incapacitating and can require hospitalization,³⁶ and convalescence may take several months because of ongoing asthenia or depression.³⁷ Furthermore, certain critical findings should alert the clinician to possible impending severe dengue and should lead to hospitalization for further observation until evolution to severe dengue has been ruled out (Table 2).

Severe dengue: Hemorrhagic fever and shock syndrome

In a very small subset of patients, dengue infection develops into a severe, potentially lifethreatening illness. Fortunately, this has rarely been reported in travelers.³⁸

Dengue hemorrhagic fever and dengue shock syndrome arise just as fever is subsiding. They constitute a spectrum of severe illness (Figure 4). Dengue hemorrhagic fever is poorly understood because its hemorrhagic manifestations are not of themselves diagnostic of the condition, as petechiae, epistaxis, and gingival bleeding may be seen in classic dengue fever (Figure 3) without progression to more severe illness.

The differentiating characteristic of severe dengue, in addition to hemorrhagic manifestations, is objective evidence of plasma leakage.39 Impending shock is suggested by the new onset of severe abdominal pain, restlessness, hepatomegaly, hypothermia, and diaphoresis.

The mechanisms causing the severe hemorrhagic manifestations characteristic of dengue hemorrhagic fever and the sudden onset of vascular permeability underlying dengue shock syndrome are not understood.⁴⁰ Many hypotheses have been generated and risk factors identified from observational and retrospective analyses. These include T-cell immune-pathologic responses involving receptors, antibodies, and cytokines,⁴¹ as well as specific host-genetic characteristics,^42,43 age,^44,45 sex,⁴⁶ comorbid conditions,^47–49 dengue virus virulence factors,⁵⁰ sequence of dengue infection, and infection parity.⁵¹ One hypothesis is that antibody-dependent enhancement of virus occurs during infection with a second dengue serotype after infection with a different serotype in the past, and that this may be the root cause of dengue hemorrhagic fever and dengue shock syndrome. However, this has not been proven.⁴⁰

DIAGNOSTIC TESTS FOR VIRUS, ANTIGENIC FRAGMENTS, ANTIBODIES

The appropriate test to confirm dengue virus infection is based on the natural history of the infection (Figure 4) coupled with the exposure risk in the returned traveler. These tests include isolation of the virus using cell culture, identification of antigenic fragments (test not available in the United States), and serologic tests for specific immunoglobulin M (IgM) and IgG antibodies using enzyme-linked immunosorbent assay (ELISA) or neutralization assays.⁵² During primary infection, viremia and antigenemia usually parallel fever, but when a person is later infected with a different dengue serotype, the period of viremia may be as short as 2 to 3 days, with antigens persisting in the serum for several more days.⁴⁰ Virus isolation is not routinely available but is both sensitive and specific for the diagnosis of dengue virus infection during the viremic period.

If polymerase chain reaction testing, dengue antigen capture ELISA, or virus isolation testing is not available, the ideal confirmatory procedure is to test for dengue IgM (looking for conversion from negative to positive), IgG (looking for a fourfold rise in antibody), or both, in paired serum samples collected 2 weeks apart, with the initial sample collected less than 5 days after the onset of symptoms. A presumptive diagnosis can be made if a single blood sample collected more than 7 days after symptom onset is found to have dengue IgM antibody. A single blood sample for IgM collected earlier than 7 days after the onset of illness may give a false-negative result (Figure 4) in infected persons.

DIFFERENTIAL DIAGNOSES: INFECTIOUS AND NONINFECTIOUS

The differential diagnosis of uncomplicated dengue in a traveler returning from an endemic area includes viral, bacterial, and protozoal infections as well as noninfectious conditions (Table 1).⁵³

Although most dengue virus infections are self-limiting, the clinical presentation may be severe enough to warrant hospitalization so that potentially life-threatening conditions can be systematically dismissed from the differential diagnosis.

Infections that can be rapidly fatal, such as malaria and enteric fever, need to be considered in patients who have traveled to endemic areas who present with undifferentiated fever. In cases of fever and maculopapular eruption, the differential diagnosis should include other causes of rash illness, such as measles and rubella. If hemorrhagic features are present, potentially fatal conditions need to be considered, including the classic viral hemorrhagic fevers caused by the Ebola and Marburg viruses, meningococcemia, the icterohemorrhagic form of leptospirosis, or other causes of bacterial sepsis. Other nonfatal infections should also be considered.

TREATMENT IS SUPPORTIVE

There is no antiviral treatment for dengue across the spectrum of disease presentations. Treatment is supportive and based on clinical presentation.

Acetaminophen (Tylenol) can be used to control fever, but aspirin and nonsteroidal anti-inflammatory drugs should not be used because they can make bleeding worse. Corticosteroids do not improve the outcome in severe dengue.²

Scrupulous attention to fluid and electrolyte balance is critical in severe dengue cases. Proper support and fluid resuscitation, including blood transfusion if needed, result in rapid recovery from dengue hemorrhagic fever with or without shock.

Suspected, probable, or confirmed cases of dengue should be reported to the local health department on the basis of published criteria (Table 3).

ADVICE TO TRAVELERS: DON’T GET BITTEN

There is currently no commercially available dengue vaccine, although several are under development.⁵⁴ Therefore, pretravel counseling on how to avoid mosquito bites when traveling to dengue-endemic areas is the key dengue prevention strategy. Proactive prevention strategies include use of insect repellents such as those containing diethyltoluamide (DEET) or permethrin⁵⁵ and elimination of outdoor locations where mosquitoes lay eggs, such as flower planter dishes, to reduce local mosquito breeding.⁵⁶

Patients who have had a previous dengue infection should be counseled about the possible increased risk of severe disease if infected with a second dengue serotype.
 

Acknowledgment: The author thanks Chester G. Moore, PhD, of Colorado State University for assistance in creating Figure 2, based on data contained in the CDC, ArboNET, and Exotic/Invasive databases.

Why do primary care physicians in nontropical parts of the world need to be on the lookout for tropical diseases such as dengue?

First, more people are traveling than ever before, and second, more people are traveling to parts of the world where dengue and other tropical diseases are endemic. Thus, dengue should now be included in the differential diagnosis of fever in anyone returning from travel to a part of the world where dengue is endemic (Table 1).¹

The number of cases of dengue in returning US travelers has been increasing steadily over the past 25 years.^2,3 Dengue is a more common cause of febrile illness than malaria in US travelers returning from all tropical and subtropical regions except Africa.⁴

Moreover, dengue may be gaining a permanent foothold in the United States, and in areas of the country where mosquitoes that can transmit the virus are found, primary care physicians are the first line of defense in public health. Specifically, to prevent the virus from becoming established locally, primary care physicians need to quickly identify and report cases to public health authorities, who can promptly follow up and initiate prevention measures.⁵

Underscoring the importance of dengue, this infection was added in 2009 to the list of nationally notifiable infectious diseases in the United States.⁶

A COMMON INFECTION WORLDWIDE

Dengue causes up to 100 million new infections, 500,000 hospitalizations, and 25,000 deaths every year in the 2.5 billion people who live in subtropical and tropical areas of the world.⁷ It is transmitted by mosquitoes of the genus Aedes (which, unlike most other mosquitoes, often bite in the daytime), and it is the most common arboviral infection (ie, transmitted mainly by arthropods) worldwide.

The dengue virus belongs to the family of flaviviruses, which includes West Nile virus, St. Louis encephalitis, and yellow fever. It has four closely related but antigenically distinct serotypes, designated DENV-1, DENV-2, DENV-3, and DENV-4. Infection with one serotype induces lifetime homotypic immunity but only short-lived heterotypic immunity to the other dengue serotypes.⁸ Hence, a person can be infected over time with each of the four serotypes. The first infection is termed the primary infection; subsequent infection with any of the remaining three serotypes is termed a secondary infection.

Source: US Centers for Disease Control and Prevention.

Dengue virus transmission has been expanding since the end of World War II in Asia and since the 1980s in the Americas following the end of many regional vector-control programs.⁹ Although dengue is known to occur in tropical Africa, its epidemiology is less well defined on that continent (Figure 1).¹⁰

Explosive epidemics of dengue occur when there are enough mosquitoes and a susceptible population across a broad age range, ie, both children and adults.^11,12 Transmission can be halted with vigorous vector-control programs, or it slows and stops when the pool of susceptible people is exhausted.^13–15

On the other hand, hyperendemic transmission occurs in areas in which multiple virus serotypes continuously circulate in a large pool of susceptible people. In these areas, dengue seroprevalence increases with age, and most adults are immune.

Anyone of any age who travels from a nonendemic area to an epidemic or hyperendemic area is at risk of infection.¹⁶

DENGUE IN THE UNITED STATES

Reported cases of dengue in South and Central America, the Caribbean, and Mexico, common destinations for US travelers, have increased more than fourfold since the 1980s. There were a total of approximately 1 million cases in the 10-year period ending in 1989, compared with more than 4.5 million in the 8-year period from 2000 to 2007.¹¹

The geographic proximity of these areas to the continental United States and the large numbers of US residents travelling to these areas have raised concern that dengue could emerge in the continental United States in areas where potential vectors exist (see below).¹⁷ Furthermore, several US territories and former territories where tourism is an economic mainstay, including the Commonwealth of Puerto Rico,¹⁸ the US Virgin Islands,¹⁹ American Samoa, and other smaller Pacific island jurisdictions such as Palau,²⁰ have reported dengue virus circulation.

Arbonet database, maintained by the US Centers for Disease Control and Prevention.

Adding to the concern that dengue could gain a persistent foothold in the United States, competent dengue vectors are found here. Two mosquito vectors, Aedes aegypti21 and Aedes albopictus,^22,23 live in some areas of the southwestern and southeastern United States (Figure 2), with Aedes aegypti being the more competent transmitter. Both vectors may be abundant in warmer months. This raises a concern that a returning dengue-infected traveler could initiate an outbreak of autochthonous transmission.²⁴

Notably, endemic dengue transmission has occurred in the past in the United States, and the virus is circulating here again at low levels. From 1946 to 1980, no cases of dengue were acquired in the continental United States. However, since 1980 there have been seven outbreaks of laboratory-confirmed, locally acquired dengue along the Texas-Mexico border.^25–27 More recently, local transmission emerged in Key West, Florida,^28,29 the first outbreak since 1945 of dengue in the continental United States not to occur near the Texas-Mexico border. And in early 2011, nonsustained but locally acquired transmission was confirmed in Hawaii after a transmission-free decade.³⁰

THE CLINICAL SPECTRUM OF DENGUE VIRUS INFECTION

Most primary and secondary dengue infections are asymptomatic.⁸ The common forms of clinically apparent disease include self-limited, undifferentiated fever and classic dengue fever (Figure 3). Severe disease, manifesting as either dengue hemorrhagic fever or dengue shock syndrome, is a rare outcome of dengue virus infection, estimated to occur in 1% of cases worldwide.³¹ However, the true proportion of severe infection among all dengue cases seen in travelers is difficult to assess reliably.

Asymptomatic infection

Clinical dengue disease is relatively uncommon, as between 60% and 80% of infections are asymptomatic, particularly in children and adults who never have been infected before.³²

In the recent outbreak in Key West, 28 symptomatic cases of locally acquired dengue were detected. The US Centers for Disease Control and Prevention (CDC) conducted a serologic survey of 240 healthy, randomly selected residents of Key West and found evidence of recent infection in 5.4% of those tested.²⁸ Based on this finding, the CDC estimated that 1,000 people had been infected, of whom more than 90% had no symptoms. However, no attempt was made to differentiate primary from secondary infection in this serosurvey. (Previous infection with one dengue serotype places some individuals at risk for severe dengue if infected with a different serotype in the future.)

Uncomplicated dengue infection

Undifferentiated fever³³ and classic dengue fever are the most common manifestations of clinical dengue infection. Also known as breakbone fever, classic dengue fever is a fever-arthralgia-rash syndrome.¹

The onset is acute, with a high fever (though rarely greater than 40.5°C [104.9°F]) 3 to 14 days (usually 5 to 9 days) after the patient was bitten by an Aedes mosquito. Therefore, a febrile illness beginning more than 2 weeks after returning from travel to an endemic area is unlikely to be dengue, and another diagnosis should be sought.

A prodrome of headache, backache, fatigue, chills, anorexia, and occasionally a rash may precede the onset of fever by about 12 hours.

With fever comes a severe frontal headache, associated retro-orbital pain with eye movement, and conjunctival injection.

Some patients develop a bright, erythematous, maculopapular eruption 2 to 6 days into the illness that appears first on the trunk and then spreads to the face and extremities, with characteristic islands of unaffected skin throughout the involved area.³⁴

Severe back or groin pain occurs in about 60% of adult patients.³⁵

Anorexia, nausea, and vomiting are common.

Patients remain febrile for about 5 days, although some experience a biphasic (saddleback) fever that declines after 2 to 3 days, only to recur in about 24 hours.

In some patients, relative bradycardia is seen 2 to 3 days after fever onset.

Lymphadenopathy, sore throat, diarrhea or constipation, cutaneous hypesthesia, dysuria, dysgeusia, hepatitis, aseptic meningitis, and encephalopathy with delirium have been reported. Splenomegaly is rare.

In classic dengue fever, initial neutropenia and lymphopenia with subsequent lymphocytosis and monocytosis are often noted.

Mild hepatitis can be seen; aspartate aminotransferase and alanine aminotransferase levels can be two to three times the upper limit of normal.

Hemorrhagic manifestations, eg, petechiae, gingival bleeding, and epistaxis, may be seen in patients with mild thrombocytopenia even if they have no evidence of hemoconcentration or evidence of vascular instability.¹⁸

Although clinical dengue infection is usually self-limiting, acute symptoms can be incapacitating and can require hospitalization,³⁶ and convalescence may take several months because of ongoing asthenia or depression.³⁷ Furthermore, certain critical findings should alert the clinician to possible impending severe dengue and should lead to hospitalization for further observation until evolution to severe dengue has been ruled out (Table 2).

Severe dengue: Hemorrhagic fever and shock syndrome

In a very small subset of patients, dengue infection develops into a severe, potentially lifethreatening illness. Fortunately, this has rarely been reported in travelers.³⁸

Dengue hemorrhagic fever and dengue shock syndrome arise just as fever is subsiding. They constitute a spectrum of severe illness (Figure 4). Dengue hemorrhagic fever is poorly understood because its hemorrhagic manifestations are not of themselves diagnostic of the condition, as petechiae, epistaxis, and gingival bleeding may be seen in classic dengue fever (Figure 3) without progression to more severe illness.

The differentiating characteristic of severe dengue, in addition to hemorrhagic manifestations, is objective evidence of plasma leakage.39 Impending shock is suggested by the new onset of severe abdominal pain, restlessness, hepatomegaly, hypothermia, and diaphoresis.

The mechanisms causing the severe hemorrhagic manifestations characteristic of dengue hemorrhagic fever and the sudden onset of vascular permeability underlying dengue shock syndrome are not understood.⁴⁰ Many hypotheses have been generated and risk factors identified from observational and retrospective analyses. These include T-cell immune-pathologic responses involving receptors, antibodies, and cytokines,⁴¹ as well as specific host-genetic characteristics,^42,43 age,^44,45 sex,⁴⁶ comorbid conditions,^47–49 dengue virus virulence factors,⁵⁰ sequence of dengue infection, and infection parity.⁵¹ One hypothesis is that antibody-dependent enhancement of virus occurs during infection with a second dengue serotype after infection with a different serotype in the past, and that this may be the root cause of dengue hemorrhagic fever and dengue shock syndrome. However, this has not been proven.⁴⁰

DIAGNOSTIC TESTS FOR VIRUS, ANTIGENIC FRAGMENTS, ANTIBODIES

The appropriate test to confirm dengue virus infection is based on the natural history of the infection (Figure 4) coupled with the exposure risk in the returned traveler. These tests include isolation of the virus using cell culture, identification of antigenic fragments (test not available in the United States), and serologic tests for specific immunoglobulin M (IgM) and IgG antibodies using enzyme-linked immunosorbent assay (ELISA) or neutralization assays.⁵² During primary infection, viremia and antigenemia usually parallel fever, but when a person is later infected with a different dengue serotype, the period of viremia may be as short as 2 to 3 days, with antigens persisting in the serum for several more days.⁴⁰ Virus isolation is not routinely available but is both sensitive and specific for the diagnosis of dengue virus infection during the viremic period.

If polymerase chain reaction testing, dengue antigen capture ELISA, or virus isolation testing is not available, the ideal confirmatory procedure is to test for dengue IgM (looking for conversion from negative to positive), IgG (looking for a fourfold rise in antibody), or both, in paired serum samples collected 2 weeks apart, with the initial sample collected less than 5 days after the onset of symptoms. A presumptive diagnosis can be made if a single blood sample collected more than 7 days after symptom onset is found to have dengue IgM antibody. A single blood sample for IgM collected earlier than 7 days after the onset of illness may give a false-negative result (Figure 4) in infected persons.

DIFFERENTIAL DIAGNOSES: INFECTIOUS AND NONINFECTIOUS

The differential diagnosis of uncomplicated dengue in a traveler returning from an endemic area includes viral, bacterial, and protozoal infections as well as noninfectious conditions (Table 1).⁵³

Although most dengue virus infections are self-limiting, the clinical presentation may be severe enough to warrant hospitalization so that potentially life-threatening conditions can be systematically dismissed from the differential diagnosis.

Infections that can be rapidly fatal, such as malaria and enteric fever, need to be considered in patients who have traveled to endemic areas who present with undifferentiated fever. In cases of fever and maculopapular eruption, the differential diagnosis should include other causes of rash illness, such as measles and rubella. If hemorrhagic features are present, potentially fatal conditions need to be considered, including the classic viral hemorrhagic fevers caused by the Ebola and Marburg viruses, meningococcemia, the icterohemorrhagic form of leptospirosis, or other causes of bacterial sepsis. Other nonfatal infections should also be considered.

TREATMENT IS SUPPORTIVE

There is no antiviral treatment for dengue across the spectrum of disease presentations. Treatment is supportive and based on clinical presentation.

Acetaminophen (Tylenol) can be used to control fever, but aspirin and nonsteroidal anti-inflammatory drugs should not be used because they can make bleeding worse. Corticosteroids do not improve the outcome in severe dengue.²

Scrupulous attention to fluid and electrolyte balance is critical in severe dengue cases. Proper support and fluid resuscitation, including blood transfusion if needed, result in rapid recovery from dengue hemorrhagic fever with or without shock.

Suspected, probable, or confirmed cases of dengue should be reported to the local health department on the basis of published criteria (Table 3).

ADVICE TO TRAVELERS: DON’T GET BITTEN

There is currently no commercially available dengue vaccine, although several are under development.⁵⁴ Therefore, pretravel counseling on how to avoid mosquito bites when traveling to dengue-endemic areas is the key dengue prevention strategy. Proactive prevention strategies include use of insect repellents such as those containing diethyltoluamide (DEET) or permethrin⁵⁵ and elimination of outdoor locations where mosquitoes lay eggs, such as flower planter dishes, to reduce local mosquito breeding.⁵⁶

Patients who have had a previous dengue infection should be counseled about the possible increased risk of severe disease if infected with a second dengue serotype.
 

Acknowledgment: The author thanks Chester G. Moore, PhD, of Colorado State University for assistance in creating Figure 2, based on data contained in the CDC, ArboNET, and Exotic/Invasive databases.

Why do primary care physicians in nontropical parts of the world need to be on the lookout for tropical diseases such as dengue?

First, more people are traveling than ever before, and second, more people are traveling to parts of the world where dengue and other tropical diseases are endemic. Thus, dengue should now be included in the differential diagnosis of fever in anyone returning from travel to a part of the world where dengue is endemic (Table 1).¹

The number of cases of dengue in returning US travelers has been increasing steadily over the past 25 years.^2,3 Dengue is a more common cause of febrile illness than malaria in US travelers returning from all tropical and subtropical regions except Africa.⁴

Moreover, dengue may be gaining a permanent foothold in the United States, and in areas of the country where mosquitoes that can transmit the virus are found, primary care physicians are the first line of defense in public health. Specifically, to prevent the virus from becoming established locally, primary care physicians need to quickly identify and report cases to public health authorities, who can promptly follow up and initiate prevention measures.⁵

Underscoring the importance of dengue, this infection was added in 2009 to the list of nationally notifiable infectious diseases in the United States.⁶

A COMMON INFECTION WORLDWIDE

Dengue causes up to 100 million new infections, 500,000 hospitalizations, and 25,000 deaths every year in the 2.5 billion people who live in subtropical and tropical areas of the world.⁷ It is transmitted by mosquitoes of the genus Aedes (which, unlike most other mosquitoes, often bite in the daytime), and it is the most common arboviral infection (ie, transmitted mainly by arthropods) worldwide.

The dengue virus belongs to the family of flaviviruses, which includes West Nile virus, St. Louis encephalitis, and yellow fever. It has four closely related but antigenically distinct serotypes, designated DENV-1, DENV-2, DENV-3, and DENV-4. Infection with one serotype induces lifetime homotypic immunity but only short-lived heterotypic immunity to the other dengue serotypes.⁸ Hence, a person can be infected over time with each of the four serotypes. The first infection is termed the primary infection; subsequent infection with any of the remaining three serotypes is termed a secondary infection.

Source: US Centers for Disease Control and Prevention.

Dengue virus transmission has been expanding since the end of World War II in Asia and since the 1980s in the Americas following the end of many regional vector-control programs.⁹ Although dengue is known to occur in tropical Africa, its epidemiology is less well defined on that continent (Figure 1).¹⁰

Explosive epidemics of dengue occur when there are enough mosquitoes and a susceptible population across a broad age range, ie, both children and adults.^11,12 Transmission can be halted with vigorous vector-control programs, or it slows and stops when the pool of susceptible people is exhausted.^13–15

On the other hand, hyperendemic transmission occurs in areas in which multiple virus serotypes continuously circulate in a large pool of susceptible people. In these areas, dengue seroprevalence increases with age, and most adults are immune.

Anyone of any age who travels from a nonendemic area to an epidemic or hyperendemic area is at risk of infection.¹⁶

DENGUE IN THE UNITED STATES

Reported cases of dengue in South and Central America, the Caribbean, and Mexico, common destinations for US travelers, have increased more than fourfold since the 1980s. There were a total of approximately 1 million cases in the 10-year period ending in 1989, compared with more than 4.5 million in the 8-year period from 2000 to 2007.¹¹

The geographic proximity of these areas to the continental United States and the large numbers of US residents travelling to these areas have raised concern that dengue could emerge in the continental United States in areas where potential vectors exist (see below).¹⁷ Furthermore, several US territories and former territories where tourism is an economic mainstay, including the Commonwealth of Puerto Rico,¹⁸ the US Virgin Islands,¹⁹ American Samoa, and other smaller Pacific island jurisdictions such as Palau,²⁰ have reported dengue virus circulation.

Arbonet database, maintained by the US Centers for Disease Control and Prevention.

Adding to the concern that dengue could gain a persistent foothold in the United States, competent dengue vectors are found here. Two mosquito vectors, Aedes aegypti21 and Aedes albopictus,^22,23 live in some areas of the southwestern and southeastern United States (Figure 2), with Aedes aegypti being the more competent transmitter. Both vectors may be abundant in warmer months. This raises a concern that a returning dengue-infected traveler could initiate an outbreak of autochthonous transmission.²⁴

Notably, endemic dengue transmission has occurred in the past in the United States, and the virus is circulating here again at low levels. From 1946 to 1980, no cases of dengue were acquired in the continental United States. However, since 1980 there have been seven outbreaks of laboratory-confirmed, locally acquired dengue along the Texas-Mexico border.^25–27 More recently, local transmission emerged in Key West, Florida,^28,29 the first outbreak since 1945 of dengue in the continental United States not to occur near the Texas-Mexico border. And in early 2011, nonsustained but locally acquired transmission was confirmed in Hawaii after a transmission-free decade.³⁰

THE CLINICAL SPECTRUM OF DENGUE VIRUS INFECTION

Most primary and secondary dengue infections are asymptomatic.⁸ The common forms of clinically apparent disease include self-limited, undifferentiated fever and classic dengue fever (Figure 3). Severe disease, manifesting as either dengue hemorrhagic fever or dengue shock syndrome, is a rare outcome of dengue virus infection, estimated to occur in 1% of cases worldwide.³¹ However, the true proportion of severe infection among all dengue cases seen in travelers is difficult to assess reliably.

Asymptomatic infection

Clinical dengue disease is relatively uncommon, as between 60% and 80% of infections are asymptomatic, particularly in children and adults who never have been infected before.³²

In the recent outbreak in Key West, 28 symptomatic cases of locally acquired dengue were detected. The US Centers for Disease Control and Prevention (CDC) conducted a serologic survey of 240 healthy, randomly selected residents of Key West and found evidence of recent infection in 5.4% of those tested.²⁸ Based on this finding, the CDC estimated that 1,000 people had been infected, of whom more than 90% had no symptoms. However, no attempt was made to differentiate primary from secondary infection in this serosurvey. (Previous infection with one dengue serotype places some individuals at risk for severe dengue if infected with a different serotype in the future.)

Uncomplicated dengue infection

Undifferentiated fever³³ and classic dengue fever are the most common manifestations of clinical dengue infection. Also known as breakbone fever, classic dengue fever is a fever-arthralgia-rash syndrome.¹

The onset is acute, with a high fever (though rarely greater than 40.5°C [104.9°F]) 3 to 14 days (usually 5 to 9 days) after the patient was bitten by an Aedes mosquito. Therefore, a febrile illness beginning more than 2 weeks after returning from travel to an endemic area is unlikely to be dengue, and another diagnosis should be sought.

A prodrome of headache, backache, fatigue, chills, anorexia, and occasionally a rash may precede the onset of fever by about 12 hours.

With fever comes a severe frontal headache, associated retro-orbital pain with eye movement, and conjunctival injection.

Some patients develop a bright, erythematous, maculopapular eruption 2 to 6 days into the illness that appears first on the trunk and then spreads to the face and extremities, with characteristic islands of unaffected skin throughout the involved area.³⁴

Severe back or groin pain occurs in about 60% of adult patients.³⁵

Anorexia, nausea, and vomiting are common.

Patients remain febrile for about 5 days, although some experience a biphasic (saddleback) fever that declines after 2 to 3 days, only to recur in about 24 hours.

In some patients, relative bradycardia is seen 2 to 3 days after fever onset.

Lymphadenopathy, sore throat, diarrhea or constipation, cutaneous hypesthesia, dysuria, dysgeusia, hepatitis, aseptic meningitis, and encephalopathy with delirium have been reported. Splenomegaly is rare.

In classic dengue fever, initial neutropenia and lymphopenia with subsequent lymphocytosis and monocytosis are often noted.

Mild hepatitis can be seen; aspartate aminotransferase and alanine aminotransferase levels can be two to three times the upper limit of normal.

Hemorrhagic manifestations, eg, petechiae, gingival bleeding, and epistaxis, may be seen in patients with mild thrombocytopenia even if they have no evidence of hemoconcentration or evidence of vascular instability.¹⁸

Although clinical dengue infection is usually self-limiting, acute symptoms can be incapacitating and can require hospitalization,³⁶ and convalescence may take several months because of ongoing asthenia or depression.³⁷ Furthermore, certain critical findings should alert the clinician to possible impending severe dengue and should lead to hospitalization for further observation until evolution to severe dengue has been ruled out (Table 2).

Severe dengue: Hemorrhagic fever and shock syndrome

In a very small subset of patients, dengue infection develops into a severe, potentially lifethreatening illness. Fortunately, this has rarely been reported in travelers.³⁸

Dengue hemorrhagic fever and dengue shock syndrome arise just as fever is subsiding. They constitute a spectrum of severe illness (Figure 4). Dengue hemorrhagic fever is poorly understood because its hemorrhagic manifestations are not of themselves diagnostic of the condition, as petechiae, epistaxis, and gingival bleeding may be seen in classic dengue fever (Figure 3) without progression to more severe illness.

The differentiating characteristic of severe dengue, in addition to hemorrhagic manifestations, is objective evidence of plasma leakage.39 Impending shock is suggested by the new onset of severe abdominal pain, restlessness, hepatomegaly, hypothermia, and diaphoresis.

The mechanisms causing the severe hemorrhagic manifestations characteristic of dengue hemorrhagic fever and the sudden onset of vascular permeability underlying dengue shock syndrome are not understood.⁴⁰ Many hypotheses have been generated and risk factors identified from observational and retrospective analyses. These include T-cell immune-pathologic responses involving receptors, antibodies, and cytokines,⁴¹ as well as specific host-genetic characteristics,^42,43 age,^44,45 sex,⁴⁶ comorbid conditions,^47–49 dengue virus virulence factors,⁵⁰ sequence of dengue infection, and infection parity.⁵¹ One hypothesis is that antibody-dependent enhancement of virus occurs during infection with a second dengue serotype after infection with a different serotype in the past, and that this may be the root cause of dengue hemorrhagic fever and dengue shock syndrome. However, this has not been proven.⁴⁰

DIAGNOSTIC TESTS FOR VIRUS, ANTIGENIC FRAGMENTS, ANTIBODIES

The appropriate test to confirm dengue virus infection is based on the natural history of the infection (Figure 4) coupled with the exposure risk in the returned traveler. These tests include isolation of the virus using cell culture, identification of antigenic fragments (test not available in the United States), and serologic tests for specific immunoglobulin M (IgM) and IgG antibodies using enzyme-linked immunosorbent assay (ELISA) or neutralization assays.⁵² During primary infection, viremia and antigenemia usually parallel fever, but when a person is later infected with a different dengue serotype, the period of viremia may be as short as 2 to 3 days, with antigens persisting in the serum for several more days.⁴⁰ Virus isolation is not routinely available but is both sensitive and specific for the diagnosis of dengue virus infection during the viremic period.

If polymerase chain reaction testing, dengue antigen capture ELISA, or virus isolation testing is not available, the ideal confirmatory procedure is to test for dengue IgM (looking for conversion from negative to positive), IgG (looking for a fourfold rise in antibody), or both, in paired serum samples collected 2 weeks apart, with the initial sample collected less than 5 days after the onset of symptoms. A presumptive diagnosis can be made if a single blood sample collected more than 7 days after symptom onset is found to have dengue IgM antibody. A single blood sample for IgM collected earlier than 7 days after the onset of illness may give a false-negative result (Figure 4) in infected persons.

DIFFERENTIAL DIAGNOSES: INFECTIOUS AND NONINFECTIOUS

The differential diagnosis of uncomplicated dengue in a traveler returning from an endemic area includes viral, bacterial, and protozoal infections as well as noninfectious conditions (Table 1).⁵³

Although most dengue virus infections are self-limiting, the clinical presentation may be severe enough to warrant hospitalization so that potentially life-threatening conditions can be systematically dismissed from the differential diagnosis.

Infections that can be rapidly fatal, such as malaria and enteric fever, need to be considered in patients who have traveled to endemic areas who present with undifferentiated fever. In cases of fever and maculopapular eruption, the differential diagnosis should include other causes of rash illness, such as measles and rubella. If hemorrhagic features are present, potentially fatal conditions need to be considered, including the classic viral hemorrhagic fevers caused by the Ebola and Marburg viruses, meningococcemia, the icterohemorrhagic form of leptospirosis, or other causes of bacterial sepsis. Other nonfatal infections should also be considered.

TREATMENT IS SUPPORTIVE

There is no antiviral treatment for dengue across the spectrum of disease presentations. Treatment is supportive and based on clinical presentation.

Acetaminophen (Tylenol) can be used to control fever, but aspirin and nonsteroidal anti-inflammatory drugs should not be used because they can make bleeding worse. Corticosteroids do not improve the outcome in severe dengue.²

Scrupulous attention to fluid and electrolyte balance is critical in severe dengue cases. Proper support and fluid resuscitation, including blood transfusion if needed, result in rapid recovery from dengue hemorrhagic fever with or without shock.

Suspected, probable, or confirmed cases of dengue should be reported to the local health department on the basis of published criteria (Table 3).

ADVICE TO TRAVELERS: DON’T GET BITTEN

There is currently no commercially available dengue vaccine, although several are under development.⁵⁴ Therefore, pretravel counseling on how to avoid mosquito bites when traveling to dengue-endemic areas is the key dengue prevention strategy. Proactive prevention strategies include use of insect repellents such as those containing diethyltoluamide (DEET) or permethrin⁵⁵ and elimination of outdoor locations where mosquitoes lay eggs, such as flower planter dishes, to reduce local mosquito breeding.⁵⁶

Patients who have had a previous dengue infection should be counseled about the possible increased risk of severe disease if infected with a second dengue serotype.
 

Acknowledgment: The author thanks Chester G. Moore, PhD, of Colorado State University for assistance in creating Figure 2, based on data contained in the CDC, ArboNET, and Exotic/Invasive databases.

A 12-year-old girl has a large, irregular area of hair loss over the central frontoparietal scalp. Physical examination reveals scattered short hairs of varying lengths and a few small crusts throughout the area of alopecia (Figure 1). The remainder of the scalp appears normal.

Q: Which diagnosis is most likely?

• Alopecia areata
• Lichen planopilaris
• Discoid lupus erythematosus
• Trichotillomania
• Follicular degeneration syndrome

A: The correct answer is trichotillomania, the compulsive pulling out of one’s own hair. Irregularly shaped areas of alopecia containing short hairs of varied lengths and excoriation should raise clinical suspicion of trichotillomania. Biopsy can confirm the diagnosis when follicles devoid of hair shafts, hemorrhage, and misshapen fragments of scalp hair (pigment casts) are seen.

DIAGNOSTIC CLUES

Trichotillomania may present as striking hair loss (alopecia) with an irregular pattern, often with sharp angles or scalloped borders.¹ Short and broken hairs within involved areas are typically seen because regenerating hairs are too short to be grasped and pulled out.² Although hair loss on the scalp may be most evident, hair loss on any hair-bearing area of the body may be noted, including eyebrows and eyelashes.

Family members and the affected individual are often aware of compulsive manipulation of hair.

Depression, anxiety, and other grooming behaviors such as skin-picking and nail-biting may be associated with trichotillomania. Affected individuals often feel a sense of gratification from pulling out hairs. Although systemic complications are rare, some individuals ingest the removed hairs (trichophagy), and the hairs may be caught in the gastric folds and eventually form a trichobezoar.³

The diagnosis is usually based on clinical findings and by asking the patient about hair-pulling. Asking the patient if the habit is due to the feel of the hair, a need to calm himself or herself, or other factors may be revealing. The majority of cases can be diagnosed without biopsy. Biopsy from affected areas reveals changes related to trauma such as empty hair follicles, hemorrhage, and hair shaft fragments in the dermis² (Figure 2). The number of catagen follicles is increased. Other causes of patchy alopecia are associated with different findings on biopsy.

Alopecia areata may be associated with an increased number of catagen hairs but is characterized by a peribulbar lymphocytic infiltrate.

Biopsy of lichen planopilaris typically reveals vacuolar changes along the dermal-follicular junction and necrotic keratinocytes.

Cutaneous lupus erythematosus is associated with thickening of the basement membrane zone, increased mucin in the dermis, follicular plugging by keratin, and vacuolar changes along the dermal-epidermal junction.

Biopsy of follicular degeneration syndrome exhibits premature desquamation of the internal root sheath as well as an increased number of fibrous tracts marking the sites of lost hairs.

The etiology of trichotillomania remains largely unknown, and the prognosis varies.^4,5 There may be a family history, as there appears to be a genetic component to this disease. The disorder may also occur in the absence of external stressors.⁵

TREATMENT OPTIONS

Young children often develop trichotillomania that is transient in nature and most often does not require formal intervention. Older children may benefit from psychotherapy.⁵

Clomipramine (Anafranil) has been shown to be more effective than placebo.⁶ Selective serotonin reuptake inhibitors are no more effective than placebo.^6,7 Pimozide (Orap), haloperidol (Haldol), and other agents have been reported to be of benefit in some instances. Although no large randomized clinical trials in children have been performed, N-acetylcysteine (Acetadote) seems to be a very promising form of therapy in adults.⁸ A multidisciplinary approach is usually helpful in finding the best treatment option for a particular patient.

A 12-year-old girl has a large, irregular area of hair loss over the central frontoparietal scalp. Physical examination reveals scattered short hairs of varying lengths and a few small crusts throughout the area of alopecia (Figure 1). The remainder of the scalp appears normal.

Q: Which diagnosis is most likely?

• Alopecia areata
• Lichen planopilaris
• Discoid lupus erythematosus
• Trichotillomania
• Follicular degeneration syndrome

A: The correct answer is trichotillomania, the compulsive pulling out of one’s own hair. Irregularly shaped areas of alopecia containing short hairs of varied lengths and excoriation should raise clinical suspicion of trichotillomania. Biopsy can confirm the diagnosis when follicles devoid of hair shafts, hemorrhage, and misshapen fragments of scalp hair (pigment casts) are seen.

DIAGNOSTIC CLUES

Trichotillomania may present as striking hair loss (alopecia) with an irregular pattern, often with sharp angles or scalloped borders.¹ Short and broken hairs within involved areas are typically seen because regenerating hairs are too short to be grasped and pulled out.² Although hair loss on the scalp may be most evident, hair loss on any hair-bearing area of the body may be noted, including eyebrows and eyelashes.

Family members and the affected individual are often aware of compulsive manipulation of hair.

Depression, anxiety, and other grooming behaviors such as skin-picking and nail-biting may be associated with trichotillomania. Affected individuals often feel a sense of gratification from pulling out hairs. Although systemic complications are rare, some individuals ingest the removed hairs (trichophagy), and the hairs may be caught in the gastric folds and eventually form a trichobezoar.³

The diagnosis is usually based on clinical findings and by asking the patient about hair-pulling. Asking the patient if the habit is due to the feel of the hair, a need to calm himself or herself, or other factors may be revealing. The majority of cases can be diagnosed without biopsy. Biopsy from affected areas reveals changes related to trauma such as empty hair follicles, hemorrhage, and hair shaft fragments in the dermis² (Figure 2). The number of catagen follicles is increased. Other causes of patchy alopecia are associated with different findings on biopsy.

Alopecia areata may be associated with an increased number of catagen hairs but is characterized by a peribulbar lymphocytic infiltrate.

Biopsy of lichen planopilaris typically reveals vacuolar changes along the dermal-follicular junction and necrotic keratinocytes.

Cutaneous lupus erythematosus is associated with thickening of the basement membrane zone, increased mucin in the dermis, follicular plugging by keratin, and vacuolar changes along the dermal-epidermal junction.

Biopsy of follicular degeneration syndrome exhibits premature desquamation of the internal root sheath as well as an increased number of fibrous tracts marking the sites of lost hairs.

The etiology of trichotillomania remains largely unknown, and the prognosis varies.^4,5 There may be a family history, as there appears to be a genetic component to this disease. The disorder may also occur in the absence of external stressors.⁵

TREATMENT OPTIONS

Young children often develop trichotillomania that is transient in nature and most often does not require formal intervention. Older children may benefit from psychotherapy.⁵

Clomipramine (Anafranil) has been shown to be more effective than placebo.⁶ Selective serotonin reuptake inhibitors are no more effective than placebo.^6,7 Pimozide (Orap), haloperidol (Haldol), and other agents have been reported to be of benefit in some instances. Although no large randomized clinical trials in children have been performed, N-acetylcysteine (Acetadote) seems to be a very promising form of therapy in adults.⁸ A multidisciplinary approach is usually helpful in finding the best treatment option for a particular patient.

A 12-year-old girl has a large, irregular area of hair loss over the central frontoparietal scalp. Physical examination reveals scattered short hairs of varying lengths and a few small crusts throughout the area of alopecia (Figure 1). The remainder of the scalp appears normal.

Q: Which diagnosis is most likely?

• Alopecia areata
• Lichen planopilaris
• Discoid lupus erythematosus
• Trichotillomania
• Follicular degeneration syndrome

A: The correct answer is trichotillomania, the compulsive pulling out of one’s own hair. Irregularly shaped areas of alopecia containing short hairs of varied lengths and excoriation should raise clinical suspicion of trichotillomania. Biopsy can confirm the diagnosis when follicles devoid of hair shafts, hemorrhage, and misshapen fragments of scalp hair (pigment casts) are seen.

DIAGNOSTIC CLUES

Trichotillomania may present as striking hair loss (alopecia) with an irregular pattern, often with sharp angles or scalloped borders.¹ Short and broken hairs within involved areas are typically seen because regenerating hairs are too short to be grasped and pulled out.² Although hair loss on the scalp may be most evident, hair loss on any hair-bearing area of the body may be noted, including eyebrows and eyelashes.

Family members and the affected individual are often aware of compulsive manipulation of hair.

Depression, anxiety, and other grooming behaviors such as skin-picking and nail-biting may be associated with trichotillomania. Affected individuals often feel a sense of gratification from pulling out hairs. Although systemic complications are rare, some individuals ingest the removed hairs (trichophagy), and the hairs may be caught in the gastric folds and eventually form a trichobezoar.³

The diagnosis is usually based on clinical findings and by asking the patient about hair-pulling. Asking the patient if the habit is due to the feel of the hair, a need to calm himself or herself, or other factors may be revealing. The majority of cases can be diagnosed without biopsy. Biopsy from affected areas reveals changes related to trauma such as empty hair follicles, hemorrhage, and hair shaft fragments in the dermis² (Figure 2). The number of catagen follicles is increased. Other causes of patchy alopecia are associated with different findings on biopsy.

Alopecia areata may be associated with an increased number of catagen hairs but is characterized by a peribulbar lymphocytic infiltrate.

Biopsy of lichen planopilaris typically reveals vacuolar changes along the dermal-follicular junction and necrotic keratinocytes.

Cutaneous lupus erythematosus is associated with thickening of the basement membrane zone, increased mucin in the dermis, follicular plugging by keratin, and vacuolar changes along the dermal-epidermal junction.

Biopsy of follicular degeneration syndrome exhibits premature desquamation of the internal root sheath as well as an increased number of fibrous tracts marking the sites of lost hairs.

The etiology of trichotillomania remains largely unknown, and the prognosis varies.^4,5 There may be a family history, as there appears to be a genetic component to this disease. The disorder may also occur in the absence of external stressors.⁵

TREATMENT OPTIONS

Young children often develop trichotillomania that is transient in nature and most often does not require formal intervention. Older children may benefit from psychotherapy.⁵

Clomipramine (Anafranil) has been shown to be more effective than placebo.⁶ Selective serotonin reuptake inhibitors are no more effective than placebo.^6,7 Pimozide (Orap), haloperidol (Haldol), and other agents have been reported to be of benefit in some instances. Although no large randomized clinical trials in children have been performed, N-acetylcysteine (Acetadote) seems to be a very promising form of therapy in adults.⁸ A multidisciplinary approach is usually helpful in finding the best treatment option for a particular patient.