Implantation of a deep brain stimulator is the most common surgical procedure performed in the United States and industrialized world for the management of advanced movement disorders. These procedures are US Food and Drug Administration (FDA)–approved for the management of the symptoms of Parkinson disease (PD) and essential tremor. Deep brain stimulation (DBS) is also approved for managing primary generalized dystonia and torticollis under a humanitarian device exemption.

Deep brain stimulation has largely replaced ablative procedures such as thalamotomy and pallidotomy. While ablative procedures can be effective for the symptoms of movement disorders, they cause a permanent lesion in the targeted nuclei and are therefore not reversible. DBS is considered safer because it can be adjusted over time and the location of the leads can be revised.¹ On the other hand, regular maintenance of implanted hardware may be considered a disadvantage of DBS.

HARDWARE AND TARGETS

While ablative procedures do not require implantable hardware, DBS consists of permanently implanted neurostimulation systems. The battery-powered pulse generators typically last for several years but require multiple replacements during a lifetime. In addition, if other hardware components fail, surgical revision may be required to maintain treatment efficacy. Surgery involving implantation of hardware carries a higher risk of infection than does a nonimplantation procedure. If infections occur, removal of the hardware is often required, with reimplantation performed after the infection clears. In addition, the expense of DBS hardware may limit availability in some cases.

Three components

Target nuclei

Several nodes or nuclei can serve as targets for DBS. In patients with PD, the most common surgical target is the subthalamic nucleus (STN), either unilaterally or bilaterally.² The globus pallidus pars interna (GPi) is also a viable target and is preferred for some patients with PD. The most common target for managing essential tremor is the ventral intermediate nucleus (VIM) of the thalamus, which can also be the target of choice for patients with tremor-predominant PD. However, the GPi and STN are usually preferred over the VIM in patients with PD because stimulation of these targets can relieve symptoms other than tremor, such as rigidity and bradykinesia. Bilateral stimulation of the GPi is the most frequent approach in patients with generalized torsion dystonia and torticollis, although the STN and thalamic nuclei (off-label) are also considered options.

PATIENT SELECTION

Patients are evaluated in our center at Cleveland Clinic by a multidisciplinary team that includes a movement disorder neurologist, a subspecialized neurosurgeon, a movement disorder neuropsychologist, and a psychiatrist with special interest in the behavioral comorbidities of movement disorders.³ Neuroimaging is included in this assessment. We have also included physical therapy as part of the initial evaluation in order to gain insight into the patient’s limitations and develop rehabilitation strategies that may enhance the outcomes of surgery or provide alternatives should surgery not be indicated. This evaluation provides extensive data that are then reviewed by the team in a conference dedicated to discussing candidacy for DBS or options for managing the symptoms of advanced movement disorders. Behavioral and cognitive issues are assessed in detail and, in our experience, are the most common reasons for not recommending DBS.

An important part of the evaluation of patients with PD is a formal test with rating of the motor section of the Unified Parkinson’s Disease Rating Scale (UPDRS) with the patient off medications for 8 to 12 hours and then after a test dose of levodopa. At our center, this off/on test is videotaped so that the responsiveness of individual symptoms to levodopa can be reviewed later in conference.

Risk of cognitive decline

While DBS is considered safe and effective, there is a risk of cognitive decline in some patients. In most patients, long-term stimulation-related cognitive decline may be detected with formal measures but is not clinically significant and is outweighed by the motor and quality-of-life benefits of surgery. In some patients, long-term cognitive decline can be significant and can limit function. Cognitive neuropsychologic testing provides valuable information in this regard. Patients with preserved cognitive function seldom experience significant decline with DBS while those with substantial baseline impairment are thought to be at greater risk. Patients who meet criteria for dementia are usually not considered candidates for DBS, but exceptions exist. Transient perioperative cognitive difficulties are more common than persistent deficits, and typically resolve within a few weeks (see “Complications of deep brain stimulation”).

Benefits in Parkinson disease

Deep brain stimulation can address several symptoms of PD but with varying effects. Tremor, rigidity, and bradykinesia usually improve substantially. Gait has a more variable response, and balance is typically refractory. A general rule is that symptoms that improve with a single dose of levodopa should also improve with DBS. (Tremor, however, will most often respond to DBS even if refractory to medication.) Good candidates for surgery typically have a greater than 30% improvement in UPDRS motor score with levodopa challenge, but sometimes, improvement in the total score is less informative than evaluation of the effects of levodopa on particular symptoms. Treatment effects can be compared with the patient’s expectations for surgery in order to infer whether the goals for symptom improvement are realistic.

After programming, DBS can provide PD symptom control similar to that of medication “on time,” but with fewer on-off fluctuations and less on-time dyskinesia. Good surgical candidates are patients who once responded well to dopaminergic medications but who, after several years with the disease, present with increased duration of “off time,” unpredictable duration of on time, and medication side effects such as on-time dyskinesia. Patients who do not respond well to levodopa even in subscores of the UPDRS may not be good candidates for DBS, and in some cases the diagnosis itself needs to be reviewed.

Deep brain stimulation can improve quality of life and alleviate symptoms of essential tremor. Tremor control is best for the upper extremities and tends to be better for distal tremors than for proximal ones. Patients who are good candidates for surgery often have severe tremors. A substantial improvement in these symptoms often has a dramatic, positive effect on work and quality of life. In some patients, surgery is considered for mild tremor if it seriously disrupts the patient’s lifestyle or occupation and cannot be well controlled with medications. Often, in these cases, tremor that appears relatively mild to the examiner is significantly limiting for the patient.

Very severe and proximal tremor is more refractory, though it may also improve. The changes can be well documented with objective measures. In these cases, however, residual tremor can still be moderate to severe and can be functionally limiting. Head or vocal tremors are typically refractory. They may be improved with bilateral implantation, but this cannot be accurately predicted. Patients who present with head-only or head-predominant tremor are thought to be less likely to benefit than those with limb tremor. Nonetheless, tremors of the head can severely impair quality of life. Because there are few other treatment options, some patients choose DBS with the understanding that the outcome is uncertain and the benefit may be limited.

Tremor resulting from multiple sclerosis or other causes can be medically refractory and disabling. In our experience, DBS can be an off-label option for managing secondary tremors and good outcomes have been observed. However, outcomes are much less predictable and tremor control less effective than in patients with essential tremor.

Patients with primary generalized dystonia can be considered candidates for DBS and may experience improved symptom control and quality of life.⁴ Patients with the DYT1 mutation are more likely to respond well to DBS, as are those with other forms of primary generalized dystonia. In contrast to that seen in patients with PD and tremor, symptomatic improvement is frequently not observed during intraoperative testing. Several months of stimulation and programming may be required before significant improvements are detected.⁵ Surgery can also be considered for off-label use in the treatment of patients with secondary dystonia—such as that following injury or associated with cerebral palsy—but outcomes are less predictable and usually more limited. A possible exception may be seen in cases of tardive dystonia, for which there is increasing evidence⁶ for the effectiveness of DBS. This remains an off-label use of DBS.

Realistic expectations

An important aspect of the multidisciplinary evaluation includes a discussion of the expectations for surgery, the risks, and the requirements for postoperative care. As discussed above, DBS is reversible and adjustable, so outcomes depend not only on accurate implantation of the hardware but also on postoperative programming. Also, monitoring and maintenance of the implanted hardware are required in these patients. It is important that patients and families appreciate the fact that specialized, long-term postoperative follow-up is as much a part of the treatment as is the implantation itself.

UNILATERAL VERSUS BILATERAL DBS

Most patients with generalized dystonia undergo bilateral DBS. However, patients with PD or essential tremor may receive bilateral, staged, or unilateral implants. Some patients with PD present with either near-complete predominance of symptoms on one side or with symptoms that affect mostly the dominant extremity. In these patients, unilateral implantation is often recommended because it has less risk than the bilateral approach and may be sufficient to address the most limiting symptoms.

As the disease advances, an additional surgery may be required to accomplish bilateral symptom control. Nevertheless, we do not routinely recommend preventive implantation because it is not known whether second-side symptoms will become severe enough to require it. This strategy allows for deferring surgical risk, which is in itself advantageous. In our experience, bilateral implantation is often recommended to PD patients who present with symptoms such as freezing of gait.

Patients who have essential tremor often present with bilateral symptoms. Although many patients will indicate that they need symptom relief on both upper extremities in order to perform activities of daily living, our practice is to recommend surgery on one side at first and to suggest the patient consider contralateral implantation after weeks or months. Bilateral implantation may carry a risk for dysarthria and the risk is thought to be reduced if bilateral procedures are staged. Although high rates of dysarthria have been reported following bilateral surgery for tremor, its occurrence has been infrequent in our experience with bilateral staged DBS. Benefits of treating tremor in the dominant extremity usually exceed those of treating nondominant tremor, so most patients prefer that the dominant side be the first one treated.

TECHNICAL OPTIONS

There are several technical options for implantation of DBS systems. Stereotactic procedures rely on co-registration of preoperative imaging with external and internal fiducials, or points of reference. Targeting of the intended structures is performed by combining direct and indirect methods. Direct methods rely on identification of the target structures with imaging, such as visualization of the STN and GPi on preoperative magnetic resonance imaging (MRI). Indirect targeting relies on cadaveric anatomic atlases and coordinate systems that infer the location of the intended structures in relation to anatomical points of reference.

Frame-based systems

Frameless systems

The key advantage of the frameless system over the frame-based system is greater mobility of the head. Another important advantage is easier access to the airway, should an emergency situation occur. In our practice, patients with experience of both frameless and frame-based systems did not report significantly less discomfort with the frameless system.

The frameless system also has disadvantages, including less secure fixation of the head, which can add risk to the procedure. In addition, because of its lightweight, plastic construction, it provides less robust support to the instrumentation entering the brain than do metallic head frames and, in some cases, there is less flexibility for adjusting targets if needed during surgery. In addition, frameless systems are nonreusable and represent a substantial additional cost.

Physiologic verification of anatomic targets identified by imaging can be accomplished with microelectrode recording (MER). This technique involves placing fine, high-impedance electrodes through the target area, so that anatomic structures can be recognized by characteristic electrical activity of individual neurons or groups of neurons. The locations of the structures are identified and the lengths of the electrode trajectories through the different structures—as well as the gaps between these structures—are recorded. The distances are then compared with the anatomy and a best-fit model is created to infer the location of the trajectory in the target area. Additional MER penetrations are made in order to further delineate the anatomy. Once a location for implantation has been selected, the DBS lead is inserted into the target area.

Electrode implantation

Lead implantation is often performed under fluoroscopic guidance in order to ensure accuracy and stability. When implanted, the electrode may cause a microlesional effect, manifested by transient improvement in symptoms.

The DBS leads are then connected to external pulse generators and assessed for clinical benefits and side effects. Amplitude, pulse width, and frequency are adjusted to test the therapeutic window of stimulation (clinical improvement thresholds versus side effect thresholds). Some PD patients develop dyskinesia during test stimulation, which may be a positive indicator for lead location. If good effects and a therapeutic window are observed, the location of the lead is considered to be satisfactory and the procedure is completed.

Pulse generator implantation

During the final step of surgery, performed under general anesthesia, the pulse generator is implanted. The extension cable that connects the DBS lead to the implantable pulse generator is tunneled subcutaneously, connecting the DBS lead to the pulse generator in the chest.

Intraoperative, real-time MRI stereotaxis

Real-time intraoperative MRI has become available for DBS implantation with devices recently cleared for use by the FDA. The procedure, typically performed in a diagnostic MRI suite, uses MR images acquired during surgery to guide DBS lead implantation in the target area and to verify implantation accuracy.⁸

Implantation of a deep brain stimulator is the most common surgical procedure performed in the United States and industrialized world for the management of advanced movement disorders. These procedures are US Food and Drug Administration (FDA)–approved for the management of the symptoms of Parkinson disease (PD) and essential tremor. Deep brain stimulation (DBS) is also approved for managing primary generalized dystonia and torticollis under a humanitarian device exemption.

Deep brain stimulation has largely replaced ablative procedures such as thalamotomy and pallidotomy. While ablative procedures can be effective for the symptoms of movement disorders, they cause a permanent lesion in the targeted nuclei and are therefore not reversible. DBS is considered safer because it can be adjusted over time and the location of the leads can be revised.¹ On the other hand, regular maintenance of implanted hardware may be considered a disadvantage of DBS.

HARDWARE AND TARGETS

While ablative procedures do not require implantable hardware, DBS consists of permanently implanted neurostimulation systems. The battery-powered pulse generators typically last for several years but require multiple replacements during a lifetime. In addition, if other hardware components fail, surgical revision may be required to maintain treatment efficacy. Surgery involving implantation of hardware carries a higher risk of infection than does a nonimplantation procedure. If infections occur, removal of the hardware is often required, with reimplantation performed after the infection clears. In addition, the expense of DBS hardware may limit availability in some cases.

Three components

Target nuclei

Several nodes or nuclei can serve as targets for DBS. In patients with PD, the most common surgical target is the subthalamic nucleus (STN), either unilaterally or bilaterally.² The globus pallidus pars interna (GPi) is also a viable target and is preferred for some patients with PD. The most common target for managing essential tremor is the ventral intermediate nucleus (VIM) of the thalamus, which can also be the target of choice for patients with tremor-predominant PD. However, the GPi and STN are usually preferred over the VIM in patients with PD because stimulation of these targets can relieve symptoms other than tremor, such as rigidity and bradykinesia. Bilateral stimulation of the GPi is the most frequent approach in patients with generalized torsion dystonia and torticollis, although the STN and thalamic nuclei (off-label) are also considered options.

PATIENT SELECTION

Patients are evaluated in our center at Cleveland Clinic by a multidisciplinary team that includes a movement disorder neurologist, a subspecialized neurosurgeon, a movement disorder neuropsychologist, and a psychiatrist with special interest in the behavioral comorbidities of movement disorders.³ Neuroimaging is included in this assessment. We have also included physical therapy as part of the initial evaluation in order to gain insight into the patient’s limitations and develop rehabilitation strategies that may enhance the outcomes of surgery or provide alternatives should surgery not be indicated. This evaluation provides extensive data that are then reviewed by the team in a conference dedicated to discussing candidacy for DBS or options for managing the symptoms of advanced movement disorders. Behavioral and cognitive issues are assessed in detail and, in our experience, are the most common reasons for not recommending DBS.

An important part of the evaluation of patients with PD is a formal test with rating of the motor section of the Unified Parkinson’s Disease Rating Scale (UPDRS) with the patient off medications for 8 to 12 hours and then after a test dose of levodopa. At our center, this off/on test is videotaped so that the responsiveness of individual symptoms to levodopa can be reviewed later in conference.

Risk of cognitive decline

While DBS is considered safe and effective, there is a risk of cognitive decline in some patients. In most patients, long-term stimulation-related cognitive decline may be detected with formal measures but is not clinically significant and is outweighed by the motor and quality-of-life benefits of surgery. In some patients, long-term cognitive decline can be significant and can limit function. Cognitive neuropsychologic testing provides valuable information in this regard. Patients with preserved cognitive function seldom experience significant decline with DBS while those with substantial baseline impairment are thought to be at greater risk. Patients who meet criteria for dementia are usually not considered candidates for DBS, but exceptions exist. Transient perioperative cognitive difficulties are more common than persistent deficits, and typically resolve within a few weeks (see “Complications of deep brain stimulation”).

Benefits in Parkinson disease

Deep brain stimulation can address several symptoms of PD but with varying effects. Tremor, rigidity, and bradykinesia usually improve substantially. Gait has a more variable response, and balance is typically refractory. A general rule is that symptoms that improve with a single dose of levodopa should also improve with DBS. (Tremor, however, will most often respond to DBS even if refractory to medication.) Good candidates for surgery typically have a greater than 30% improvement in UPDRS motor score with levodopa challenge, but sometimes, improvement in the total score is less informative than evaluation of the effects of levodopa on particular symptoms. Treatment effects can be compared with the patient’s expectations for surgery in order to infer whether the goals for symptom improvement are realistic.

After programming, DBS can provide PD symptom control similar to that of medication “on time,” but with fewer on-off fluctuations and less on-time dyskinesia. Good surgical candidates are patients who once responded well to dopaminergic medications but who, after several years with the disease, present with increased duration of “off time,” unpredictable duration of on time, and medication side effects such as on-time dyskinesia. Patients who do not respond well to levodopa even in subscores of the UPDRS may not be good candidates for DBS, and in some cases the diagnosis itself needs to be reviewed.

Deep brain stimulation can improve quality of life and alleviate symptoms of essential tremor. Tremor control is best for the upper extremities and tends to be better for distal tremors than for proximal ones. Patients who are good candidates for surgery often have severe tremors. A substantial improvement in these symptoms often has a dramatic, positive effect on work and quality of life. In some patients, surgery is considered for mild tremor if it seriously disrupts the patient’s lifestyle or occupation and cannot be well controlled with medications. Often, in these cases, tremor that appears relatively mild to the examiner is significantly limiting for the patient.

Very severe and proximal tremor is more refractory, though it may also improve. The changes can be well documented with objective measures. In these cases, however, residual tremor can still be moderate to severe and can be functionally limiting. Head or vocal tremors are typically refractory. They may be improved with bilateral implantation, but this cannot be accurately predicted. Patients who present with head-only or head-predominant tremor are thought to be less likely to benefit than those with limb tremor. Nonetheless, tremors of the head can severely impair quality of life. Because there are few other treatment options, some patients choose DBS with the understanding that the outcome is uncertain and the benefit may be limited.

Tremor resulting from multiple sclerosis or other causes can be medically refractory and disabling. In our experience, DBS can be an off-label option for managing secondary tremors and good outcomes have been observed. However, outcomes are much less predictable and tremor control less effective than in patients with essential tremor.

Patients with primary generalized dystonia can be considered candidates for DBS and may experience improved symptom control and quality of life.⁴ Patients with the DYT1 mutation are more likely to respond well to DBS, as are those with other forms of primary generalized dystonia. In contrast to that seen in patients with PD and tremor, symptomatic improvement is frequently not observed during intraoperative testing. Several months of stimulation and programming may be required before significant improvements are detected.⁵ Surgery can also be considered for off-label use in the treatment of patients with secondary dystonia—such as that following injury or associated with cerebral palsy—but outcomes are less predictable and usually more limited. A possible exception may be seen in cases of tardive dystonia, for which there is increasing evidence⁶ for the effectiveness of DBS. This remains an off-label use of DBS.

Realistic expectations

An important aspect of the multidisciplinary evaluation includes a discussion of the expectations for surgery, the risks, and the requirements for postoperative care. As discussed above, DBS is reversible and adjustable, so outcomes depend not only on accurate implantation of the hardware but also on postoperative programming. Also, monitoring and maintenance of the implanted hardware are required in these patients. It is important that patients and families appreciate the fact that specialized, long-term postoperative follow-up is as much a part of the treatment as is the implantation itself.

UNILATERAL VERSUS BILATERAL DBS

Most patients with generalized dystonia undergo bilateral DBS. However, patients with PD or essential tremor may receive bilateral, staged, or unilateral implants. Some patients with PD present with either near-complete predominance of symptoms on one side or with symptoms that affect mostly the dominant extremity. In these patients, unilateral implantation is often recommended because it has less risk than the bilateral approach and may be sufficient to address the most limiting symptoms.

As the disease advances, an additional surgery may be required to accomplish bilateral symptom control. Nevertheless, we do not routinely recommend preventive implantation because it is not known whether second-side symptoms will become severe enough to require it. This strategy allows for deferring surgical risk, which is in itself advantageous. In our experience, bilateral implantation is often recommended to PD patients who present with symptoms such as freezing of gait.

Patients who have essential tremor often present with bilateral symptoms. Although many patients will indicate that they need symptom relief on both upper extremities in order to perform activities of daily living, our practice is to recommend surgery on one side at first and to suggest the patient consider contralateral implantation after weeks or months. Bilateral implantation may carry a risk for dysarthria and the risk is thought to be reduced if bilateral procedures are staged. Although high rates of dysarthria have been reported following bilateral surgery for tremor, its occurrence has been infrequent in our experience with bilateral staged DBS. Benefits of treating tremor in the dominant extremity usually exceed those of treating nondominant tremor, so most patients prefer that the dominant side be the first one treated.

TECHNICAL OPTIONS

There are several technical options for implantation of DBS systems. Stereotactic procedures rely on co-registration of preoperative imaging with external and internal fiducials, or points of reference. Targeting of the intended structures is performed by combining direct and indirect methods. Direct methods rely on identification of the target structures with imaging, such as visualization of the STN and GPi on preoperative magnetic resonance imaging (MRI). Indirect targeting relies on cadaveric anatomic atlases and coordinate systems that infer the location of the intended structures in relation to anatomical points of reference.

Frame-based systems

Frameless systems

The key advantage of the frameless system over the frame-based system is greater mobility of the head. Another important advantage is easier access to the airway, should an emergency situation occur. In our practice, patients with experience of both frameless and frame-based systems did not report significantly less discomfort with the frameless system.

The frameless system also has disadvantages, including less secure fixation of the head, which can add risk to the procedure. In addition, because of its lightweight, plastic construction, it provides less robust support to the instrumentation entering the brain than do metallic head frames and, in some cases, there is less flexibility for adjusting targets if needed during surgery. In addition, frameless systems are nonreusable and represent a substantial additional cost.

Physiologic verification of anatomic targets identified by imaging can be accomplished with microelectrode recording (MER). This technique involves placing fine, high-impedance electrodes through the target area, so that anatomic structures can be recognized by characteristic electrical activity of individual neurons or groups of neurons. The locations of the structures are identified and the lengths of the electrode trajectories through the different structures—as well as the gaps between these structures—are recorded. The distances are then compared with the anatomy and a best-fit model is created to infer the location of the trajectory in the target area. Additional MER penetrations are made in order to further delineate the anatomy. Once a location for implantation has been selected, the DBS lead is inserted into the target area.

Electrode implantation

Lead implantation is often performed under fluoroscopic guidance in order to ensure accuracy and stability. When implanted, the electrode may cause a microlesional effect, manifested by transient improvement in symptoms.

The DBS leads are then connected to external pulse generators and assessed for clinical benefits and side effects. Amplitude, pulse width, and frequency are adjusted to test the therapeutic window of stimulation (clinical improvement thresholds versus side effect thresholds). Some PD patients develop dyskinesia during test stimulation, which may be a positive indicator for lead location. If good effects and a therapeutic window are observed, the location of the lead is considered to be satisfactory and the procedure is completed.

Pulse generator implantation

During the final step of surgery, performed under general anesthesia, the pulse generator is implanted. The extension cable that connects the DBS lead to the implantable pulse generator is tunneled subcutaneously, connecting the DBS lead to the pulse generator in the chest.

Intraoperative, real-time MRI stereotaxis

Real-time intraoperative MRI has become available for DBS implantation with devices recently cleared for use by the FDA. The procedure, typically performed in a diagnostic MRI suite, uses MR images acquired during surgery to guide DBS lead implantation in the target area and to verify implantation accuracy.⁸

Implantation of a deep brain stimulator is the most common surgical procedure performed in the United States and industrialized world for the management of advanced movement disorders. These procedures are US Food and Drug Administration (FDA)–approved for the management of the symptoms of Parkinson disease (PD) and essential tremor. Deep brain stimulation (DBS) is also approved for managing primary generalized dystonia and torticollis under a humanitarian device exemption.

Deep brain stimulation has largely replaced ablative procedures such as thalamotomy and pallidotomy. While ablative procedures can be effective for the symptoms of movement disorders, they cause a permanent lesion in the targeted nuclei and are therefore not reversible. DBS is considered safer because it can be adjusted over time and the location of the leads can be revised.¹ On the other hand, regular maintenance of implanted hardware may be considered a disadvantage of DBS.

HARDWARE AND TARGETS

While ablative procedures do not require implantable hardware, DBS consists of permanently implanted neurostimulation systems. The battery-powered pulse generators typically last for several years but require multiple replacements during a lifetime. In addition, if other hardware components fail, surgical revision may be required to maintain treatment efficacy. Surgery involving implantation of hardware carries a higher risk of infection than does a nonimplantation procedure. If infections occur, removal of the hardware is often required, with reimplantation performed after the infection clears. In addition, the expense of DBS hardware may limit availability in some cases.

Three components

Target nuclei

Several nodes or nuclei can serve as targets for DBS. In patients with PD, the most common surgical target is the subthalamic nucleus (STN), either unilaterally or bilaterally.² The globus pallidus pars interna (GPi) is also a viable target and is preferred for some patients with PD. The most common target for managing essential tremor is the ventral intermediate nucleus (VIM) of the thalamus, which can also be the target of choice for patients with tremor-predominant PD. However, the GPi and STN are usually preferred over the VIM in patients with PD because stimulation of these targets can relieve symptoms other than tremor, such as rigidity and bradykinesia. Bilateral stimulation of the GPi is the most frequent approach in patients with generalized torsion dystonia and torticollis, although the STN and thalamic nuclei (off-label) are also considered options.

PATIENT SELECTION

Patients are evaluated in our center at Cleveland Clinic by a multidisciplinary team that includes a movement disorder neurologist, a subspecialized neurosurgeon, a movement disorder neuropsychologist, and a psychiatrist with special interest in the behavioral comorbidities of movement disorders.³ Neuroimaging is included in this assessment. We have also included physical therapy as part of the initial evaluation in order to gain insight into the patient’s limitations and develop rehabilitation strategies that may enhance the outcomes of surgery or provide alternatives should surgery not be indicated. This evaluation provides extensive data that are then reviewed by the team in a conference dedicated to discussing candidacy for DBS or options for managing the symptoms of advanced movement disorders. Behavioral and cognitive issues are assessed in detail and, in our experience, are the most common reasons for not recommending DBS.

An important part of the evaluation of patients with PD is a formal test with rating of the motor section of the Unified Parkinson’s Disease Rating Scale (UPDRS) with the patient off medications for 8 to 12 hours and then after a test dose of levodopa. At our center, this off/on test is videotaped so that the responsiveness of individual symptoms to levodopa can be reviewed later in conference.

Risk of cognitive decline

While DBS is considered safe and effective, there is a risk of cognitive decline in some patients. In most patients, long-term stimulation-related cognitive decline may be detected with formal measures but is not clinically significant and is outweighed by the motor and quality-of-life benefits of surgery. In some patients, long-term cognitive decline can be significant and can limit function. Cognitive neuropsychologic testing provides valuable information in this regard. Patients with preserved cognitive function seldom experience significant decline with DBS while those with substantial baseline impairment are thought to be at greater risk. Patients who meet criteria for dementia are usually not considered candidates for DBS, but exceptions exist. Transient perioperative cognitive difficulties are more common than persistent deficits, and typically resolve within a few weeks (see “Complications of deep brain stimulation”).

Benefits in Parkinson disease

Deep brain stimulation can address several symptoms of PD but with varying effects. Tremor, rigidity, and bradykinesia usually improve substantially. Gait has a more variable response, and balance is typically refractory. A general rule is that symptoms that improve with a single dose of levodopa should also improve with DBS. (Tremor, however, will most often respond to DBS even if refractory to medication.) Good candidates for surgery typically have a greater than 30% improvement in UPDRS motor score with levodopa challenge, but sometimes, improvement in the total score is less informative than evaluation of the effects of levodopa on particular symptoms. Treatment effects can be compared with the patient’s expectations for surgery in order to infer whether the goals for symptom improvement are realistic.

After programming, DBS can provide PD symptom control similar to that of medication “on time,” but with fewer on-off fluctuations and less on-time dyskinesia. Good surgical candidates are patients who once responded well to dopaminergic medications but who, after several years with the disease, present with increased duration of “off time,” unpredictable duration of on time, and medication side effects such as on-time dyskinesia. Patients who do not respond well to levodopa even in subscores of the UPDRS may not be good candidates for DBS, and in some cases the diagnosis itself needs to be reviewed.

Deep brain stimulation can improve quality of life and alleviate symptoms of essential tremor. Tremor control is best for the upper extremities and tends to be better for distal tremors than for proximal ones. Patients who are good candidates for surgery often have severe tremors. A substantial improvement in these symptoms often has a dramatic, positive effect on work and quality of life. In some patients, surgery is considered for mild tremor if it seriously disrupts the patient’s lifestyle or occupation and cannot be well controlled with medications. Often, in these cases, tremor that appears relatively mild to the examiner is significantly limiting for the patient.

Very severe and proximal tremor is more refractory, though it may also improve. The changes can be well documented with objective measures. In these cases, however, residual tremor can still be moderate to severe and can be functionally limiting. Head or vocal tremors are typically refractory. They may be improved with bilateral implantation, but this cannot be accurately predicted. Patients who present with head-only or head-predominant tremor are thought to be less likely to benefit than those with limb tremor. Nonetheless, tremors of the head can severely impair quality of life. Because there are few other treatment options, some patients choose DBS with the understanding that the outcome is uncertain and the benefit may be limited.

Tremor resulting from multiple sclerosis or other causes can be medically refractory and disabling. In our experience, DBS can be an off-label option for managing secondary tremors and good outcomes have been observed. However, outcomes are much less predictable and tremor control less effective than in patients with essential tremor.

Patients with primary generalized dystonia can be considered candidates for DBS and may experience improved symptom control and quality of life.⁴ Patients with the DYT1 mutation are more likely to respond well to DBS, as are those with other forms of primary generalized dystonia. In contrast to that seen in patients with PD and tremor, symptomatic improvement is frequently not observed during intraoperative testing. Several months of stimulation and programming may be required before significant improvements are detected.⁵ Surgery can also be considered for off-label use in the treatment of patients with secondary dystonia—such as that following injury or associated with cerebral palsy—but outcomes are less predictable and usually more limited. A possible exception may be seen in cases of tardive dystonia, for which there is increasing evidence⁶ for the effectiveness of DBS. This remains an off-label use of DBS.

Realistic expectations

An important aspect of the multidisciplinary evaluation includes a discussion of the expectations for surgery, the risks, and the requirements for postoperative care. As discussed above, DBS is reversible and adjustable, so outcomes depend not only on accurate implantation of the hardware but also on postoperative programming. Also, monitoring and maintenance of the implanted hardware are required in these patients. It is important that patients and families appreciate the fact that specialized, long-term postoperative follow-up is as much a part of the treatment as is the implantation itself.

UNILATERAL VERSUS BILATERAL DBS

Most patients with generalized dystonia undergo bilateral DBS. However, patients with PD or essential tremor may receive bilateral, staged, or unilateral implants. Some patients with PD present with either near-complete predominance of symptoms on one side or with symptoms that affect mostly the dominant extremity. In these patients, unilateral implantation is often recommended because it has less risk than the bilateral approach and may be sufficient to address the most limiting symptoms.

As the disease advances, an additional surgery may be required to accomplish bilateral symptom control. Nevertheless, we do not routinely recommend preventive implantation because it is not known whether second-side symptoms will become severe enough to require it. This strategy allows for deferring surgical risk, which is in itself advantageous. In our experience, bilateral implantation is often recommended to PD patients who present with symptoms such as freezing of gait.

Patients who have essential tremor often present with bilateral symptoms. Although many patients will indicate that they need symptom relief on both upper extremities in order to perform activities of daily living, our practice is to recommend surgery on one side at first and to suggest the patient consider contralateral implantation after weeks or months. Bilateral implantation may carry a risk for dysarthria and the risk is thought to be reduced if bilateral procedures are staged. Although high rates of dysarthria have been reported following bilateral surgery for tremor, its occurrence has been infrequent in our experience with bilateral staged DBS. Benefits of treating tremor in the dominant extremity usually exceed those of treating nondominant tremor, so most patients prefer that the dominant side be the first one treated.

TECHNICAL OPTIONS

There are several technical options for implantation of DBS systems. Stereotactic procedures rely on co-registration of preoperative imaging with external and internal fiducials, or points of reference. Targeting of the intended structures is performed by combining direct and indirect methods. Direct methods rely on identification of the target structures with imaging, such as visualization of the STN and GPi on preoperative magnetic resonance imaging (MRI). Indirect targeting relies on cadaveric anatomic atlases and coordinate systems that infer the location of the intended structures in relation to anatomical points of reference.

Frame-based systems

Frameless systems

The key advantage of the frameless system over the frame-based system is greater mobility of the head. Another important advantage is easier access to the airway, should an emergency situation occur. In our practice, patients with experience of both frameless and frame-based systems did not report significantly less discomfort with the frameless system.

The frameless system also has disadvantages, including less secure fixation of the head, which can add risk to the procedure. In addition, because of its lightweight, plastic construction, it provides less robust support to the instrumentation entering the brain than do metallic head frames and, in some cases, there is less flexibility for adjusting targets if needed during surgery. In addition, frameless systems are nonreusable and represent a substantial additional cost.

Physiologic verification of anatomic targets identified by imaging can be accomplished with microelectrode recording (MER). This technique involves placing fine, high-impedance electrodes through the target area, so that anatomic structures can be recognized by characteristic electrical activity of individual neurons or groups of neurons. The locations of the structures are identified and the lengths of the electrode trajectories through the different structures—as well as the gaps between these structures—are recorded. The distances are then compared with the anatomy and a best-fit model is created to infer the location of the trajectory in the target area. Additional MER penetrations are made in order to further delineate the anatomy. Once a location for implantation has been selected, the DBS lead is inserted into the target area.

Electrode implantation

Lead implantation is often performed under fluoroscopic guidance in order to ensure accuracy and stability. When implanted, the electrode may cause a microlesional effect, manifested by transient improvement in symptoms.

The DBS leads are then connected to external pulse generators and assessed for clinical benefits and side effects. Amplitude, pulse width, and frequency are adjusted to test the therapeutic window of stimulation (clinical improvement thresholds versus side effect thresholds). Some PD patients develop dyskinesia during test stimulation, which may be a positive indicator for lead location. If good effects and a therapeutic window are observed, the location of the lead is considered to be satisfactory and the procedure is completed.

Pulse generator implantation

During the final step of surgery, performed under general anesthesia, the pulse generator is implanted. The extension cable that connects the DBS lead to the implantable pulse generator is tunneled subcutaneously, connecting the DBS lead to the pulse generator in the chest.

Intraoperative, real-time MRI stereotaxis

Real-time intraoperative MRI has become available for DBS implantation with devices recently cleared for use by the FDA. The procedure, typically performed in a diagnostic MRI suite, uses MR images acquired during surgery to guide DBS lead implantation in the target area and to verify implantation accuracy.⁸

Dystonia is a movement disorder in which involuntary sustained muscle contractions cause twisting movements that place the body in abnormal, sometimes painful, positions. Dystonia is believed to arise from an abnormality in the basal ganglia and an inherent or acquired defect in the processing of neurotransmitters.¹

Dystonia is uncommon, although its exact prevalence is unknown. Nutt et al concluded that at least 250,000 people were affected by idiopathic dystonia in the United States, but prevalence is likely higher because misdiagnosis is not uncommon.² A more recent European study found the prevalence of primary dystonia in the general population aged 50 years or more to be 732 per 100,000.³ The Epidemiological Study of Dystonia in Europe (ESDE) Collaborative Group found that the estimated prevalence of cervical dystonia was 50 to 200 per 1 million individuals.⁴ Also known as spasmodic torticollis, this is the most commonly diagnosed form of focal dystonia.

CLASSIFICATION OF DYSTONIA

Accurate classification of dystonia is important, since this informs approaches to management as well as prognosis. The three most important means by which dystonia is classified are (1) etiology, including primary dystonia, which encompasses a variety of genetic variables, and secondary dystonia; (2) bodily distribution of symptoms; and (3) age at onset.

Etiology

Most primary or idiopathic dystonia appears to be hereditary. Early-onset primary dystonia is most frequently caused by a mutation in the DYT1 gene, although other genetic mutations are possible.⁵ Patients with primary dystonia have no other underlying disorder; involuntary muscle contractions are the sole symptom. A thorough history should include a review of perinatal and early developmental history, prior neurologic illness, and exposure to drugs known to cause acquired dystonia. Physical examinations (encompassing intellectual, pyramidal, cerebellar, and sensory domains) and laboratory tests reveal no specific cause for the dystonic symptoms. Primary dystonia is also most frequently action-induced; at rest, the affected body region may appear to be normal.

Secondary dystonia occurs as a symptom of another disease process. Multiple sclerosis or any one of several hereditary neurologic disorders, such as Wilson disease, may be implicated. Secondary dystonia also may result from trauma to the brain, as might occur during an automobile accident; from heavy-metal or carbon monoxide poisoning; or as an adverse effect of medication. It may be psychogenic or related to Parkinson disease or Parkinson-plus syndromes, a group of neurodegenerative disorders with parkinsonian features. Tardive dystonia, the most common adult form of secondary dystonia, may occur follow ing exposure to certain neuroleptic drugs; tardive dystonia is a type of tardive dyskinesia that describes any involuntary neurologic movement disorder.

Bodily distribution

Dystonia is further classified by location of symptoms. Focal dystonias, which are usually primary dystonias, describe symptoms that are limited to a region of the body, such as a specific arm. There are several variations. Cervical dystonia affects the head and neck, is the most common adult-onset dystonia, and affects more women than men. Blepharospasm, or involuntary contractions of the eyelids, potentially leads to extended eye closure and functional blindness and often involves other facial muscles. Laryngeal dystonia affects the muscles in the larynx. Limb dystonia, such as writer’s or musician’s cramp, affects muscles in the arm, hand, leg, or foot. Limb dystonia is often task-specific action dystonia, and can be primary or secondary.

Segmental dystonia describes a group of involved muscles that are contiguous, such as cranial to neck to cervical to arm. Oromandibular dystonia, affecting the face, mouth, and jaw, often with unusual tongue movements (ie, lingual dystonia), is a type of segmental dystonia, although some consider it a focal dystonia. Meige syndrome is the combination of blepharospasm and oromandibular dystonia. Certain limb and cranial dystonias are considered segmental dystonias. Dystonia that affects two or more noncontiguous muscle groups in different parts of the body is multifocal. Hemi dystonia describes unilateral symptoms.

Symptoms that have advanced from a focal presentation to affect additional regions of the body characterize generalized dystonia. The symptoms potentially advance to include the trunk and limbs. The muscular contractions are usually sustained, are often both repetitive and painful, and worsen with activity.⁶ In severe cases, muscular contractions may occur even while resting. Early-onset myoclonus dystonia is a generalized hereditary dystonia whose symptoms include dystonic contractions of the neck and shoulders and rapid jerking movements.⁷ Of note diagnostically, early-onset dystonia in a leg typically begins at age 8 to 9 years and is more likely than other early-onset presentations to progress to generalized dystonia. Early-onset dystonia that begins in an arm typically presents later, at age 12 to 14 years, and is less likely to progress to generalized dystonia. Late-onset dystonia (> 27 years of age), by contrast, rarely begins in a leg and tends to remain either focal or segmental.⁸

Age of onset

A third useful classification scheme identifies early-onset (childhood to young adult) and late-onset varieties of dystonia.

THE DIAGNOSTIC CHALLENGE

Consider primary dystonia if perinatal and developmental histories, intellect, strength, and perception of sensations are normal. There should be no prior history of neurologic illness or exposure to neuroleptic drugs whose adverse effects include secondary dystonia. In primary dystonia, diagnostic studies are negative and dystonia is the only symptom. If onset of symptoms is associated with activity, then primary dystonia should be considered. In the case of early- or late-onset limb dystonia, testing should be performed for the DYT1 gene. If the results are negative, then a trial for dopa-responsive dystonia should be undertaken with levodopa.

Consider secondary dystonia if the patient has been exposed to neuroleptic drugs, symptoms are distributed unilaterally, or the presentation is unusual for age or distribution of symptoms. For example, cranial dystonia in a child would raise the index of suspicion for secondary dystonia. If tardive dystonia is part of the differential diagnosis, consider magnetic resonance imaging (MRI), serum ceruloplasmin measurement, or slit-lamp diagnostic testing. Suspicion of a structural lesion affecting the central nervous system warrants examination with MRI, computed tomography, or angiography. Certain metabolic and neurologic hereditary disorders cause secondary dystonia, in which case dopa-responsive dystonia should be ruled out. Psychometric testing should also be considered.

SYMPTOMATIC TREATMENT WITH CHEMODENERVATION

In the absence of a cure, treatment options for dystonia are necessarily symptomatic and supportive. Titratable chemo denervation agents are injected directly into the muscle or motor nerve, temporarily weakening the local muscle and easing dystonia symptoms. Chemo denervation agents include phenol, ethyl alcohol, and botulinum toxin types A (BTX-A; onabotulinumtoxinA, abo botulinumtoxinA, and incobotulinumtoxinA) and B (BTX-B; rimabotulinum toxinB).

Phenol and ethyl alcohol injections targeted perineurally or as a motor point block have been employed for dystonia and cause nonselective tissue destruction, muscle necrosis, and highly variable durations of response. Perineural microcirculation may be damaged, possibly leading to long-term defects.

Clostridium botulinum bacteria produce seven serologically distinct neuroparalytic toxins. They are the most powerful such toxins currently known and temporarily prevent acetylcholine vesicles from docking into the presynaptic neuromuscular junction. Use of BTX-A for treatment of dystonia was recommended in a National Institutes of Health consensus statement in 1990.⁹ It has been studied for a variety of dystonias, including blepharospasm, hemifacial spasm, laryngeal dystonia, oromandibular dystonia, and cervical dystonia, among other focal dystonias. Lew et al reported in 1997 on the successful use of BTX-B for cervical dystonia in a double-blind, single-treatment study,¹⁰ and confirmatory studies followed.^11,12

Varying indications for botulinum toxin

US Food and Drug Administration–approved indications for the toxins vary. The three BTX-A products and the single BTX-B product are approved for the treatment of cervical dystonia in adults to reduce the severity of abnormal head position and neck pain. OnabotulinumtoxinA is approved for treatment of blepharospasm and strabismus associated with dystonia; and incobotulinumtoxinA is approved for blepharospasm in patients who have previously been treated with onabotulinumtoxinA. BTX-A has also been found to be safe and effective for the management of focal dystonias. These botulinum toxin agents are not equivalent in dosing units, so caution must be observed when switching brands.

Patients selected to receive BTX for dystonia should meet three criteria:

• The dystonia should interfere with their functioning, comfort, or care to the degree that causes impairment and affects activities of daily living;
• Focal weakening following administration of the drug should not decrease their level of function; and
• The patient should understand that use of BTX may not completely address positioning, posturing, or secondary deformities.

Contraindications include pregnancy, lactation, comorbid neuromuscular disease (eg, amyotrophic lateral sclerosis or myasthenia gravis), and use of an aminoglycoside.

The need for BTX therapy should be reevaluated prior to each treatment; clinical benefit lasts 3 months or more. Electromyography may facilitate the location of target muscles, particularly since involved musculature may not be palpable and is often not superficial.¹³ In-office tools that help document baseline and posttreatment results, including videotaping dystonic limb movements and the use of rating scales, can be important for evaluating the patient’s progress.¹⁴

Relief for cervical dystonia

The treatment of choice for focal dystonias and focal aspects of generalized dystonia is BTX. Both BTX-A and BTX-B offer effective palliative treatments for cervical dystonia by improving neck position, reducing pain, and decreasing disability in sufferers.^11,15–18 The BTX solution is injected directly into the dystonic muscle at several locations, temporarily weakening the overactive muscle. The BTX dose is approximately proportional to the size of the muscle, although smaller muscles typically responsible for precision movement may require a relatively larger dose (Table 2). Doses may be modified according to clinical factors such as muscle bulk and severity of dystonia (Table 3).

Relief following BTX injection for cervical dystonia occurs about 1 week later, with the greatest effect seen at about 2 to 6 weeks following injection; relief may last 12 to 16 weeks. Reinjections are not normally administered prior to 12 weeks’ duration in order to reduce the possibility of antibody formation. Concomitant interventions addressing depression and anxiety may have a significant effect on overall quality of life.¹⁹ Patients may also try several sensory tricks, called gestes antagoniste, which may temporarily reduce or alleviate the dystonia. However, these tactile procedures—such as placing a hand on top of the head—lose their effectiveness over time.

Treatment of blepharospasm, focal limb dystonia

The use of BTX-A for blepharospasm is a significant improvement over the former clinical reliance on various oral medications, which, with the exception of baclofen, proved largely ineffective.²⁰ Surgical treatments result in damage to muscular and nervous tissues, and so are reserved only for nonresponders to BTX-A therapy.²¹

BTX-A can provide effective relief and is the treatment of choice for focal limb dystonias.²² Goals of treatment include functional improvement, correction of abnormal posture, and relief from discomfort. Although a variety of oral medications may also be prescribed, drug toxicity and adverse effects can outweigh the benefit and are usually only used in cases of severe dystonia. Oral medications used for limb dystonia include anticholinergics, dopamine agonists and antagonists, baclofen, clonazepam or other benzodiazepines, and muscle relaxants.

Antibodies may bind to the drug in a small percentage of patients who regularly receive injections of BTX, rendering additional injections of that specific serotype of BTX ineffective. This immunoresistance can be avoided if clinicians inject only the smallest quantity of BTX that achieves clinical efficacy, avoid administering booster injections before the end of the minimum 12-week lockout period, and extend the period between treatments as long as possible. If immunoresistance does occur, the BTX should be exchanged for a different serotype.

Testing for nonresponse

Patients are said to be nonresponders to BTX therapy if at 4 to 6 weeks following injection they show no reduction in muscle tone. A functional test for nonresponse is to inject a small amount of BTX into either the frontalis or sternocleidomastoid muscle prior to starting treatment; asymmetric weakness demonstrates a response, indicating that either injection technique or muscle selection is the problem. In addition to the development of neutralizing antibodies, other possible reasons for nonresponse include a dose that is too low or an alteration in the pattern of muscles involved in the dystonic movement.

Dystonia is a movement disorder in which involuntary sustained muscle contractions cause twisting movements that place the body in abnormal, sometimes painful, positions. Dystonia is believed to arise from an abnormality in the basal ganglia and an inherent or acquired defect in the processing of neurotransmitters.¹

Dystonia is uncommon, although its exact prevalence is unknown. Nutt et al concluded that at least 250,000 people were affected by idiopathic dystonia in the United States, but prevalence is likely higher because misdiagnosis is not uncommon.² A more recent European study found the prevalence of primary dystonia in the general population aged 50 years or more to be 732 per 100,000.³ The Epidemiological Study of Dystonia in Europe (ESDE) Collaborative Group found that the estimated prevalence of cervical dystonia was 50 to 200 per 1 million individuals.⁴ Also known as spasmodic torticollis, this is the most commonly diagnosed form of focal dystonia.

CLASSIFICATION OF DYSTONIA

Accurate classification of dystonia is important, since this informs approaches to management as well as prognosis. The three most important means by which dystonia is classified are (1) etiology, including primary dystonia, which encompasses a variety of genetic variables, and secondary dystonia; (2) bodily distribution of symptoms; and (3) age at onset.

Etiology

Most primary or idiopathic dystonia appears to be hereditary. Early-onset primary dystonia is most frequently caused by a mutation in the DYT1 gene, although other genetic mutations are possible.⁵ Patients with primary dystonia have no other underlying disorder; involuntary muscle contractions are the sole symptom. A thorough history should include a review of perinatal and early developmental history, prior neurologic illness, and exposure to drugs known to cause acquired dystonia. Physical examinations (encompassing intellectual, pyramidal, cerebellar, and sensory domains) and laboratory tests reveal no specific cause for the dystonic symptoms. Primary dystonia is also most frequently action-induced; at rest, the affected body region may appear to be normal.

Secondary dystonia occurs as a symptom of another disease process. Multiple sclerosis or any one of several hereditary neurologic disorders, such as Wilson disease, may be implicated. Secondary dystonia also may result from trauma to the brain, as might occur during an automobile accident; from heavy-metal or carbon monoxide poisoning; or as an adverse effect of medication. It may be psychogenic or related to Parkinson disease or Parkinson-plus syndromes, a group of neurodegenerative disorders with parkinsonian features. Tardive dystonia, the most common adult form of secondary dystonia, may occur follow ing exposure to certain neuroleptic drugs; tardive dystonia is a type of tardive dyskinesia that describes any involuntary neurologic movement disorder.

Bodily distribution

Dystonia is further classified by location of symptoms. Focal dystonias, which are usually primary dystonias, describe symptoms that are limited to a region of the body, such as a specific arm. There are several variations. Cervical dystonia affects the head and neck, is the most common adult-onset dystonia, and affects more women than men. Blepharospasm, or involuntary contractions of the eyelids, potentially leads to extended eye closure and functional blindness and often involves other facial muscles. Laryngeal dystonia affects the muscles in the larynx. Limb dystonia, such as writer’s or musician’s cramp, affects muscles in the arm, hand, leg, or foot. Limb dystonia is often task-specific action dystonia, and can be primary or secondary.

Segmental dystonia describes a group of involved muscles that are contiguous, such as cranial to neck to cervical to arm. Oromandibular dystonia, affecting the face, mouth, and jaw, often with unusual tongue movements (ie, lingual dystonia), is a type of segmental dystonia, although some consider it a focal dystonia. Meige syndrome is the combination of blepharospasm and oromandibular dystonia. Certain limb and cranial dystonias are considered segmental dystonias. Dystonia that affects two or more noncontiguous muscle groups in different parts of the body is multifocal. Hemi dystonia describes unilateral symptoms.

Symptoms that have advanced from a focal presentation to affect additional regions of the body characterize generalized dystonia. The symptoms potentially advance to include the trunk and limbs. The muscular contractions are usually sustained, are often both repetitive and painful, and worsen with activity.⁶ In severe cases, muscular contractions may occur even while resting. Early-onset myoclonus dystonia is a generalized hereditary dystonia whose symptoms include dystonic contractions of the neck and shoulders and rapid jerking movements.⁷ Of note diagnostically, early-onset dystonia in a leg typically begins at age 8 to 9 years and is more likely than other early-onset presentations to progress to generalized dystonia. Early-onset dystonia that begins in an arm typically presents later, at age 12 to 14 years, and is less likely to progress to generalized dystonia. Late-onset dystonia (> 27 years of age), by contrast, rarely begins in a leg and tends to remain either focal or segmental.⁸

Age of onset

A third useful classification scheme identifies early-onset (childhood to young adult) and late-onset varieties of dystonia.

THE DIAGNOSTIC CHALLENGE

Consider primary dystonia if perinatal and developmental histories, intellect, strength, and perception of sensations are normal. There should be no prior history of neurologic illness or exposure to neuroleptic drugs whose adverse effects include secondary dystonia. In primary dystonia, diagnostic studies are negative and dystonia is the only symptom. If onset of symptoms is associated with activity, then primary dystonia should be considered. In the case of early- or late-onset limb dystonia, testing should be performed for the DYT1 gene. If the results are negative, then a trial for dopa-responsive dystonia should be undertaken with levodopa.

Consider secondary dystonia if the patient has been exposed to neuroleptic drugs, symptoms are distributed unilaterally, or the presentation is unusual for age or distribution of symptoms. For example, cranial dystonia in a child would raise the index of suspicion for secondary dystonia. If tardive dystonia is part of the differential diagnosis, consider magnetic resonance imaging (MRI), serum ceruloplasmin measurement, or slit-lamp diagnostic testing. Suspicion of a structural lesion affecting the central nervous system warrants examination with MRI, computed tomography, or angiography. Certain metabolic and neurologic hereditary disorders cause secondary dystonia, in which case dopa-responsive dystonia should be ruled out. Psychometric testing should also be considered.

SYMPTOMATIC TREATMENT WITH CHEMODENERVATION

In the absence of a cure, treatment options for dystonia are necessarily symptomatic and supportive. Titratable chemo denervation agents are injected directly into the muscle or motor nerve, temporarily weakening the local muscle and easing dystonia symptoms. Chemo denervation agents include phenol, ethyl alcohol, and botulinum toxin types A (BTX-A; onabotulinumtoxinA, abo botulinumtoxinA, and incobotulinumtoxinA) and B (BTX-B; rimabotulinum toxinB).

Phenol and ethyl alcohol injections targeted perineurally or as a motor point block have been employed for dystonia and cause nonselective tissue destruction, muscle necrosis, and highly variable durations of response. Perineural microcirculation may be damaged, possibly leading to long-term defects.

Clostridium botulinum bacteria produce seven serologically distinct neuroparalytic toxins. They are the most powerful such toxins currently known and temporarily prevent acetylcholine vesicles from docking into the presynaptic neuromuscular junction. Use of BTX-A for treatment of dystonia was recommended in a National Institutes of Health consensus statement in 1990.⁹ It has been studied for a variety of dystonias, including blepharospasm, hemifacial spasm, laryngeal dystonia, oromandibular dystonia, and cervical dystonia, among other focal dystonias. Lew et al reported in 1997 on the successful use of BTX-B for cervical dystonia in a double-blind, single-treatment study,¹⁰ and confirmatory studies followed.^11,12

Varying indications for botulinum toxin

US Food and Drug Administration–approved indications for the toxins vary. The three BTX-A products and the single BTX-B product are approved for the treatment of cervical dystonia in adults to reduce the severity of abnormal head position and neck pain. OnabotulinumtoxinA is approved for treatment of blepharospasm and strabismus associated with dystonia; and incobotulinumtoxinA is approved for blepharospasm in patients who have previously been treated with onabotulinumtoxinA. BTX-A has also been found to be safe and effective for the management of focal dystonias. These botulinum toxin agents are not equivalent in dosing units, so caution must be observed when switching brands.

Patients selected to receive BTX for dystonia should meet three criteria:

• The dystonia should interfere with their functioning, comfort, or care to the degree that causes impairment and affects activities of daily living;
• Focal weakening following administration of the drug should not decrease their level of function; and
• The patient should understand that use of BTX may not completely address positioning, posturing, or secondary deformities.

Contraindications include pregnancy, lactation, comorbid neuromuscular disease (eg, amyotrophic lateral sclerosis or myasthenia gravis), and use of an aminoglycoside.

The need for BTX therapy should be reevaluated prior to each treatment; clinical benefit lasts 3 months or more. Electromyography may facilitate the location of target muscles, particularly since involved musculature may not be palpable and is often not superficial.¹³ In-office tools that help document baseline and posttreatment results, including videotaping dystonic limb movements and the use of rating scales, can be important for evaluating the patient’s progress.¹⁴

Relief for cervical dystonia

The treatment of choice for focal dystonias and focal aspects of generalized dystonia is BTX. Both BTX-A and BTX-B offer effective palliative treatments for cervical dystonia by improving neck position, reducing pain, and decreasing disability in sufferers.^11,15–18 The BTX solution is injected directly into the dystonic muscle at several locations, temporarily weakening the overactive muscle. The BTX dose is approximately proportional to the size of the muscle, although smaller muscles typically responsible for precision movement may require a relatively larger dose (Table 2). Doses may be modified according to clinical factors such as muscle bulk and severity of dystonia (Table 3).

Relief following BTX injection for cervical dystonia occurs about 1 week later, with the greatest effect seen at about 2 to 6 weeks following injection; relief may last 12 to 16 weeks. Reinjections are not normally administered prior to 12 weeks’ duration in order to reduce the possibility of antibody formation. Concomitant interventions addressing depression and anxiety may have a significant effect on overall quality of life.¹⁹ Patients may also try several sensory tricks, called gestes antagoniste, which may temporarily reduce or alleviate the dystonia. However, these tactile procedures—such as placing a hand on top of the head—lose their effectiveness over time.

Treatment of blepharospasm, focal limb dystonia

The use of BTX-A for blepharospasm is a significant improvement over the former clinical reliance on various oral medications, which, with the exception of baclofen, proved largely ineffective.²⁰ Surgical treatments result in damage to muscular and nervous tissues, and so are reserved only for nonresponders to BTX-A therapy.²¹

BTX-A can provide effective relief and is the treatment of choice for focal limb dystonias.²² Goals of treatment include functional improvement, correction of abnormal posture, and relief from discomfort. Although a variety of oral medications may also be prescribed, drug toxicity and adverse effects can outweigh the benefit and are usually only used in cases of severe dystonia. Oral medications used for limb dystonia include anticholinergics, dopamine agonists and antagonists, baclofen, clonazepam or other benzodiazepines, and muscle relaxants.

Antibodies may bind to the drug in a small percentage of patients who regularly receive injections of BTX, rendering additional injections of that specific serotype of BTX ineffective. This immunoresistance can be avoided if clinicians inject only the smallest quantity of BTX that achieves clinical efficacy, avoid administering booster injections before the end of the minimum 12-week lockout period, and extend the period between treatments as long as possible. If immunoresistance does occur, the BTX should be exchanged for a different serotype.

Testing for nonresponse

Patients are said to be nonresponders to BTX therapy if at 4 to 6 weeks following injection they show no reduction in muscle tone. A functional test for nonresponse is to inject a small amount of BTX into either the frontalis or sternocleidomastoid muscle prior to starting treatment; asymmetric weakness demonstrates a response, indicating that either injection technique or muscle selection is the problem. In addition to the development of neutralizing antibodies, other possible reasons for nonresponse include a dose that is too low or an alteration in the pattern of muscles involved in the dystonic movement.

Dystonia is a movement disorder in which involuntary sustained muscle contractions cause twisting movements that place the body in abnormal, sometimes painful, positions. Dystonia is believed to arise from an abnormality in the basal ganglia and an inherent or acquired defect in the processing of neurotransmitters.¹

Dystonia is uncommon, although its exact prevalence is unknown. Nutt et al concluded that at least 250,000 people were affected by idiopathic dystonia in the United States, but prevalence is likely higher because misdiagnosis is not uncommon.² A more recent European study found the prevalence of primary dystonia in the general population aged 50 years or more to be 732 per 100,000.³ The Epidemiological Study of Dystonia in Europe (ESDE) Collaborative Group found that the estimated prevalence of cervical dystonia was 50 to 200 per 1 million individuals.⁴ Also known as spasmodic torticollis, this is the most commonly diagnosed form of focal dystonia.

CLASSIFICATION OF DYSTONIA

Accurate classification of dystonia is important, since this informs approaches to management as well as prognosis. The three most important means by which dystonia is classified are (1) etiology, including primary dystonia, which encompasses a variety of genetic variables, and secondary dystonia; (2) bodily distribution of symptoms; and (3) age at onset.

Etiology

Most primary or idiopathic dystonia appears to be hereditary. Early-onset primary dystonia is most frequently caused by a mutation in the DYT1 gene, although other genetic mutations are possible.⁵ Patients with primary dystonia have no other underlying disorder; involuntary muscle contractions are the sole symptom. A thorough history should include a review of perinatal and early developmental history, prior neurologic illness, and exposure to drugs known to cause acquired dystonia. Physical examinations (encompassing intellectual, pyramidal, cerebellar, and sensory domains) and laboratory tests reveal no specific cause for the dystonic symptoms. Primary dystonia is also most frequently action-induced; at rest, the affected body region may appear to be normal.

Secondary dystonia occurs as a symptom of another disease process. Multiple sclerosis or any one of several hereditary neurologic disorders, such as Wilson disease, may be implicated. Secondary dystonia also may result from trauma to the brain, as might occur during an automobile accident; from heavy-metal or carbon monoxide poisoning; or as an adverse effect of medication. It may be psychogenic or related to Parkinson disease or Parkinson-plus syndromes, a group of neurodegenerative disorders with parkinsonian features. Tardive dystonia, the most common adult form of secondary dystonia, may occur follow ing exposure to certain neuroleptic drugs; tardive dystonia is a type of tardive dyskinesia that describes any involuntary neurologic movement disorder.

Bodily distribution

Dystonia is further classified by location of symptoms. Focal dystonias, which are usually primary dystonias, describe symptoms that are limited to a region of the body, such as a specific arm. There are several variations. Cervical dystonia affects the head and neck, is the most common adult-onset dystonia, and affects more women than men. Blepharospasm, or involuntary contractions of the eyelids, potentially leads to extended eye closure and functional blindness and often involves other facial muscles. Laryngeal dystonia affects the muscles in the larynx. Limb dystonia, such as writer’s or musician’s cramp, affects muscles in the arm, hand, leg, or foot. Limb dystonia is often task-specific action dystonia, and can be primary or secondary.

Segmental dystonia describes a group of involved muscles that are contiguous, such as cranial to neck to cervical to arm. Oromandibular dystonia, affecting the face, mouth, and jaw, often with unusual tongue movements (ie, lingual dystonia), is a type of segmental dystonia, although some consider it a focal dystonia. Meige syndrome is the combination of blepharospasm and oromandibular dystonia. Certain limb and cranial dystonias are considered segmental dystonias. Dystonia that affects two or more noncontiguous muscle groups in different parts of the body is multifocal. Hemi dystonia describes unilateral symptoms.

Symptoms that have advanced from a focal presentation to affect additional regions of the body characterize generalized dystonia. The symptoms potentially advance to include the trunk and limbs. The muscular contractions are usually sustained, are often both repetitive and painful, and worsen with activity.⁶ In severe cases, muscular contractions may occur even while resting. Early-onset myoclonus dystonia is a generalized hereditary dystonia whose symptoms include dystonic contractions of the neck and shoulders and rapid jerking movements.⁷ Of note diagnostically, early-onset dystonia in a leg typically begins at age 8 to 9 years and is more likely than other early-onset presentations to progress to generalized dystonia. Early-onset dystonia that begins in an arm typically presents later, at age 12 to 14 years, and is less likely to progress to generalized dystonia. Late-onset dystonia (> 27 years of age), by contrast, rarely begins in a leg and tends to remain either focal or segmental.⁸

Age of onset

A third useful classification scheme identifies early-onset (childhood to young adult) and late-onset varieties of dystonia.

THE DIAGNOSTIC CHALLENGE

Consider primary dystonia if perinatal and developmental histories, intellect, strength, and perception of sensations are normal. There should be no prior history of neurologic illness or exposure to neuroleptic drugs whose adverse effects include secondary dystonia. In primary dystonia, diagnostic studies are negative and dystonia is the only symptom. If onset of symptoms is associated with activity, then primary dystonia should be considered. In the case of early- or late-onset limb dystonia, testing should be performed for the DYT1 gene. If the results are negative, then a trial for dopa-responsive dystonia should be undertaken with levodopa.

Consider secondary dystonia if the patient has been exposed to neuroleptic drugs, symptoms are distributed unilaterally, or the presentation is unusual for age or distribution of symptoms. For example, cranial dystonia in a child would raise the index of suspicion for secondary dystonia. If tardive dystonia is part of the differential diagnosis, consider magnetic resonance imaging (MRI), serum ceruloplasmin measurement, or slit-lamp diagnostic testing. Suspicion of a structural lesion affecting the central nervous system warrants examination with MRI, computed tomography, or angiography. Certain metabolic and neurologic hereditary disorders cause secondary dystonia, in which case dopa-responsive dystonia should be ruled out. Psychometric testing should also be considered.

SYMPTOMATIC TREATMENT WITH CHEMODENERVATION

In the absence of a cure, treatment options for dystonia are necessarily symptomatic and supportive. Titratable chemo denervation agents are injected directly into the muscle or motor nerve, temporarily weakening the local muscle and easing dystonia symptoms. Chemo denervation agents include phenol, ethyl alcohol, and botulinum toxin types A (BTX-A; onabotulinumtoxinA, abo botulinumtoxinA, and incobotulinumtoxinA) and B (BTX-B; rimabotulinum toxinB).

Phenol and ethyl alcohol injections targeted perineurally or as a motor point block have been employed for dystonia and cause nonselective tissue destruction, muscle necrosis, and highly variable durations of response. Perineural microcirculation may be damaged, possibly leading to long-term defects.

Clostridium botulinum bacteria produce seven serologically distinct neuroparalytic toxins. They are the most powerful such toxins currently known and temporarily prevent acetylcholine vesicles from docking into the presynaptic neuromuscular junction. Use of BTX-A for treatment of dystonia was recommended in a National Institutes of Health consensus statement in 1990.⁹ It has been studied for a variety of dystonias, including blepharospasm, hemifacial spasm, laryngeal dystonia, oromandibular dystonia, and cervical dystonia, among other focal dystonias. Lew et al reported in 1997 on the successful use of BTX-B for cervical dystonia in a double-blind, single-treatment study,¹⁰ and confirmatory studies followed.^11,12

Varying indications for botulinum toxin

US Food and Drug Administration–approved indications for the toxins vary. The three BTX-A products and the single BTX-B product are approved for the treatment of cervical dystonia in adults to reduce the severity of abnormal head position and neck pain. OnabotulinumtoxinA is approved for treatment of blepharospasm and strabismus associated with dystonia; and incobotulinumtoxinA is approved for blepharospasm in patients who have previously been treated with onabotulinumtoxinA. BTX-A has also been found to be safe and effective for the management of focal dystonias. These botulinum toxin agents are not equivalent in dosing units, so caution must be observed when switching brands.

Patients selected to receive BTX for dystonia should meet three criteria:

• The dystonia should interfere with their functioning, comfort, or care to the degree that causes impairment and affects activities of daily living;
• Focal weakening following administration of the drug should not decrease their level of function; and
• The patient should understand that use of BTX may not completely address positioning, posturing, or secondary deformities.

Contraindications include pregnancy, lactation, comorbid neuromuscular disease (eg, amyotrophic lateral sclerosis or myasthenia gravis), and use of an aminoglycoside.

The need for BTX therapy should be reevaluated prior to each treatment; clinical benefit lasts 3 months or more. Electromyography may facilitate the location of target muscles, particularly since involved musculature may not be palpable and is often not superficial.¹³ In-office tools that help document baseline and posttreatment results, including videotaping dystonic limb movements and the use of rating scales, can be important for evaluating the patient’s progress.¹⁴

Relief for cervical dystonia

The treatment of choice for focal dystonias and focal aspects of generalized dystonia is BTX. Both BTX-A and BTX-B offer effective palliative treatments for cervical dystonia by improving neck position, reducing pain, and decreasing disability in sufferers.^11,15–18 The BTX solution is injected directly into the dystonic muscle at several locations, temporarily weakening the overactive muscle. The BTX dose is approximately proportional to the size of the muscle, although smaller muscles typically responsible for precision movement may require a relatively larger dose (Table 2). Doses may be modified according to clinical factors such as muscle bulk and severity of dystonia (Table 3).

Relief following BTX injection for cervical dystonia occurs about 1 week later, with the greatest effect seen at about 2 to 6 weeks following injection; relief may last 12 to 16 weeks. Reinjections are not normally administered prior to 12 weeks’ duration in order to reduce the possibility of antibody formation. Concomitant interventions addressing depression and anxiety may have a significant effect on overall quality of life.¹⁹ Patients may also try several sensory tricks, called gestes antagoniste, which may temporarily reduce or alleviate the dystonia. However, these tactile procedures—such as placing a hand on top of the head—lose their effectiveness over time.

Treatment of blepharospasm, focal limb dystonia

The use of BTX-A for blepharospasm is a significant improvement over the former clinical reliance on various oral medications, which, with the exception of baclofen, proved largely ineffective.²⁰ Surgical treatments result in damage to muscular and nervous tissues, and so are reserved only for nonresponders to BTX-A therapy.²¹

BTX-A can provide effective relief and is the treatment of choice for focal limb dystonias.²² Goals of treatment include functional improvement, correction of abnormal posture, and relief from discomfort. Although a variety of oral medications may also be prescribed, drug toxicity and adverse effects can outweigh the benefit and are usually only used in cases of severe dystonia. Oral medications used for limb dystonia include anticholinergics, dopamine agonists and antagonists, baclofen, clonazepam or other benzodiazepines, and muscle relaxants.

Antibodies may bind to the drug in a small percentage of patients who regularly receive injections of BTX, rendering additional injections of that specific serotype of BTX ineffective. This immunoresistance can be avoided if clinicians inject only the smallest quantity of BTX that achieves clinical efficacy, avoid administering booster injections before the end of the minimum 12-week lockout period, and extend the period between treatments as long as possible. If immunoresistance does occur, the BTX should be exchanged for a different serotype.

Testing for nonresponse

Patients are said to be nonresponders to BTX therapy if at 4 to 6 weeks following injection they show no reduction in muscle tone. A functional test for nonresponse is to inject a small amount of BTX into either the frontalis or sternocleidomastoid muscle prior to starting treatment; asymmetric weakness demonstrates a response, indicating that either injection technique or muscle selection is the problem. In addition to the development of neutralizing antibodies, other possible reasons for nonresponse include a dose that is too low or an alteration in the pattern of muscles involved in the dystonic movement.

Chorea is characterized by continuous, random, brief, involuntary muscle contractions that result from a variety of causes.¹ These involuntary movements are nonstereotyped and irregular. Although choreic disorders are among the most common involuntary movement disorders, their diagnosis and treatment present several important challenges, including identifying and removing the cause if possible, controlling and preventing motor symptoms, and managing neuropsychologic complications.¹ This article provides an overview of the diagnosis and treatment of choreic disorders, using Sydenham chorea to illustrate the management of autoimmune choreas and Huntington disease as the model for the management of heritable choreas.

Management of choreic disorders begins with a first-pass diagnosis and the use of symptomatic therapies. Even if this first pass yields no firm diagnosis, it at least rules out causes that have the most practical significance. A subsequent second-pass evaluation can be undertaken to look for rarer causes. Symptomatic therapies are continued throughout the diagnostic period. More specific therapies can be administered if an etiologic or pathogenic mechanism is determined (eg, postinfectious, autoimmune, metabolic).

CHOREIC DISORDERS: A PRACTICAL DIAGNOSTIC APPROACH

In general, choreic disorders may be subdivided into six categories:

1. Heredodegenerative disorders, such as Huntington disease and other genetically heterogeneous choreas, include Huntington disease–like 2 (HDL2) and benign hereditary chorea.¹ Sporadic cases include those of unknown paternity; X-linked disorders (eg, McLeod syndrome); and autosomal-recessive disorders such as chorea-acanthocytosis, which is characterized by chorea, dystonia with prominent orofacial involvement, self-mutilation, myopathy, and neuropathy.²

2. Drug-induced choreas include neurolepticinduced tardive dyskinesia and nontardive hyperkinetic drug-related choreas, the most common of which is levodopa-induced dyskinesias.³ Tardive drug-induced choreas may occur while using the culprit drug, while tapering the drug, or after it has been discontinued. The culprit drugs are represented by dopamine-receptor blockers and include the first-generation neuroleptics (eg, phenothiazines, haloperidol), antidepressants (lox-apine), and gastrointestinal agents (metoclopramide, prochlorperazine). Drug-induced choreas are possible with a wide range of pharmacologic agents, including antiparkinsonian drugs (eg, levodopa, dopamine agonists, anticholinergics), sympathomimetics (eg, amphetamines, cocaine), anticonvulsants, calcium channel blockers, and oral contraceptives.¹

3. Autoimmune choreas include Sydenham chorea, systemic lupus erythematosus, and antiphospholipid antibody syndromes. The latter encompass lupus anticoagulant and anticardiolipin antibodies.

4. Metabolic choreas are most often associated with hyperthyroidism, although case reports have described choreas in patients with vitamin B₁₂ deficiency. A variety of hereditary metabolic diseases are also included in this category.

5. Vascular choreas include polycythemia vera and cerebrovascular accidents, the latter frequently presenting as hemiballismus. Polycythemia vera is associated with a high incidence of neurologic symptoms, including a reported incidence of chorea of 0.5% to 5%,⁴ and should be considered as a potential cause of chorea.

6. Other choreic disorders include a variety of entities such as rare paraneoplastic disorder/syndrome, and posttraumatic and postanoxic presentations.

The first-pass diagnostic approach includes a family history, drug history, and brain magnetic resonance imaging to identify potential structural causes of chorea. Genetic testing for Huntington disease or other choreic disorders may also be performed, although it is essential to consider the potential implications of a positive test result. Intensive pretest and posttest counseling is important both for the patient and for currently asymptomatic family members who may also be affected.¹

Other testing includes:

• Complete blood count
• Creatine phosphokinase
• Peripheral smear for acanthocytes
• Comprehensive metabolic panel
• Ceruloplasmin level
• Measurement of thyroxine (T₄) and triiodothyronine (T₃)
• B₁₂ tests
• Antinuclear antibody sedimentation rate
• Lupus anticoagulant-anticardiolipin antibodies
• Antistreptolysin O (ASO) titer
• Anti-DNase-B titer.

GENERAL CONSIDERATIONS FOR THE TREATMENT OF CHOREIC DISORDERS

In some cases, choreic disorders have a treatable underlying etiology, such as thyroid disease or vitamin B₁₂ deficiency. Tardive syndromes may require treatment beyond drug discontinuation, including use of dopamine depleters for the classic tardive dyskinetic syndromes and anticholinergics for the tardive dystonic syndromes. Levodopa dyskinesia may be treated using amantadine, clozapine, or deep brain stimulation.^5,6 The treatment of patients with autoimmune choreas is not well defined. It may include anticoagulation in patients with positive anticardiolipin antibodies to prevent venous or arterial thromboembolism,⁷ but the risk of arterial thromboembolism is uncertain, and it is unclear whether chorea is truly a harbinger of vascular events.

A negative ASO titer does not exclude Sydenham chorea, a result of childhood infection with group-A beta-hemolytic streptococcus, and antibiotics should be considered in the appropriate context. Some researchers have argued that immune responses associated with acute infections may result in autoimmune neuropsychiatric symptoms. In pediatric patients, this has been referred to as pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS).⁸

A related phenomenon has been proposed as a potential mechanism of some types of chorea, although the relationship between acute infection and chorea is controversial. Patients with elevated ASO titer or anti-DNase-B titers may be candidates for antibiotics. By 6 weeks after the onset of infection, these titers will fall and a diagnosis of Sydenham chorea can be postulated or based exclusively on clinical judgment.

SYDENHAM CHOREA

Unique to Sydenham chorea is the use of penicillin as prophylaxis. Other than that, the management of Sydenham chorea exemplifies the management approach for the larger category of autoimmune choreic disorders. Pathogenic-based treatment options include immune modulation with cortico steroids, intravenous immunoglobulin (IVIG), and plasma exchange; all treatments must be administered in the appropriate clinical context.

One double-blind clinical trial examined the effectiveness of corticosteroid treatment in children with Sydenham chorea randomly assigned to receive either prednisone (n = 22) or placebo (n = 15).⁹ Prednisone was administered at a dose of 2 mg/kg/day for 4 weeks, followed by gradual tapering and discontinuation. The median time to remission of chorea was significantly lower for patients in the prednisone group (54.3 days) compared with those in the placebo group (119.9 days; P < .001). Patients in the prednisone group also exhibited significantly better scores on a chorea intensity rating scale at 8 weeks and 12 weeks (P < .001). Potential limitations of this approach include relapse of chorea symptoms and corticosteroid-related adverse events (eg, Cushing syndrome, hypertension).

A second study compared the effectiveness of three modalities: IVIG at a dose of 1 g/kg/day for 2 days (n = 4), plasma exchange (n = 8), and prednisone (n = 6).¹⁰ Although differences between treatment groups were not statistically significant, the authors noted that the clinical improvement in chorea symptoms tended to be greater for patients who received IVIG or plasma exchange than for those who received prednisone. Mean chorea scores improved from baseline by 72% for the IVIG group, 50% for the plasma exchange group, and 29% for the prednisone group.

After etiology-dependent treatments have been considered, several other options may be effective regardless of the specific etiology. These include symptomatic treatments such as haloperidol, atypical neuroleptics, and amantadine.¹¹ Antiepileptic medications or benzodiazepines may also help to control symptoms, although less information is available about the use of these agents for the treatment of Sydenham chorea. Tetrabenazine may be considered for patients who will require long-term treatment.

HUNTINGTON DISEASE

Pharmacotherapy of Huntington disease may be unnecessary if symptoms are mild or not bothersome. Symptomatic treatment options include tetrabenazine, amantadine, and either first-generation neuroleptics (eg, haloperidol) or second-generation atypical neuroleptics (eg, olanzapine, quetiapine, risperidone, ziprasidone).

Treating choreas with tetrabenazine or amantadine

Considerable recent attention has focused on the efficacy and safety of tetrabenazine for the treatment of Huntington disease and other choreic disorders. Tetrabenazine is a central monoamine depleter that reversibly binds to the type-2 vesicular monoamine transporter.¹² The TETRA-HD study examined the efficacy and safety of tetrabenazine for the short- and long-term control of Huntington disease.¹² An initial study compared tetrabenazine with placebo in 75 patients who were treated for up to 13 weeks. In an extension study, all patients received individualized tetrabenazine doses for up to 80 weeks.

The mean total maximal chorea (TMC) scores from the Unified Huntington Disease Rating Scale (UHDRS) decreased markedly during the first 10 weeks of tetrabenazine treatment, remained lower than baseline throughout 80 weeks, and then returned to baseline levels after tetrabenazine discontinuation (Figure 1). At week 80, the mean TMC score was reduced by 4.6 UHDRS units compared with baseline (P < .001) The long-term extension phase was completed by 45 of 75 patients. Treatment-related adverse events that prompted discontinuation included depression, delusions, and vocal tics. The most commonly reported adverse events included sedation or somnolence (n = 18), depressed mood (n = 17), anxiety (n = 13), insomnia (n = 10), and akathisia (n = 9). Scores of parkinsonism and dysphagia increased significantly from baseline over the 80-week study.

Amantadine is an option for patients who cannot tolerate tetrabenazine. A double-blind, placebo-controlled study performed by researchers at the National Institutes of Health (NIH) examined the efficacy and safety of amantadine in 24 patients with Huntington disease.¹³ Patients were treated with oral amantadine 400 mg/day or placebo for 2 weeks, and were then crossed over to the other treatment. Amantadine was associated with a median reduction in extremity chorea score at rest of 36% from baseline (P = .04), versus 0% improvement with placebo. The mean improvement with amantadine was 56% for the 10 patients with the highest drug plasma levels.

Improvement in chorea scores from baseline for amantadine compared with placebo was rated with four different methods: (1) maximal chorea severity measured from video recordings; (2) maximal chorea severity measured by live raters; (3), chorea severity at rest measured from video recordings; and (4) extremity chorea at rest measured from video recordings. Amantadine was superior to placebo according to all four rating methods. Treatment was generally safe and well tolerated, and no consistent changes in cognitive function were noted with amantadine therapy.

Reprinted with permission from Neurology (Lucetti C, et al. IV amantadine improves chorea in Huntington’s disease: an acute randomized, controlled study. Neurology 2003; 60:1995–1997). Copyright © 2003 by AAN Enterprises, Inc.

In addition to addressing chorea, it is also important to manage nonmotor complications of Huntington disease, including cognition, mood, and thought disorders. Rivastigmine was assessed for the treatment of motor symptoms, functional disability, and cognitive impairment associated with Huntington disease in an open-label study of 18 patients; 11 received rivastigmine 6 mg/day and 7 control patients did not.¹⁵ Motor and cognitive function were assessed for up to 2 years by raters who were blinded to treatment assignment. Ratings on a global motor performance scale were significantly better for patients who received rivastigmine than for control subjects. Rivastigmine treatment was also associated with trends toward improvements in functional disability and cognitive impairment, although these differences were not statistically significant.

A small open-label study examined the effects of donepezil for movement and cognitive symptoms associated with Huntington disease.¹⁶ Donepezil did not significantly improve cognitive symptoms, although the study enrolled only eight patients. All patients tolerated oral donepezil at a dose of 5 mg/day, but four patients withdrew from the study when the dose was increased to 10 mg/day. In two patients, chorea worsened and falls increased, moderate to severe diarrhea developed in three patients, and one patient reported anxiety and irritability.

Depression is another common complication of Huntington disease. The incidence of depression among patients with Huntington disease is approximately 40%, and the risk of suicide is at least eightfold greater than that among the general population.¹⁷ Treatment must be guided by clinical judgment. Selective serotonin reuptake inhibitor antidepressants have been recommended. Other options to manage depression include mirtazapine, monoamine oxidase inhibitors, or electroshock therapy. Mood-stabilizing agents (eg, carbamazepine, lamotrigine, valproate) may also be indicated in helping with impulse control. Haloperidol and second-generation antipsychotics are used for the treatment of a broad range of psychiatric conditions, many of which may overlap with Huntington disease, including schizophrenia and schizophreniform disorder, schizoaffective disorder, bipolar disorder, dementia, and disruptive behavior.¹⁸ The risk of tardive dyskinesia may be as much as fivefold lower with second-generation antipsychotics.¹⁸ Many patients with Huntington disease require treatment for aggression. A variety of approaches are available, including behavior modification, the antidepressant sertraline, buspirone, antipsychotic agents (eg, risperidone, olanzapine), propranolol, and lithium (combined with haloperidol).

Long-term care considerations

As a consequence of the diverse clinical manifestations of choreic disorders in movement, function, mood, and cognition, the treatment of Huntington disease requires a multidisciplinary approach that involves a number of different health care specialties across the long-term course of the disorder. Members of the Huntington disease treatment team may include neurologists, psychiatrists, nurses, social workers, geneticists, physical therapists, occupational therapists, speech therapists, dietitians, and other supporting groups or professional societies. The clinical manifestations of Huntington disease may evolve over time, as symptoms such as bradykinesia, dystonia, rigidity, cognitive decline, and gait instability become more significant.¹⁹ As a result, optimal management strategies for patients with Huntington disease may change significantly across the long-term course of the disease. During the early course of the disease, the typical clinical presentation is largely hyperkinesis, irritability, and distractibility. These patients will require initiation of drug therapy and linkage to sources of support. In the later stages of the disease, the presentation shifts to a more hypokinetic and apathetic profile, and patients are more likely to require drug regimen review and modification, nursing home placement, and palliative care services.¹⁹

Another important concern in Huntington disease treatment is care of the caregiver. Surveys show that the key concerns of caregivers include the expertise of the health care professionals who are treating the patient and the availability of sufficient services in the community.²⁰ Several resources are available for Huntington disease caregivers, including local support groups, the Huntington’s Disease Society of America, Q Foundation, and the Huntington Study Group. The Lundbeck pharmaceutical company operates a patient assistance program (LundbeckShare. com) as well as an information center that can be accessed toll free at (888)457-4273. Approximately 90% of patients who request copayment assistance qualify for aid, regardless of the type of insurance they carry.

SUMMARY AND CONCLUSIONS

The approach to a patient with chorea starts with a search for specifically treatable etiologies. Autoimmune, metabolic, and vascular causes should be sought first and treated. The symptomatic treatment of all choreas is based on the model described here for Huntington disease, and includes attention to cognitive, psychiatric, and social support issues. The recommended approach is multidisciplinary, with a change in the mix of services as the disease progresses. It is also important to recognize the burden of Huntington disease on the caregiver and consider steps to make this burden more manageable.

Chorea is characterized by continuous, random, brief, involuntary muscle contractions that result from a variety of causes.¹ These involuntary movements are nonstereotyped and irregular. Although choreic disorders are among the most common involuntary movement disorders, their diagnosis and treatment present several important challenges, including identifying and removing the cause if possible, controlling and preventing motor symptoms, and managing neuropsychologic complications.¹ This article provides an overview of the diagnosis and treatment of choreic disorders, using Sydenham chorea to illustrate the management of autoimmune choreas and Huntington disease as the model for the management of heritable choreas.

Management of choreic disorders begins with a first-pass diagnosis and the use of symptomatic therapies. Even if this first pass yields no firm diagnosis, it at least rules out causes that have the most practical significance. A subsequent second-pass evaluation can be undertaken to look for rarer causes. Symptomatic therapies are continued throughout the diagnostic period. More specific therapies can be administered if an etiologic or pathogenic mechanism is determined (eg, postinfectious, autoimmune, metabolic).

CHOREIC DISORDERS: A PRACTICAL DIAGNOSTIC APPROACH

In general, choreic disorders may be subdivided into six categories:

1. Heredodegenerative disorders, such as Huntington disease and other genetically heterogeneous choreas, include Huntington disease–like 2 (HDL2) and benign hereditary chorea.¹ Sporadic cases include those of unknown paternity; X-linked disorders (eg, McLeod syndrome); and autosomal-recessive disorders such as chorea-acanthocytosis, which is characterized by chorea, dystonia with prominent orofacial involvement, self-mutilation, myopathy, and neuropathy.²

2. Drug-induced choreas include neurolepticinduced tardive dyskinesia and nontardive hyperkinetic drug-related choreas, the most common of which is levodopa-induced dyskinesias.³ Tardive drug-induced choreas may occur while using the culprit drug, while tapering the drug, or after it has been discontinued. The culprit drugs are represented by dopamine-receptor blockers and include the first-generation neuroleptics (eg, phenothiazines, haloperidol), antidepressants (lox-apine), and gastrointestinal agents (metoclopramide, prochlorperazine). Drug-induced choreas are possible with a wide range of pharmacologic agents, including antiparkinsonian drugs (eg, levodopa, dopamine agonists, anticholinergics), sympathomimetics (eg, amphetamines, cocaine), anticonvulsants, calcium channel blockers, and oral contraceptives.¹

3. Autoimmune choreas include Sydenham chorea, systemic lupus erythematosus, and antiphospholipid antibody syndromes. The latter encompass lupus anticoagulant and anticardiolipin antibodies.

4. Metabolic choreas are most often associated with hyperthyroidism, although case reports have described choreas in patients with vitamin B₁₂ deficiency. A variety of hereditary metabolic diseases are also included in this category.

5. Vascular choreas include polycythemia vera and cerebrovascular accidents, the latter frequently presenting as hemiballismus. Polycythemia vera is associated with a high incidence of neurologic symptoms, including a reported incidence of chorea of 0.5% to 5%,⁴ and should be considered as a potential cause of chorea.

6. Other choreic disorders include a variety of entities such as rare paraneoplastic disorder/syndrome, and posttraumatic and postanoxic presentations.

The first-pass diagnostic approach includes a family history, drug history, and brain magnetic resonance imaging to identify potential structural causes of chorea. Genetic testing for Huntington disease or other choreic disorders may also be performed, although it is essential to consider the potential implications of a positive test result. Intensive pretest and posttest counseling is important both for the patient and for currently asymptomatic family members who may also be affected.¹

Other testing includes:

• Complete blood count
• Creatine phosphokinase
• Peripheral smear for acanthocytes
• Comprehensive metabolic panel
• Ceruloplasmin level
• Measurement of thyroxine (T₄) and triiodothyronine (T₃)
• B₁₂ tests
• Antinuclear antibody sedimentation rate
• Lupus anticoagulant-anticardiolipin antibodies
• Antistreptolysin O (ASO) titer
• Anti-DNase-B titer.

GENERAL CONSIDERATIONS FOR THE TREATMENT OF CHOREIC DISORDERS

In some cases, choreic disorders have a treatable underlying etiology, such as thyroid disease or vitamin B₁₂ deficiency. Tardive syndromes may require treatment beyond drug discontinuation, including use of dopamine depleters for the classic tardive dyskinetic syndromes and anticholinergics for the tardive dystonic syndromes. Levodopa dyskinesia may be treated using amantadine, clozapine, or deep brain stimulation.^5,6 The treatment of patients with autoimmune choreas is not well defined. It may include anticoagulation in patients with positive anticardiolipin antibodies to prevent venous or arterial thromboembolism,⁷ but the risk of arterial thromboembolism is uncertain, and it is unclear whether chorea is truly a harbinger of vascular events.

A negative ASO titer does not exclude Sydenham chorea, a result of childhood infection with group-A beta-hemolytic streptococcus, and antibiotics should be considered in the appropriate context. Some researchers have argued that immune responses associated with acute infections may result in autoimmune neuropsychiatric symptoms. In pediatric patients, this has been referred to as pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS).⁸

A related phenomenon has been proposed as a potential mechanism of some types of chorea, although the relationship between acute infection and chorea is controversial. Patients with elevated ASO titer or anti-DNase-B titers may be candidates for antibiotics. By 6 weeks after the onset of infection, these titers will fall and a diagnosis of Sydenham chorea can be postulated or based exclusively on clinical judgment.

SYDENHAM CHOREA

Unique to Sydenham chorea is the use of penicillin as prophylaxis. Other than that, the management of Sydenham chorea exemplifies the management approach for the larger category of autoimmune choreic disorders. Pathogenic-based treatment options include immune modulation with cortico steroids, intravenous immunoglobulin (IVIG), and plasma exchange; all treatments must be administered in the appropriate clinical context.

One double-blind clinical trial examined the effectiveness of corticosteroid treatment in children with Sydenham chorea randomly assigned to receive either prednisone (n = 22) or placebo (n = 15).⁹ Prednisone was administered at a dose of 2 mg/kg/day for 4 weeks, followed by gradual tapering and discontinuation. The median time to remission of chorea was significantly lower for patients in the prednisone group (54.3 days) compared with those in the placebo group (119.9 days; P < .001). Patients in the prednisone group also exhibited significantly better scores on a chorea intensity rating scale at 8 weeks and 12 weeks (P < .001). Potential limitations of this approach include relapse of chorea symptoms and corticosteroid-related adverse events (eg, Cushing syndrome, hypertension).

A second study compared the effectiveness of three modalities: IVIG at a dose of 1 g/kg/day for 2 days (n = 4), plasma exchange (n = 8), and prednisone (n = 6).¹⁰ Although differences between treatment groups were not statistically significant, the authors noted that the clinical improvement in chorea symptoms tended to be greater for patients who received IVIG or plasma exchange than for those who received prednisone. Mean chorea scores improved from baseline by 72% for the IVIG group, 50% for the plasma exchange group, and 29% for the prednisone group.

After etiology-dependent treatments have been considered, several other options may be effective regardless of the specific etiology. These include symptomatic treatments such as haloperidol, atypical neuroleptics, and amantadine.¹¹ Antiepileptic medications or benzodiazepines may also help to control symptoms, although less information is available about the use of these agents for the treatment of Sydenham chorea. Tetrabenazine may be considered for patients who will require long-term treatment.

HUNTINGTON DISEASE

Pharmacotherapy of Huntington disease may be unnecessary if symptoms are mild or not bothersome. Symptomatic treatment options include tetrabenazine, amantadine, and either first-generation neuroleptics (eg, haloperidol) or second-generation atypical neuroleptics (eg, olanzapine, quetiapine, risperidone, ziprasidone).

Treating choreas with tetrabenazine or amantadine

Considerable recent attention has focused on the efficacy and safety of tetrabenazine for the treatment of Huntington disease and other choreic disorders. Tetrabenazine is a central monoamine depleter that reversibly binds to the type-2 vesicular monoamine transporter.¹² The TETRA-HD study examined the efficacy and safety of tetrabenazine for the short- and long-term control of Huntington disease.¹² An initial study compared tetrabenazine with placebo in 75 patients who were treated for up to 13 weeks. In an extension study, all patients received individualized tetrabenazine doses for up to 80 weeks.

The mean total maximal chorea (TMC) scores from the Unified Huntington Disease Rating Scale (UHDRS) decreased markedly during the first 10 weeks of tetrabenazine treatment, remained lower than baseline throughout 80 weeks, and then returned to baseline levels after tetrabenazine discontinuation (Figure 1). At week 80, the mean TMC score was reduced by 4.6 UHDRS units compared with baseline (P < .001) The long-term extension phase was completed by 45 of 75 patients. Treatment-related adverse events that prompted discontinuation included depression, delusions, and vocal tics. The most commonly reported adverse events included sedation or somnolence (n = 18), depressed mood (n = 17), anxiety (n = 13), insomnia (n = 10), and akathisia (n = 9). Scores of parkinsonism and dysphagia increased significantly from baseline over the 80-week study.

Amantadine is an option for patients who cannot tolerate tetrabenazine. A double-blind, placebo-controlled study performed by researchers at the National Institutes of Health (NIH) examined the efficacy and safety of amantadine in 24 patients with Huntington disease.¹³ Patients were treated with oral amantadine 400 mg/day or placebo for 2 weeks, and were then crossed over to the other treatment. Amantadine was associated with a median reduction in extremity chorea score at rest of 36% from baseline (P = .04), versus 0% improvement with placebo. The mean improvement with amantadine was 56% for the 10 patients with the highest drug plasma levels.

Improvement in chorea scores from baseline for amantadine compared with placebo was rated with four different methods: (1) maximal chorea severity measured from video recordings; (2) maximal chorea severity measured by live raters; (3), chorea severity at rest measured from video recordings; and (4) extremity chorea at rest measured from video recordings. Amantadine was superior to placebo according to all four rating methods. Treatment was generally safe and well tolerated, and no consistent changes in cognitive function were noted with amantadine therapy.

Reprinted with permission from Neurology (Lucetti C, et al. IV amantadine improves chorea in Huntington’s disease: an acute randomized, controlled study. Neurology 2003; 60:1995–1997). Copyright © 2003 by AAN Enterprises, Inc.

In addition to addressing chorea, it is also important to manage nonmotor complications of Huntington disease, including cognition, mood, and thought disorders. Rivastigmine was assessed for the treatment of motor symptoms, functional disability, and cognitive impairment associated with Huntington disease in an open-label study of 18 patients; 11 received rivastigmine 6 mg/day and 7 control patients did not.¹⁵ Motor and cognitive function were assessed for up to 2 years by raters who were blinded to treatment assignment. Ratings on a global motor performance scale were significantly better for patients who received rivastigmine than for control subjects. Rivastigmine treatment was also associated with trends toward improvements in functional disability and cognitive impairment, although these differences were not statistically significant.

A small open-label study examined the effects of donepezil for movement and cognitive symptoms associated with Huntington disease.¹⁶ Donepezil did not significantly improve cognitive symptoms, although the study enrolled only eight patients. All patients tolerated oral donepezil at a dose of 5 mg/day, but four patients withdrew from the study when the dose was increased to 10 mg/day. In two patients, chorea worsened and falls increased, moderate to severe diarrhea developed in three patients, and one patient reported anxiety and irritability.

Depression is another common complication of Huntington disease. The incidence of depression among patients with Huntington disease is approximately 40%, and the risk of suicide is at least eightfold greater than that among the general population.¹⁷ Treatment must be guided by clinical judgment. Selective serotonin reuptake inhibitor antidepressants have been recommended. Other options to manage depression include mirtazapine, monoamine oxidase inhibitors, or electroshock therapy. Mood-stabilizing agents (eg, carbamazepine, lamotrigine, valproate) may also be indicated in helping with impulse control. Haloperidol and second-generation antipsychotics are used for the treatment of a broad range of psychiatric conditions, many of which may overlap with Huntington disease, including schizophrenia and schizophreniform disorder, schizoaffective disorder, bipolar disorder, dementia, and disruptive behavior.¹⁸ The risk of tardive dyskinesia may be as much as fivefold lower with second-generation antipsychotics.¹⁸ Many patients with Huntington disease require treatment for aggression. A variety of approaches are available, including behavior modification, the antidepressant sertraline, buspirone, antipsychotic agents (eg, risperidone, olanzapine), propranolol, and lithium (combined with haloperidol).

Long-term care considerations

As a consequence of the diverse clinical manifestations of choreic disorders in movement, function, mood, and cognition, the treatment of Huntington disease requires a multidisciplinary approach that involves a number of different health care specialties across the long-term course of the disorder. Members of the Huntington disease treatment team may include neurologists, psychiatrists, nurses, social workers, geneticists, physical therapists, occupational therapists, speech therapists, dietitians, and other supporting groups or professional societies. The clinical manifestations of Huntington disease may evolve over time, as symptoms such as bradykinesia, dystonia, rigidity, cognitive decline, and gait instability become more significant.¹⁹ As a result, optimal management strategies for patients with Huntington disease may change significantly across the long-term course of the disease. During the early course of the disease, the typical clinical presentation is largely hyperkinesis, irritability, and distractibility. These patients will require initiation of drug therapy and linkage to sources of support. In the later stages of the disease, the presentation shifts to a more hypokinetic and apathetic profile, and patients are more likely to require drug regimen review and modification, nursing home placement, and palliative care services.¹⁹

Another important concern in Huntington disease treatment is care of the caregiver. Surveys show that the key concerns of caregivers include the expertise of the health care professionals who are treating the patient and the availability of sufficient services in the community.²⁰ Several resources are available for Huntington disease caregivers, including local support groups, the Huntington’s Disease Society of America, Q Foundation, and the Huntington Study Group. The Lundbeck pharmaceutical company operates a patient assistance program (LundbeckShare. com) as well as an information center that can be accessed toll free at (888)457-4273. Approximately 90% of patients who request copayment assistance qualify for aid, regardless of the type of insurance they carry.

SUMMARY AND CONCLUSIONS

The approach to a patient with chorea starts with a search for specifically treatable etiologies. Autoimmune, metabolic, and vascular causes should be sought first and treated. The symptomatic treatment of all choreas is based on the model described here for Huntington disease, and includes attention to cognitive, psychiatric, and social support issues. The recommended approach is multidisciplinary, with a change in the mix of services as the disease progresses. It is also important to recognize the burden of Huntington disease on the caregiver and consider steps to make this burden more manageable.

Chorea is characterized by continuous, random, brief, involuntary muscle contractions that result from a variety of causes.¹ These involuntary movements are nonstereotyped and irregular. Although choreic disorders are among the most common involuntary movement disorders, their diagnosis and treatment present several important challenges, including identifying and removing the cause if possible, controlling and preventing motor symptoms, and managing neuropsychologic complications.¹ This article provides an overview of the diagnosis and treatment of choreic disorders, using Sydenham chorea to illustrate the management of autoimmune choreas and Huntington disease as the model for the management of heritable choreas.

Management of choreic disorders begins with a first-pass diagnosis and the use of symptomatic therapies. Even if this first pass yields no firm diagnosis, it at least rules out causes that have the most practical significance. A subsequent second-pass evaluation can be undertaken to look for rarer causes. Symptomatic therapies are continued throughout the diagnostic period. More specific therapies can be administered if an etiologic or pathogenic mechanism is determined (eg, postinfectious, autoimmune, metabolic).

CHOREIC DISORDERS: A PRACTICAL DIAGNOSTIC APPROACH

In general, choreic disorders may be subdivided into six categories:

1. Heredodegenerative disorders, such as Huntington disease and other genetically heterogeneous choreas, include Huntington disease–like 2 (HDL2) and benign hereditary chorea.¹ Sporadic cases include those of unknown paternity; X-linked disorders (eg, McLeod syndrome); and autosomal-recessive disorders such as chorea-acanthocytosis, which is characterized by chorea, dystonia with prominent orofacial involvement, self-mutilation, myopathy, and neuropathy.²

2. Drug-induced choreas include neurolepticinduced tardive dyskinesia and nontardive hyperkinetic drug-related choreas, the most common of which is levodopa-induced dyskinesias.³ Tardive drug-induced choreas may occur while using the culprit drug, while tapering the drug, or after it has been discontinued. The culprit drugs are represented by dopamine-receptor blockers and include the first-generation neuroleptics (eg, phenothiazines, haloperidol), antidepressants (lox-apine), and gastrointestinal agents (metoclopramide, prochlorperazine). Drug-induced choreas are possible with a wide range of pharmacologic agents, including antiparkinsonian drugs (eg, levodopa, dopamine agonists, anticholinergics), sympathomimetics (eg, amphetamines, cocaine), anticonvulsants, calcium channel blockers, and oral contraceptives.¹

3. Autoimmune choreas include Sydenham chorea, systemic lupus erythematosus, and antiphospholipid antibody syndromes. The latter encompass lupus anticoagulant and anticardiolipin antibodies.

4. Metabolic choreas are most often associated with hyperthyroidism, although case reports have described choreas in patients with vitamin B₁₂ deficiency. A variety of hereditary metabolic diseases are also included in this category.

5. Vascular choreas include polycythemia vera and cerebrovascular accidents, the latter frequently presenting as hemiballismus. Polycythemia vera is associated with a high incidence of neurologic symptoms, including a reported incidence of chorea of 0.5% to 5%,⁴ and should be considered as a potential cause of chorea.

6. Other choreic disorders include a variety of entities such as rare paraneoplastic disorder/syndrome, and posttraumatic and postanoxic presentations.

The first-pass diagnostic approach includes a family history, drug history, and brain magnetic resonance imaging to identify potential structural causes of chorea. Genetic testing for Huntington disease or other choreic disorders may also be performed, although it is essential to consider the potential implications of a positive test result. Intensive pretest and posttest counseling is important both for the patient and for currently asymptomatic family members who may also be affected.¹

Other testing includes:

• Complete blood count
• Creatine phosphokinase
• Peripheral smear for acanthocytes
• Comprehensive metabolic panel
• Ceruloplasmin level
• Measurement of thyroxine (T₄) and triiodothyronine (T₃)
• B₁₂ tests
• Antinuclear antibody sedimentation rate
• Lupus anticoagulant-anticardiolipin antibodies
• Antistreptolysin O (ASO) titer
• Anti-DNase-B titer.

GENERAL CONSIDERATIONS FOR THE TREATMENT OF CHOREIC DISORDERS

In some cases, choreic disorders have a treatable underlying etiology, such as thyroid disease or vitamin B₁₂ deficiency. Tardive syndromes may require treatment beyond drug discontinuation, including use of dopamine depleters for the classic tardive dyskinetic syndromes and anticholinergics for the tardive dystonic syndromes. Levodopa dyskinesia may be treated using amantadine, clozapine, or deep brain stimulation.^5,6 The treatment of patients with autoimmune choreas is not well defined. It may include anticoagulation in patients with positive anticardiolipin antibodies to prevent venous or arterial thromboembolism,⁷ but the risk of arterial thromboembolism is uncertain, and it is unclear whether chorea is truly a harbinger of vascular events.

A negative ASO titer does not exclude Sydenham chorea, a result of childhood infection with group-A beta-hemolytic streptococcus, and antibiotics should be considered in the appropriate context. Some researchers have argued that immune responses associated with acute infections may result in autoimmune neuropsychiatric symptoms. In pediatric patients, this has been referred to as pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS).⁸

A related phenomenon has been proposed as a potential mechanism of some types of chorea, although the relationship between acute infection and chorea is controversial. Patients with elevated ASO titer or anti-DNase-B titers may be candidates for antibiotics. By 6 weeks after the onset of infection, these titers will fall and a diagnosis of Sydenham chorea can be postulated or based exclusively on clinical judgment.

SYDENHAM CHOREA

Unique to Sydenham chorea is the use of penicillin as prophylaxis. Other than that, the management of Sydenham chorea exemplifies the management approach for the larger category of autoimmune choreic disorders. Pathogenic-based treatment options include immune modulation with cortico steroids, intravenous immunoglobulin (IVIG), and plasma exchange; all treatments must be administered in the appropriate clinical context.

One double-blind clinical trial examined the effectiveness of corticosteroid treatment in children with Sydenham chorea randomly assigned to receive either prednisone (n = 22) or placebo (n = 15).⁹ Prednisone was administered at a dose of 2 mg/kg/day for 4 weeks, followed by gradual tapering and discontinuation. The median time to remission of chorea was significantly lower for patients in the prednisone group (54.3 days) compared with those in the placebo group (119.9 days; P < .001). Patients in the prednisone group also exhibited significantly better scores on a chorea intensity rating scale at 8 weeks and 12 weeks (P < .001). Potential limitations of this approach include relapse of chorea symptoms and corticosteroid-related adverse events (eg, Cushing syndrome, hypertension).

A second study compared the effectiveness of three modalities: IVIG at a dose of 1 g/kg/day for 2 days (n = 4), plasma exchange (n = 8), and prednisone (n = 6).¹⁰ Although differences between treatment groups were not statistically significant, the authors noted that the clinical improvement in chorea symptoms tended to be greater for patients who received IVIG or plasma exchange than for those who received prednisone. Mean chorea scores improved from baseline by 72% for the IVIG group, 50% for the plasma exchange group, and 29% for the prednisone group.

After etiology-dependent treatments have been considered, several other options may be effective regardless of the specific etiology. These include symptomatic treatments such as haloperidol, atypical neuroleptics, and amantadine.¹¹ Antiepileptic medications or benzodiazepines may also help to control symptoms, although less information is available about the use of these agents for the treatment of Sydenham chorea. Tetrabenazine may be considered for patients who will require long-term treatment.

HUNTINGTON DISEASE

Pharmacotherapy of Huntington disease may be unnecessary if symptoms are mild or not bothersome. Symptomatic treatment options include tetrabenazine, amantadine, and either first-generation neuroleptics (eg, haloperidol) or second-generation atypical neuroleptics (eg, olanzapine, quetiapine, risperidone, ziprasidone).

Treating choreas with tetrabenazine or amantadine

Considerable recent attention has focused on the efficacy and safety of tetrabenazine for the treatment of Huntington disease and other choreic disorders. Tetrabenazine is a central monoamine depleter that reversibly binds to the type-2 vesicular monoamine transporter.¹² The TETRA-HD study examined the efficacy and safety of tetrabenazine for the short- and long-term control of Huntington disease.¹² An initial study compared tetrabenazine with placebo in 75 patients who were treated for up to 13 weeks. In an extension study, all patients received individualized tetrabenazine doses for up to 80 weeks.

The mean total maximal chorea (TMC) scores from the Unified Huntington Disease Rating Scale (UHDRS) decreased markedly during the first 10 weeks of tetrabenazine treatment, remained lower than baseline throughout 80 weeks, and then returned to baseline levels after tetrabenazine discontinuation (Figure 1). At week 80, the mean TMC score was reduced by 4.6 UHDRS units compared with baseline (P < .001) The long-term extension phase was completed by 45 of 75 patients. Treatment-related adverse events that prompted discontinuation included depression, delusions, and vocal tics. The most commonly reported adverse events included sedation or somnolence (n = 18), depressed mood (n = 17), anxiety (n = 13), insomnia (n = 10), and akathisia (n = 9). Scores of parkinsonism and dysphagia increased significantly from baseline over the 80-week study.

Amantadine is an option for patients who cannot tolerate tetrabenazine. A double-blind, placebo-controlled study performed by researchers at the National Institutes of Health (NIH) examined the efficacy and safety of amantadine in 24 patients with Huntington disease.¹³ Patients were treated with oral amantadine 400 mg/day or placebo for 2 weeks, and were then crossed over to the other treatment. Amantadine was associated with a median reduction in extremity chorea score at rest of 36% from baseline (P = .04), versus 0% improvement with placebo. The mean improvement with amantadine was 56% for the 10 patients with the highest drug plasma levels.

Improvement in chorea scores from baseline for amantadine compared with placebo was rated with four different methods: (1) maximal chorea severity measured from video recordings; (2) maximal chorea severity measured by live raters; (3), chorea severity at rest measured from video recordings; and (4) extremity chorea at rest measured from video recordings. Amantadine was superior to placebo according to all four rating methods. Treatment was generally safe and well tolerated, and no consistent changes in cognitive function were noted with amantadine therapy.

Reprinted with permission from Neurology (Lucetti C, et al. IV amantadine improves chorea in Huntington’s disease: an acute randomized, controlled study. Neurology 2003; 60:1995–1997). Copyright © 2003 by AAN Enterprises, Inc.

In addition to addressing chorea, it is also important to manage nonmotor complications of Huntington disease, including cognition, mood, and thought disorders. Rivastigmine was assessed for the treatment of motor symptoms, functional disability, and cognitive impairment associated with Huntington disease in an open-label study of 18 patients; 11 received rivastigmine 6 mg/day and 7 control patients did not.¹⁵ Motor and cognitive function were assessed for up to 2 years by raters who were blinded to treatment assignment. Ratings on a global motor performance scale were significantly better for patients who received rivastigmine than for control subjects. Rivastigmine treatment was also associated with trends toward improvements in functional disability and cognitive impairment, although these differences were not statistically significant.

A small open-label study examined the effects of donepezil for movement and cognitive symptoms associated with Huntington disease.¹⁶ Donepezil did not significantly improve cognitive symptoms, although the study enrolled only eight patients. All patients tolerated oral donepezil at a dose of 5 mg/day, but four patients withdrew from the study when the dose was increased to 10 mg/day. In two patients, chorea worsened and falls increased, moderate to severe diarrhea developed in three patients, and one patient reported anxiety and irritability.

Depression is another common complication of Huntington disease. The incidence of depression among patients with Huntington disease is approximately 40%, and the risk of suicide is at least eightfold greater than that among the general population.¹⁷ Treatment must be guided by clinical judgment. Selective serotonin reuptake inhibitor antidepressants have been recommended. Other options to manage depression include mirtazapine, monoamine oxidase inhibitors, or electroshock therapy. Mood-stabilizing agents (eg, carbamazepine, lamotrigine, valproate) may also be indicated in helping with impulse control. Haloperidol and second-generation antipsychotics are used for the treatment of a broad range of psychiatric conditions, many of which may overlap with Huntington disease, including schizophrenia and schizophreniform disorder, schizoaffective disorder, bipolar disorder, dementia, and disruptive behavior.¹⁸ The risk of tardive dyskinesia may be as much as fivefold lower with second-generation antipsychotics.¹⁸ Many patients with Huntington disease require treatment for aggression. A variety of approaches are available, including behavior modification, the antidepressant sertraline, buspirone, antipsychotic agents (eg, risperidone, olanzapine), propranolol, and lithium (combined with haloperidol).

Long-term care considerations

As a consequence of the diverse clinical manifestations of choreic disorders in movement, function, mood, and cognition, the treatment of Huntington disease requires a multidisciplinary approach that involves a number of different health care specialties across the long-term course of the disorder. Members of the Huntington disease treatment team may include neurologists, psychiatrists, nurses, social workers, geneticists, physical therapists, occupational therapists, speech therapists, dietitians, and other supporting groups or professional societies. The clinical manifestations of Huntington disease may evolve over time, as symptoms such as bradykinesia, dystonia, rigidity, cognitive decline, and gait instability become more significant.¹⁹ As a result, optimal management strategies for patients with Huntington disease may change significantly across the long-term course of the disease. During the early course of the disease, the typical clinical presentation is largely hyperkinesis, irritability, and distractibility. These patients will require initiation of drug therapy and linkage to sources of support. In the later stages of the disease, the presentation shifts to a more hypokinetic and apathetic profile, and patients are more likely to require drug regimen review and modification, nursing home placement, and palliative care services.¹⁹

Another important concern in Huntington disease treatment is care of the caregiver. Surveys show that the key concerns of caregivers include the expertise of the health care professionals who are treating the patient and the availability of sufficient services in the community.²⁰ Several resources are available for Huntington disease caregivers, including local support groups, the Huntington’s Disease Society of America, Q Foundation, and the Huntington Study Group. The Lundbeck pharmaceutical company operates a patient assistance program (LundbeckShare. com) as well as an information center that can be accessed toll free at (888)457-4273. Approximately 90% of patients who request copayment assistance qualify for aid, regardless of the type of insurance they carry.

SUMMARY AND CONCLUSIONS

The approach to a patient with chorea starts with a search for specifically treatable etiologies. Autoimmune, metabolic, and vascular causes should be sought first and treated. The symptomatic treatment of all choreas is based on the model described here for Huntington disease, and includes attention to cognitive, psychiatric, and social support issues. The recommended approach is multidisciplinary, with a change in the mix of services as the disease progresses. It is also important to recognize the burden of Huntington disease on the caregiver and consider steps to make this burden more manageable.

Tourette syndrome (TS) is part of a spectrum of tic disorders. Tics are sudden, rapid, stereotyped, repetitive, nonrhythmic movements or vocalizations affecting discrete muscle groups, and are preceded by a sensory component. Patients in whom tic suppression is attempted report the experience of a sensation of inner pressure that must be released. This eventually results in the performance of motor movement or vocal sounds. TS is a disorder of childhood onset that is characterized by multiple motor and vocal tics. In some cases, there are features of obsessive compulsive disorder (OCD), attention deficit hyperactivity disorder (ADHD), or other behavioral manifestations such as coprolalia, echopraxia, palilalia, and self-injury.^1,2 The spectrum of tic disorders includes:

• Transient tics of childhood (tic duration less than 12 months)
• Chronic motor or vocal tics (lasting more than 12 months), and
• TS (variable motor and vocal tics lasting more than 12 months).

Many children meet the diagnostic criteria for TS between the ages of 6 and 9 years, but symptoms may improve by adulthood. The eventual loss of tics over time reflects the maturation of brain systems that control ballistic action.³

The tics that accompany TS may be defined as simple or complex, and as motor or vocal. Simple motor tics involve only a few muscles, such as eye blinking, shoulder shrugging, or facial grimacing. Complex motor tics involve multiple groups of muscles that are recruited in orchestrated bouts (eg, hand gestures, jumping, touching, or pressing), and may include copropraxia (a sudden tic-like vulgar, sexual, or obscene gesture) or echopraxia (involuntary, spontaneous imitation of someone else’s movements). Simple vocal tics are meaningless sounds such as throat clearing, grunting, sniffing, snorting, and chirping. Complex vocal tics involve speech and language such as sudden, spontaneous expression of single words or phrases, or speech blocking.⁴

Tics may be acquired as a consequence of other disorders, including head trauma, encephalitis, stroke, carbon monoxide poisoning, Creutzfeldt-Jakob disease, neurosyphilis, hypoglycemia, or Sydenham chorea.⁵ Genetic disorders such as Huntington disease may be associated with tics. Tics may also occur with certain chromosomal abnormalities or be associated with some neuropsychiatric disorders. Finally, tics may be caused by a large number of medications or illicit drugs, including cocaine, amphetamines, antipsychotics, and antidepressants.

The prevalence of all types of tics in childhood is approximately 6% to 12%, although the prevalence of chronic vocal tics is approximately 1 to 10 per 1,000 children and adolescents.⁶ TS is especially common among autistic children and in those with Asperger syndrome and other autistic spectrum disorders. A survey of patients at Cleveland Clinic Florida found that tics and TS accounted for 8% of all patients with movement disorders. Of patients with tics or TS who were older than 18 years, 70% were male.

PATHOPHYSIOLOGY OF TOURETTE SYNDROME: ROLE OF THE BASAL GANGLIA

Although the pathophysiology of TS is not completely understood, abnormal function of the basal ganglia is thought to be a central component of the disorder. The basal ganglia normally act to facilitate voluntary movements while suppressing competing involuntary ones. Abnormalities of basal ganglia activity are important in several disorders of motor function.⁷ Output neurons from the basal ganglia inhibit thalamic motor nuclei and midbrain neurons of the extrapyramidal motor system, and act to inhibit motor pattern generators in the cerebral cortex and brainstem. Hyperkinetic disorders, including tics, chorea, and dystonia, are thought to result at least in part from impaired inhibition of unwanted motor activity from the basal ganglia to downstream motor centers.⁷

Family heritability studies provide strong support that TS is a genetic disorder. For example, the concordance rate is 86% for monozygotic twins versus 20% for dizygotic twins.⁸ Chromosomes linked to TS include 2p32.2 and 13q31.1.^9,10 Interactions between genetics and environment are also thought to play a significant role. The concept of pediatric autoimmune neuropsychiatric disorders associated with streptococcal (PANDAS) infections has been proposed to explain an apparent temporal association between streptococcal infections and exacerbation of tics. According to this model, molecular mimicry between streptococcal antigens and endogenous brain antigens results in an autoimmune attack.¹¹ However, the identification of specific antibodies against basal ganglia cells remains controversial.¹²

MANAGEMENT OVERVIEW

Accurate diagnosis of TS is essential, and includes a complete history and neurologic examination. The tic phenomenology (complex vs simple) should be characterized, and the patient should be carefully questioned to identify the symptoms that are most bothersome (eg, motor or vocal tics, OCD, or ADHD). Pharmacotherapy should be reserved for problems that are functionally disabling and not remediable by nonpharmacologic interventions.

Treatment may also be required for other neuropsychiatric symptoms. Anxiety and depression have been reported in 19% to 80% of patients with tics, and depression is strongly correlated with the duration and severity of tics.^13,14 Episodic outburst (rage), self-injurious, OCD, antisocial, and oppositional behaviors are all more common among individuals with tic disorders.¹⁵ Personality disorders may be related to OCD, ADHD, or to family or economic issues. Tic disorders are also associated with an increased incidence of somatic complaints, as well as higher rates of academic difficulties, which may be related to ADHD or medications. Sleep disturbances affect an estimated 20% to 50% of patients, and may include difficulty initiating or maintaining sleep, restlessness, movement-related arousal, or parasomnia.¹⁶

Education is an important part of treatment, and may include the patient, family members, teachers or other school staff, and work colleagues. A number of behavioral or psychosocial approaches may help to improve tics, including conditioning techniques, relaxation training, biofeedback, habit reversal, awareness training, and hypnosis.¹⁷

PHARMACOLOGIC TREATMENT: THREE TIERS

First-tier therapies

The alpha-adrenergic blockers clonidine and guanfacine are first-tier therapies. Treatment should be initiated at a low dose and escalated gradually according to response, which is determined by the severity, and not the presence, of tics. Clonidine may be administered at a dose of 0.025 mg two or three times daily or, for maintenance, 0.1 mg three times daily; another option is 0.1, 0.2, or 0.3 mg weekly by transdermal administration. Guanfacine may be administered at a dose of 1 mg once daily. Alpha-adrenergic blockers are useful for the treatment of mild tics, and are considered first-line therapy for tic suppression. Side effects may include dry mouth, somnolence, and, rarely, blood pressure fluctuations.

Agents that affect gamma-amino butyric acid (GABA) neurotransmission have been associated with improved symptoms of tic disorders.¹⁸ For example, both clonazepam and diazepam have been reported to reduce TS symptoms.¹⁸ Both of these benzodiazepines are associated with sedation, blunting of cognition, and exacerbation of depression, however.^19,20

Second-tier therapies

Second-tier therapies, consisting of neuroleptics, induce a rapid treatment response. Haloperidol may be started at a dose of 0.25 mg once daily, with a maintenance dosage of 0.5 to 3.0 mg/day. Cognitive blunting or extrapyramidal side effects are rare in patients with TS, but the potential for these side effects should be thoroughly discussed with the patient or parent/guardian before treatment. Pimozide 0.5 mg (2 to 6 mg/day for maintenance) may be associated with tremor or parkinsonian symptoms (predominantly akinesia). Risperidone 0.25 mg/day (0.5 to 4 mg/day for maintenance), olanzapine 2.5 mg/day (5 to 10 mg/day for maintenance), and quetiapine 25 mg twice daily (100 to 300 mg/day for maintenance) are associated with potential adverse effects of extrapyramidal symptoms, weight gain, and diabetes.

Third-tier therapies

Dopamine agonists (reserpine and tetrabenazine) and botulinum toxin are third-tier therapies. Reserpine, although rarely used in current clinical practice, may be administered at doses of 0.1 to 0.25 mg/day, titrating upward on the basis of clinical response. Tetrabenazine may be administered at a starting dose of 12.5 mg/day, with higher doses as needed depending on the response to treatment. Adverse effects include hypotension, sedation, extrapyramidal symptoms (predominantly parkinsonism), and depression.²¹

The exact mechanism by which tetrabenazine produces this suppression effect is unknown, but it is believed to be related to its effect of reversibly depleting monoamines. At least three neuronal protein classes regulate the effects of dopamine on voluntary and involuntary movement.²² The two presynaptic proteins are vesicular monoamine transporter subtype 2 (VMAT2) and dopamine transporter (DAT). Postsynaptically, dopamine activity is regulated by G-protein–linked dopamine receptors (eg, the D₂ receptor). Tetrabenazine reduces the uptake of monoamines (including dopamine) into synaptic vesicles by reversibly binding to VMAT2, resulting in degradation of dopamine within axon terminals by monoamine oxidases.²³ By blocking dopamine transport, tetrabenazine depletes dopamine with greater selectivity than it does other monoamines.²⁴

The dosage of tetrabenazine for the treatment of motor disorders, particularly chorea, was established in the Huntington Study Group (HSG) clinical trial.²⁵ In the HSG trial, a starting dose of 12.5 mg on day 1 was increased to 12.5 mg twice daily on days 2 to 7, and then by 12.5 mg/day at weekly intervals until the desired clinical effect, intolerable adverse effects, or a maximum dose of 100 mg/day was reached. Daily dosages of 37.5 mg or more are administered in three divided doses. Adverse events (reported in 70% of patients who received placebo and 91% of patients who received tetrabenazine) include sedation or somnolence, insomnia, and fatigue.²¹ These findings may be carried over to patients with tics.

Botulinum toxin may also help to control tics—especially dystonic tics. The premonitory symptoms of TS are usually unaffected by botulinum toxin.²⁶ The adverse effect profile for patients with TS is similar to that of patients with dystonia or facial dyskinesia, and may include soreness, transient weakness, ptosis (if injected for eye blinking), and mild transient dysphagia (if injected into the larynx).²⁷

MANAGING COMORBID CONDITIONS

Approximately 30% of patients with TS also have OCD.²⁸ Treatment options include selective serotonin reuptake inhibitors at standard doses, and the tricyclic antidepressant clomipramine (25 mg once or twice daily, or 75 mg/day in sustained-release form). Trazodone, a serotonin antagonist and reuptake inhibitor that is associated with a lower incidence of anticholinergic effects, may be initiated at a dose of 50 mg/day and slowly increased to 150 to 400 mg/day depending on clinical response.

As many as 60% of patients with TS may also have ADHD.²⁸ Methylphenidate is helpful for the treatment of ADHD and does not exacerbate tics, but it is a restricted medication. The recommended dose is 20 mg once daily, titrated upward as needed based on response. Atomoxetine carries a warning regarding increased risk of suicide. It has also been associated with an increased risk of sexual dysfunction and behavioral changes, including aggressive behaviors, agitation, and irratibility.²⁹

DEEP BRAIN STIMULATION

Deep brain stimulation (DBS) has been shown to improve TS in single-case studies and in small series, although the long-term benefit is unclear. Potential targets of stimulation include midline thalamic centromedian-parafascicular (CM-PF) nuclei, the ventralis oralis complex of the thalamus, motor and limbic globus pallidus pars interna (GPi), and the anterior limb of the internal capsule.³⁰ In particular, stimulation of the sensorimotor GPi may ameliorate hyperkinetic states.

One report described the results of DBS implantation in a 15-year-old boy with TS who had not responded to several pharmacologic treatment options.³¹ Six months after implantation, the patient exhibited markedly improved tic severity as measured using the Yale Global Tic Severity Scale, including a 76% reduction in motor tic severity, 68% reduction in vocal tics, and a complete resolution of impairment.³¹

Published consensus criteria for the selection of suitable candidates for DBS include age greater than 25 years, chronic and severe tics with severe functional impairment for at least 12 months, tics that are frequent and noticeable in most situations most of the time, failure of conventional medical therapy, medical stability for 6 months, and willingness to participate in ongoing psychologic interventions.³² Exclusion criteria include the presence of another medical condition that could explain the tics, an unstable medical condition, being considered likely to benefit from psychologic interventions, psychosocial factors that may complicate the recovery process or make it difficult to assess outcome, and unwillingness to participate in ongoing treatment for psychosocial problems or risk factors. Other factors that should be considered include comorbidities, the variability in tic severity over time, the involvement of a multidisciplinary treatment team, results of a thorough neuropsychologic assessment, expertise of the surgical team, and access to imaging facilities for presurgical mapping and postsurgical evaluation.

SUMMARY AND CONCLUSIONS

Tourette syndrome is not uncommon among the adult population of a typical neurology practice, and should not be considered exclusively a pediatric diagnosis. Several treatment options are available, including behavioral approaches and several medications. Treatment should focus on the most disabling symptoms. Neuropsychologic assessment and psychiatric support may be necessary for some patients. The same comorbidities that are encountered in children are usually evident in adult patients as well. In medically refractory cases, DBS surgery may be helpful.

Tourette syndrome (TS) is part of a spectrum of tic disorders. Tics are sudden, rapid, stereotyped, repetitive, nonrhythmic movements or vocalizations affecting discrete muscle groups, and are preceded by a sensory component. Patients in whom tic suppression is attempted report the experience of a sensation of inner pressure that must be released. This eventually results in the performance of motor movement or vocal sounds. TS is a disorder of childhood onset that is characterized by multiple motor and vocal tics. In some cases, there are features of obsessive compulsive disorder (OCD), attention deficit hyperactivity disorder (ADHD), or other behavioral manifestations such as coprolalia, echopraxia, palilalia, and self-injury.^1,2 The spectrum of tic disorders includes:

• Transient tics of childhood (tic duration less than 12 months)
• Chronic motor or vocal tics (lasting more than 12 months), and
• TS (variable motor and vocal tics lasting more than 12 months).

Many children meet the diagnostic criteria for TS between the ages of 6 and 9 years, but symptoms may improve by adulthood. The eventual loss of tics over time reflects the maturation of brain systems that control ballistic action.³

The tics that accompany TS may be defined as simple or complex, and as motor or vocal. Simple motor tics involve only a few muscles, such as eye blinking, shoulder shrugging, or facial grimacing. Complex motor tics involve multiple groups of muscles that are recruited in orchestrated bouts (eg, hand gestures, jumping, touching, or pressing), and may include copropraxia (a sudden tic-like vulgar, sexual, or obscene gesture) or echopraxia (involuntary, spontaneous imitation of someone else’s movements). Simple vocal tics are meaningless sounds such as throat clearing, grunting, sniffing, snorting, and chirping. Complex vocal tics involve speech and language such as sudden, spontaneous expression of single words or phrases, or speech blocking.⁴

Tics may be acquired as a consequence of other disorders, including head trauma, encephalitis, stroke, carbon monoxide poisoning, Creutzfeldt-Jakob disease, neurosyphilis, hypoglycemia, or Sydenham chorea.⁵ Genetic disorders such as Huntington disease may be associated with tics. Tics may also occur with certain chromosomal abnormalities or be associated with some neuropsychiatric disorders. Finally, tics may be caused by a large number of medications or illicit drugs, including cocaine, amphetamines, antipsychotics, and antidepressants.

The prevalence of all types of tics in childhood is approximately 6% to 12%, although the prevalence of chronic vocal tics is approximately 1 to 10 per 1,000 children and adolescents.⁶ TS is especially common among autistic children and in those with Asperger syndrome and other autistic spectrum disorders. A survey of patients at Cleveland Clinic Florida found that tics and TS accounted for 8% of all patients with movement disorders. Of patients with tics or TS who were older than 18 years, 70% were male.

PATHOPHYSIOLOGY OF TOURETTE SYNDROME: ROLE OF THE BASAL GANGLIA

Although the pathophysiology of TS is not completely understood, abnormal function of the basal ganglia is thought to be a central component of the disorder. The basal ganglia normally act to facilitate voluntary movements while suppressing competing involuntary ones. Abnormalities of basal ganglia activity are important in several disorders of motor function.⁷ Output neurons from the basal ganglia inhibit thalamic motor nuclei and midbrain neurons of the extrapyramidal motor system, and act to inhibit motor pattern generators in the cerebral cortex and brainstem. Hyperkinetic disorders, including tics, chorea, and dystonia, are thought to result at least in part from impaired inhibition of unwanted motor activity from the basal ganglia to downstream motor centers.⁷

Family heritability studies provide strong support that TS is a genetic disorder. For example, the concordance rate is 86% for monozygotic twins versus 20% for dizygotic twins.⁸ Chromosomes linked to TS include 2p32.2 and 13q31.1.^9,10 Interactions between genetics and environment are also thought to play a significant role. The concept of pediatric autoimmune neuropsychiatric disorders associated with streptococcal (PANDAS) infections has been proposed to explain an apparent temporal association between streptococcal infections and exacerbation of tics. According to this model, molecular mimicry between streptococcal antigens and endogenous brain antigens results in an autoimmune attack.¹¹ However, the identification of specific antibodies against basal ganglia cells remains controversial.¹²

MANAGEMENT OVERVIEW

Accurate diagnosis of TS is essential, and includes a complete history and neurologic examination. The tic phenomenology (complex vs simple) should be characterized, and the patient should be carefully questioned to identify the symptoms that are most bothersome (eg, motor or vocal tics, OCD, or ADHD). Pharmacotherapy should be reserved for problems that are functionally disabling and not remediable by nonpharmacologic interventions.

Treatment may also be required for other neuropsychiatric symptoms. Anxiety and depression have been reported in 19% to 80% of patients with tics, and depression is strongly correlated with the duration and severity of tics.^13,14 Episodic outburst (rage), self-injurious, OCD, antisocial, and oppositional behaviors are all more common among individuals with tic disorders.¹⁵ Personality disorders may be related to OCD, ADHD, or to family or economic issues. Tic disorders are also associated with an increased incidence of somatic complaints, as well as higher rates of academic difficulties, which may be related to ADHD or medications. Sleep disturbances affect an estimated 20% to 50% of patients, and may include difficulty initiating or maintaining sleep, restlessness, movement-related arousal, or parasomnia.¹⁶

Education is an important part of treatment, and may include the patient, family members, teachers or other school staff, and work colleagues. A number of behavioral or psychosocial approaches may help to improve tics, including conditioning techniques, relaxation training, biofeedback, habit reversal, awareness training, and hypnosis.¹⁷

PHARMACOLOGIC TREATMENT: THREE TIERS

First-tier therapies

The alpha-adrenergic blockers clonidine and guanfacine are first-tier therapies. Treatment should be initiated at a low dose and escalated gradually according to response, which is determined by the severity, and not the presence, of tics. Clonidine may be administered at a dose of 0.025 mg two or three times daily or, for maintenance, 0.1 mg three times daily; another option is 0.1, 0.2, or 0.3 mg weekly by transdermal administration. Guanfacine may be administered at a dose of 1 mg once daily. Alpha-adrenergic blockers are useful for the treatment of mild tics, and are considered first-line therapy for tic suppression. Side effects may include dry mouth, somnolence, and, rarely, blood pressure fluctuations.

Agents that affect gamma-amino butyric acid (GABA) neurotransmission have been associated with improved symptoms of tic disorders.¹⁸ For example, both clonazepam and diazepam have been reported to reduce TS symptoms.¹⁸ Both of these benzodiazepines are associated with sedation, blunting of cognition, and exacerbation of depression, however.^19,20

Second-tier therapies

Second-tier therapies, consisting of neuroleptics, induce a rapid treatment response. Haloperidol may be started at a dose of 0.25 mg once daily, with a maintenance dosage of 0.5 to 3.0 mg/day. Cognitive blunting or extrapyramidal side effects are rare in patients with TS, but the potential for these side effects should be thoroughly discussed with the patient or parent/guardian before treatment. Pimozide 0.5 mg (2 to 6 mg/day for maintenance) may be associated with tremor or parkinsonian symptoms (predominantly akinesia). Risperidone 0.25 mg/day (0.5 to 4 mg/day for maintenance), olanzapine 2.5 mg/day (5 to 10 mg/day for maintenance), and quetiapine 25 mg twice daily (100 to 300 mg/day for maintenance) are associated with potential adverse effects of extrapyramidal symptoms, weight gain, and diabetes.

Third-tier therapies

Dopamine agonists (reserpine and tetrabenazine) and botulinum toxin are third-tier therapies. Reserpine, although rarely used in current clinical practice, may be administered at doses of 0.1 to 0.25 mg/day, titrating upward on the basis of clinical response. Tetrabenazine may be administered at a starting dose of 12.5 mg/day, with higher doses as needed depending on the response to treatment. Adverse effects include hypotension, sedation, extrapyramidal symptoms (predominantly parkinsonism), and depression.²¹

The exact mechanism by which tetrabenazine produces this suppression effect is unknown, but it is believed to be related to its effect of reversibly depleting monoamines. At least three neuronal protein classes regulate the effects of dopamine on voluntary and involuntary movement.²² The two presynaptic proteins are vesicular monoamine transporter subtype 2 (VMAT2) and dopamine transporter (DAT). Postsynaptically, dopamine activity is regulated by G-protein–linked dopamine receptors (eg, the D₂ receptor). Tetrabenazine reduces the uptake of monoamines (including dopamine) into synaptic vesicles by reversibly binding to VMAT2, resulting in degradation of dopamine within axon terminals by monoamine oxidases.²³ By blocking dopamine transport, tetrabenazine depletes dopamine with greater selectivity than it does other monoamines.²⁴

The dosage of tetrabenazine for the treatment of motor disorders, particularly chorea, was established in the Huntington Study Group (HSG) clinical trial.²⁵ In the HSG trial, a starting dose of 12.5 mg on day 1 was increased to 12.5 mg twice daily on days 2 to 7, and then by 12.5 mg/day at weekly intervals until the desired clinical effect, intolerable adverse effects, or a maximum dose of 100 mg/day was reached. Daily dosages of 37.5 mg or more are administered in three divided doses. Adverse events (reported in 70% of patients who received placebo and 91% of patients who received tetrabenazine) include sedation or somnolence, insomnia, and fatigue.²¹ These findings may be carried over to patients with tics.

Botulinum toxin may also help to control tics—especially dystonic tics. The premonitory symptoms of TS are usually unaffected by botulinum toxin.²⁶ The adverse effect profile for patients with TS is similar to that of patients with dystonia or facial dyskinesia, and may include soreness, transient weakness, ptosis (if injected for eye blinking), and mild transient dysphagia (if injected into the larynx).²⁷

MANAGING COMORBID CONDITIONS

Approximately 30% of patients with TS also have OCD.²⁸ Treatment options include selective serotonin reuptake inhibitors at standard doses, and the tricyclic antidepressant clomipramine (25 mg once or twice daily, or 75 mg/day in sustained-release form). Trazodone, a serotonin antagonist and reuptake inhibitor that is associated with a lower incidence of anticholinergic effects, may be initiated at a dose of 50 mg/day and slowly increased to 150 to 400 mg/day depending on clinical response.

As many as 60% of patients with TS may also have ADHD.²⁸ Methylphenidate is helpful for the treatment of ADHD and does not exacerbate tics, but it is a restricted medication. The recommended dose is 20 mg once daily, titrated upward as needed based on response. Atomoxetine carries a warning regarding increased risk of suicide. It has also been associated with an increased risk of sexual dysfunction and behavioral changes, including aggressive behaviors, agitation, and irratibility.²⁹

DEEP BRAIN STIMULATION

Deep brain stimulation (DBS) has been shown to improve TS in single-case studies and in small series, although the long-term benefit is unclear. Potential targets of stimulation include midline thalamic centromedian-parafascicular (CM-PF) nuclei, the ventralis oralis complex of the thalamus, motor and limbic globus pallidus pars interna (GPi), and the anterior limb of the internal capsule.³⁰ In particular, stimulation of the sensorimotor GPi may ameliorate hyperkinetic states.

One report described the results of DBS implantation in a 15-year-old boy with TS who had not responded to several pharmacologic treatment options.³¹ Six months after implantation, the patient exhibited markedly improved tic severity as measured using the Yale Global Tic Severity Scale, including a 76% reduction in motor tic severity, 68% reduction in vocal tics, and a complete resolution of impairment.³¹

Published consensus criteria for the selection of suitable candidates for DBS include age greater than 25 years, chronic and severe tics with severe functional impairment for at least 12 months, tics that are frequent and noticeable in most situations most of the time, failure of conventional medical therapy, medical stability for 6 months, and willingness to participate in ongoing psychologic interventions.³² Exclusion criteria include the presence of another medical condition that could explain the tics, an unstable medical condition, being considered likely to benefit from psychologic interventions, psychosocial factors that may complicate the recovery process or make it difficult to assess outcome, and unwillingness to participate in ongoing treatment for psychosocial problems or risk factors. Other factors that should be considered include comorbidities, the variability in tic severity over time, the involvement of a multidisciplinary treatment team, results of a thorough neuropsychologic assessment, expertise of the surgical team, and access to imaging facilities for presurgical mapping and postsurgical evaluation.

SUMMARY AND CONCLUSIONS

Tourette syndrome is not uncommon among the adult population of a typical neurology practice, and should not be considered exclusively a pediatric diagnosis. Several treatment options are available, including behavioral approaches and several medications. Treatment should focus on the most disabling symptoms. Neuropsychologic assessment and psychiatric support may be necessary for some patients. The same comorbidities that are encountered in children are usually evident in adult patients as well. In medically refractory cases, DBS surgery may be helpful.

Tourette syndrome (TS) is part of a spectrum of tic disorders. Tics are sudden, rapid, stereotyped, repetitive, nonrhythmic movements or vocalizations affecting discrete muscle groups, and are preceded by a sensory component. Patients in whom tic suppression is attempted report the experience of a sensation of inner pressure that must be released. This eventually results in the performance of motor movement or vocal sounds. TS is a disorder of childhood onset that is characterized by multiple motor and vocal tics. In some cases, there are features of obsessive compulsive disorder (OCD), attention deficit hyperactivity disorder (ADHD), or other behavioral manifestations such as coprolalia, echopraxia, palilalia, and self-injury.^1,2 The spectrum of tic disorders includes:

• Transient tics of childhood (tic duration less than 12 months)
• Chronic motor or vocal tics (lasting more than 12 months), and
• TS (variable motor and vocal tics lasting more than 12 months).

Many children meet the diagnostic criteria for TS between the ages of 6 and 9 years, but symptoms may improve by adulthood. The eventual loss of tics over time reflects the maturation of brain systems that control ballistic action.³

The tics that accompany TS may be defined as simple or complex, and as motor or vocal. Simple motor tics involve only a few muscles, such as eye blinking, shoulder shrugging, or facial grimacing. Complex motor tics involve multiple groups of muscles that are recruited in orchestrated bouts (eg, hand gestures, jumping, touching, or pressing), and may include copropraxia (a sudden tic-like vulgar, sexual, or obscene gesture) or echopraxia (involuntary, spontaneous imitation of someone else’s movements). Simple vocal tics are meaningless sounds such as throat clearing, grunting, sniffing, snorting, and chirping. Complex vocal tics involve speech and language such as sudden, spontaneous expression of single words or phrases, or speech blocking.⁴

Tics may be acquired as a consequence of other disorders, including head trauma, encephalitis, stroke, carbon monoxide poisoning, Creutzfeldt-Jakob disease, neurosyphilis, hypoglycemia, or Sydenham chorea.⁵ Genetic disorders such as Huntington disease may be associated with tics. Tics may also occur with certain chromosomal abnormalities or be associated with some neuropsychiatric disorders. Finally, tics may be caused by a large number of medications or illicit drugs, including cocaine, amphetamines, antipsychotics, and antidepressants.

The prevalence of all types of tics in childhood is approximately 6% to 12%, although the prevalence of chronic vocal tics is approximately 1 to 10 per 1,000 children and adolescents.⁶ TS is especially common among autistic children and in those with Asperger syndrome and other autistic spectrum disorders. A survey of patients at Cleveland Clinic Florida found that tics and TS accounted for 8% of all patients with movement disorders. Of patients with tics or TS who were older than 18 years, 70% were male.

PATHOPHYSIOLOGY OF TOURETTE SYNDROME: ROLE OF THE BASAL GANGLIA

Although the pathophysiology of TS is not completely understood, abnormal function of the basal ganglia is thought to be a central component of the disorder. The basal ganglia normally act to facilitate voluntary movements while suppressing competing involuntary ones. Abnormalities of basal ganglia activity are important in several disorders of motor function.⁷ Output neurons from the basal ganglia inhibit thalamic motor nuclei and midbrain neurons of the extrapyramidal motor system, and act to inhibit motor pattern generators in the cerebral cortex and brainstem. Hyperkinetic disorders, including tics, chorea, and dystonia, are thought to result at least in part from impaired inhibition of unwanted motor activity from the basal ganglia to downstream motor centers.⁷

Family heritability studies provide strong support that TS is a genetic disorder. For example, the concordance rate is 86% for monozygotic twins versus 20% for dizygotic twins.⁸ Chromosomes linked to TS include 2p32.2 and 13q31.1.^9,10 Interactions between genetics and environment are also thought to play a significant role. The concept of pediatric autoimmune neuropsychiatric disorders associated with streptococcal (PANDAS) infections has been proposed to explain an apparent temporal association between streptococcal infections and exacerbation of tics. According to this model, molecular mimicry between streptococcal antigens and endogenous brain antigens results in an autoimmune attack.¹¹ However, the identification of specific antibodies against basal ganglia cells remains controversial.¹²

MANAGEMENT OVERVIEW

Accurate diagnosis of TS is essential, and includes a complete history and neurologic examination. The tic phenomenology (complex vs simple) should be characterized, and the patient should be carefully questioned to identify the symptoms that are most bothersome (eg, motor or vocal tics, OCD, or ADHD). Pharmacotherapy should be reserved for problems that are functionally disabling and not remediable by nonpharmacologic interventions.

Treatment may also be required for other neuropsychiatric symptoms. Anxiety and depression have been reported in 19% to 80% of patients with tics, and depression is strongly correlated with the duration and severity of tics.^13,14 Episodic outburst (rage), self-injurious, OCD, antisocial, and oppositional behaviors are all more common among individuals with tic disorders.¹⁵ Personality disorders may be related to OCD, ADHD, or to family or economic issues. Tic disorders are also associated with an increased incidence of somatic complaints, as well as higher rates of academic difficulties, which may be related to ADHD or medications. Sleep disturbances affect an estimated 20% to 50% of patients, and may include difficulty initiating or maintaining sleep, restlessness, movement-related arousal, or parasomnia.¹⁶

Education is an important part of treatment, and may include the patient, family members, teachers or other school staff, and work colleagues. A number of behavioral or psychosocial approaches may help to improve tics, including conditioning techniques, relaxation training, biofeedback, habit reversal, awareness training, and hypnosis.¹⁷

PHARMACOLOGIC TREATMENT: THREE TIERS

First-tier therapies

The alpha-adrenergic blockers clonidine and guanfacine are first-tier therapies. Treatment should be initiated at a low dose and escalated gradually according to response, which is determined by the severity, and not the presence, of tics. Clonidine may be administered at a dose of 0.025 mg two or three times daily or, for maintenance, 0.1 mg three times daily; another option is 0.1, 0.2, or 0.3 mg weekly by transdermal administration. Guanfacine may be administered at a dose of 1 mg once daily. Alpha-adrenergic blockers are useful for the treatment of mild tics, and are considered first-line therapy for tic suppression. Side effects may include dry mouth, somnolence, and, rarely, blood pressure fluctuations.

Agents that affect gamma-amino butyric acid (GABA) neurotransmission have been associated with improved symptoms of tic disorders.¹⁸ For example, both clonazepam and diazepam have been reported to reduce TS symptoms.¹⁸ Both of these benzodiazepines are associated with sedation, blunting of cognition, and exacerbation of depression, however.^19,20

Second-tier therapies

Second-tier therapies, consisting of neuroleptics, induce a rapid treatment response. Haloperidol may be started at a dose of 0.25 mg once daily, with a maintenance dosage of 0.5 to 3.0 mg/day. Cognitive blunting or extrapyramidal side effects are rare in patients with TS, but the potential for these side effects should be thoroughly discussed with the patient or parent/guardian before treatment. Pimozide 0.5 mg (2 to 6 mg/day for maintenance) may be associated with tremor or parkinsonian symptoms (predominantly akinesia). Risperidone 0.25 mg/day (0.5 to 4 mg/day for maintenance), olanzapine 2.5 mg/day (5 to 10 mg/day for maintenance), and quetiapine 25 mg twice daily (100 to 300 mg/day for maintenance) are associated with potential adverse effects of extrapyramidal symptoms, weight gain, and diabetes.

Third-tier therapies

Dopamine agonists (reserpine and tetrabenazine) and botulinum toxin are third-tier therapies. Reserpine, although rarely used in current clinical practice, may be administered at doses of 0.1 to 0.25 mg/day, titrating upward on the basis of clinical response. Tetrabenazine may be administered at a starting dose of 12.5 mg/day, with higher doses as needed depending on the response to treatment. Adverse effects include hypotension, sedation, extrapyramidal symptoms (predominantly parkinsonism), and depression.²¹

The exact mechanism by which tetrabenazine produces this suppression effect is unknown, but it is believed to be related to its effect of reversibly depleting monoamines. At least three neuronal protein classes regulate the effects of dopamine on voluntary and involuntary movement.²² The two presynaptic proteins are vesicular monoamine transporter subtype 2 (VMAT2) and dopamine transporter (DAT). Postsynaptically, dopamine activity is regulated by G-protein–linked dopamine receptors (eg, the D₂ receptor). Tetrabenazine reduces the uptake of monoamines (including dopamine) into synaptic vesicles by reversibly binding to VMAT2, resulting in degradation of dopamine within axon terminals by monoamine oxidases.²³ By blocking dopamine transport, tetrabenazine depletes dopamine with greater selectivity than it does other monoamines.²⁴

The dosage of tetrabenazine for the treatment of motor disorders, particularly chorea, was established in the Huntington Study Group (HSG) clinical trial.²⁵ In the HSG trial, a starting dose of 12.5 mg on day 1 was increased to 12.5 mg twice daily on days 2 to 7, and then by 12.5 mg/day at weekly intervals until the desired clinical effect, intolerable adverse effects, or a maximum dose of 100 mg/day was reached. Daily dosages of 37.5 mg or more are administered in three divided doses. Adverse events (reported in 70% of patients who received placebo and 91% of patients who received tetrabenazine) include sedation or somnolence, insomnia, and fatigue.²¹ These findings may be carried over to patients with tics.

Botulinum toxin may also help to control tics—especially dystonic tics. The premonitory symptoms of TS are usually unaffected by botulinum toxin.²⁶ The adverse effect profile for patients with TS is similar to that of patients with dystonia or facial dyskinesia, and may include soreness, transient weakness, ptosis (if injected for eye blinking), and mild transient dysphagia (if injected into the larynx).²⁷

MANAGING COMORBID CONDITIONS

Approximately 30% of patients with TS also have OCD.²⁸ Treatment options include selective serotonin reuptake inhibitors at standard doses, and the tricyclic antidepressant clomipramine (25 mg once or twice daily, or 75 mg/day in sustained-release form). Trazodone, a serotonin antagonist and reuptake inhibitor that is associated with a lower incidence of anticholinergic effects, may be initiated at a dose of 50 mg/day and slowly increased to 150 to 400 mg/day depending on clinical response.

As many as 60% of patients with TS may also have ADHD.²⁸ Methylphenidate is helpful for the treatment of ADHD and does not exacerbate tics, but it is a restricted medication. The recommended dose is 20 mg once daily, titrated upward as needed based on response. Atomoxetine carries a warning regarding increased risk of suicide. It has also been associated with an increased risk of sexual dysfunction and behavioral changes, including aggressive behaviors, agitation, and irratibility.²⁹

DEEP BRAIN STIMULATION

Deep brain stimulation (DBS) has been shown to improve TS in single-case studies and in small series, although the long-term benefit is unclear. Potential targets of stimulation include midline thalamic centromedian-parafascicular (CM-PF) nuclei, the ventralis oralis complex of the thalamus, motor and limbic globus pallidus pars interna (GPi), and the anterior limb of the internal capsule.³⁰ In particular, stimulation of the sensorimotor GPi may ameliorate hyperkinetic states.

One report described the results of DBS implantation in a 15-year-old boy with TS who had not responded to several pharmacologic treatment options.³¹ Six months after implantation, the patient exhibited markedly improved tic severity as measured using the Yale Global Tic Severity Scale, including a 76% reduction in motor tic severity, 68% reduction in vocal tics, and a complete resolution of impairment.³¹

Published consensus criteria for the selection of suitable candidates for DBS include age greater than 25 years, chronic and severe tics with severe functional impairment for at least 12 months, tics that are frequent and noticeable in most situations most of the time, failure of conventional medical therapy, medical stability for 6 months, and willingness to participate in ongoing psychologic interventions.³² Exclusion criteria include the presence of another medical condition that could explain the tics, an unstable medical condition, being considered likely to benefit from psychologic interventions, psychosocial factors that may complicate the recovery process or make it difficult to assess outcome, and unwillingness to participate in ongoing treatment for psychosocial problems or risk factors. Other factors that should be considered include comorbidities, the variability in tic severity over time, the involvement of a multidisciplinary treatment team, results of a thorough neuropsychologic assessment, expertise of the surgical team, and access to imaging facilities for presurgical mapping and postsurgical evaluation.

SUMMARY AND CONCLUSIONS

Tourette syndrome is not uncommon among the adult population of a typical neurology practice, and should not be considered exclusively a pediatric diagnosis. Several treatment options are available, including behavioral approaches and several medications. Treatment should focus on the most disabling symptoms. Neuropsychologic assessment and psychiatric support may be necessary for some patients. The same comorbidities that are encountered in children are usually evident in adult patients as well. In medically refractory cases, DBS surgery may be helpful.

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.