Hypertrophic cardiomyopathy: Management of ventricular arrhythmias and sudden cardiac death risk

INTRODUCTION — 

Hypertrophic cardiomyopathy (HCM) is a genetic disease of the myocardium caused by mutations in one of several genes that encode components of the contractile apparatus of the heart. (See "Hypertrophic cardiomyopathy: Gene mutations and clinical genetic testing".)

In broad terms, the symptoms related to HCM can be categorized as those related to heart failure (HF), chest pain, or arrhythmias. Patients with HCM are prone to both atrial and ventricular arrhythmias. Many ventricular arrhythmias are asymptomatic, but some can precipitate hemodynamic collapse and sudden cardiac death (SCD).

The evaluation and management of ventricular arrhythmias and SCD risk in patients with HCM will be reviewed here.

Other issues related to HCM are discussed separately:

●(See "Hypertrophic cardiomyopathy: Clinical manifestations, diagnosis, and evaluation".)

●(See "Hypertrophic cardiomyopathy in adults: Supraventricular tachycardias including atrial fibrillation".)

●(See "Hypertrophic cardiomyopathy: Management of patients with outflow tract obstruction".)

●(See "Hypertrophic cardiomyopathy: Management of patients without outflow tract obstruction".)

PATHOGENESIS — 

An abnormal myocardial substrate comprised of myocyte disarray (picture 1), interstitial fibrosis, and replacement fibrosis provides the likely structural nidus for the generation of ventricular arrhythmias in patients with HCM. This substrate can be acted upon by factors such as myocardial ischemia, left ventricular (LV) outflow tract obstruction, and abnormal vascular response with inappropriate vasodilatation, as well as high adrenergic states (eg, exercise) that can lower the threshold for initiating ventricular tachycardia (VT) or ventricular fibrillation.

EPIDEMIOLOGY — 

Ventricular arrhythmias are common in patients with HCM and can range from isolated premature ventricular complexes/contractions (PVCs; also referred to as premature ventricular beats or premature ventricular depolarizations) to sustained VT and ventricular fibrillation. As an example, in a study of 178 patients who underwent 24-hour ambulatory monitoring, PVCs were highly prevalent (88 percent; 12 percent had ≥500 PVCs) [1].

Nonsustained VT (NSVT) is also present in 15 to 30 percent of patients with HCM [1-4]. NSVT is more likely in older patients and is associated with greater LV wall thickening and New York Heart Association class III or IV symptoms (table 1). The prevalence of NSVT is less common in young patients (<40 years old) with HCM [5].

Clinically documented sustained VT is relatively rare, with the annual incidence of sudden cardiac arrest (SCA) approximately 1 percent or less [1,6-9].

CLINICAL MANIFESTATIONS — 

The presentation of ventricular arrhythmias in patients with HCM is highly variable, ranging from an absence of symptoms to palpitations to SCA, but, in general, the presentation of ventricular arrhythmias is similar to their presentation in other types of patients without HCM.

●Most patients with premature ventricular complexes/contractions (PVCs) or nonsustained VT (NSVT) will be asymptomatic or have intermittent palpitations.

●Sustained VT most often results in palpitations, presyncope, or syncope.

●SCA, although rare, can be the initial presentation of sustained VT or ventricular fibrillation (VF).

More detailed information on the presenting symptoms of PVCs, NSVT, sustained VT, and VF is discussed separately. (See "Premature ventricular complexes: Clinical presentation and diagnostic evaluation", section on 'Symptoms' and "Nonsustained ventricular tachycardia: Clinical manifestations, evaluation, and management", section on 'History' and "Sustained monomorphic ventricular tachycardia: Clinical manifestations, diagnosis, and evaluation", section on 'History and associated symptoms'.)

THERAPY FOR SYMPTOMATIC VENTRICULAR ARRHYTHMIAS

Premature ventricular beats — Patients with HCM who have symptomatic premature ventricular complexes/contractions are typically treated similarly to patients without HCM. The approach to therapy is discussed separately. (See "Premature ventricular complexes: Treatment and prognosis".)

Symptomatic NSVT — In patients with HCM and symptomatic nonsustained VT (NSVT), we suggest initial therapy with a beta blocker rather than other agents. In patients with NSVT and symptoms despite maximal beta blocker therapy, options for therapy include sotalol and amiodarone [10,11]. The details on therapy with these agents are discussed separately. (See "Clinical uses of sotalol" and "Nonsustained ventricular tachycardia: Clinical manifestations, evaluation, and management", section on 'Treatment of patients with structurally abnormal hearts'.)

There are no data in patients with HCM to guide therapy. Evidence in patients with other forms of structural heart disease is described separately. (See "Nonsustained ventricular tachycardia: Clinical manifestations, evaluation, and management".)

Sustained ventricular tachycardia — For patients with frequent ventricular arrhythmias or multiple implantable cardioverter-defibrillator (ICD) shocks, adjunctive antiarrhythmic pharmacotherapy is indicated. Antiarrhythmic therapy for such patients is discussed separately. (See "Secondary prevention of sudden cardiac death in heart failure and cardiomyopathy", section on 'Other treatment options'.)

In patients with HCM and VT resistant to pharmacotherapy, ablation may be an option for therapy. However, ablation is more likely to be effective in patients with an apical aneurysm or another well-defined nidus for VT. In patients with an apical aneurysm, catheter ablation may be particularly effective; the structural nidus for VT is often identifiable at the junction of the aneurysm rim and LV myocardium and is accessible for ablation [12]. On the other hand, in the remainder of the HCM population, the diffuse abnormal myocardial substrate results in multiple foci for VT, which is a situation that is generally less amenable to catheter ablation [13].

Additional options for patients with VT refractory to medical therapy are discussed elsewhere. (See "Electrical storm and incessant ventricular tachycardia".)

Restriction of physical activity — In patients with HCM, activity restriction may be appropriate. This issue is covered in detail separately. (See "Athletes: Overview of sudden cardiac death risk and sport participation", section on 'Hypertrophic cardiomyopathy' and "Athletes with arrhythmias: Treatment and returning to athletic participation", section on 'Athletes with ICDs'.)

Ouflow tract obstruction and arrhythmias — While LV outflow tract (LVOT) obstruction may increase the risk of sudden cardiac death (SCD), septal reduction therapies (eg, septal myectomy, alcohol septal ablation) are only indicated for relief of symptoms related to LVOT obstruction; such therapies are not expected to decrease the risk of SCD in patients with HCM. Retrospective cohort studies suggest that the risk of SCD or appropriate ICD shocks is lower following septal myectomy compared with nonoperated patients, but selection of patients for myectomy may lead significant bias [14,15].

SUDDEN DEATH RISK ASSESSMENT

General approach to evaluation — In patients with HCM, the evaluation for SCD risk is based on components of the patient's personal and family history as well as the results of cardiovascular imaging and testing. In general, patients present in one of two broad categories:

●History of SCA or sustained VT – In patients with HCM and a history of SCA or documentation of sustained ventricular arrhythmias, an ICD for secondary prevention is indicated, and no further evaluation of SCD risk is necessary. The approach to ICD placement in such patients is discussed elsewhere in this topic. (See 'Secondary prevention' below.)

●No history of SCA or sustained VT – In patients without a history of SCA or sustained arrhythmias, the approach to assessing SCD risk is as follows:

•Initial approach – The initial approach to risk stratification depends on the method used to assign risk (eg, risk factor- or risk model-based method), but typically consists of the following:

-History – Clinicians should obtain a detailed history focused on identifying episodes of syncope likely caused by arrhythmias in the patient or a first-degree relative.

-Echocardiography – Echocardiography is typically performed to diagnose HCM. The key features on echocardiography that inform SCD risk stratification are LV hypertrophy, left atrial size, LV ejection fraction (LVEF), and LVOT gradient. The details on the exact criteria are discussed elsewhere in this topic. (See 'Primary prevention' below.)

-Ambulatory electrocardiogram monitoring – Each patient should undergo ambulatory monitoring to assess for nonsustained VT (NSVT).

-Assessment of blood pressure response to exercise – One contributor to this topic routinely assesses the blood pressure response to exercise as part of the initial assessment.

Once this assessment is complete, we estimate the likelihood of SCD using one of two approaches, which are discussed elsewhere in this topic. (See 'Primary prevention' below.)

•Additional assessment – In patients in whom the risk of SCD is unclear or low after initial evaluation, the approach to risk assessment differs between the contributors to this topic. One contributor to this topic favors obtaining cardiovascular magnetic resonance (CMR) imaging to assess for late gadolinium enhancement (LGE), while another typically reviews available testing for gene data and the blood pressure response to exercise.

Risk factors — Patient characteristics with a consistent association with SCD risk include:

●Prior episode of SCA or sustained ventricular arrhythmia in the patient or a first-degree relative

●Syncope judged to be arrhythmic origin in the patient or a first degree relative

●LV hypertrophy

●Left atrial diameter

●LVEF or fractional shortening

●Age

●NSVT on ambulatory monitoring

●CMR evidence of LGE

Risk factors whose association with SCD risk is less clear include:

●Presence of an LV apical aneurysm

●Genotype status

●Abnormal blood pressure response to exercise

In general, the strength of association between patient characteristics and SCD risk is difficult to compare between studies and varies between factors. Direct comparisons are limited by factors that include differences between study outcomes (eg, sudden death alone versus sudden death plus appropriate ICD shocks) and differences in the prevailing use of medical, surgical, and device therapies between populations and over time.

Larger studies that evaluated the association between risk factors and SCD risk using a common definition of SCD may provide a more internally consistent measurement of the relative strength of risk factors:

●In a consortium study that included 3675 patients with HCM and that evaluated the outcome of SCD or appropriate ICD shocks, there was an association between the composite outcome of SCD or appropriate ICD shocks and the following factors [16]:

•Age (hazard ratio [HR] per year 0.98, 95% CI 0.97-0.99)

•Maximal LV wall thickness (HR per mm 1.2, 95% CI 1.01-1.4)

•Left atrial diameter (HR per mm 1.03, 95% CI 1.01-1.05)

•LV outflow gradient (HR per mmHg 1.004, 95% CI 1.001-1.01)

•Family history of SCD (HR 1.6, 95% CI 1.2-2.1)

•NSVT on ambulatory monitoring (HR 2.3, 95% CI 1.6-3.2)

•Unexplained syncope (HR 2.1, 95% CI 1.5-2.8)

●In another study that included 774 patients with HCM, LGE ≥15 percent and LVEF <50 percent were associated with the risk of SCD or aborted SCD [17]. The addition of these factors added prognostic information to the 2022 European Society of Cardiology risk model estimate and the 2020 American Heart Association/American College of Cardiology risk criteria.

Risk factors whose strength of association are less clear due to conflicting data or incomplete data include:

●The association between LV apical aneurysm and SCD is unclear. In a study that compared the addition of LV apical aneurysm with other guideline-based criteria, LV apical aneurysm was not associated with the rate of SCD [17]. In a cohort of 1940 patients with HCM, patients with an apical aneurysm had a SCD rate higher than those without an apical aneurysm (6.4 versus 2.0 percent per year) [12]. The risk of SCD conferred by an apical aneurysm may extend later into life [18].

●Genotype testing is variably used in the assessment of SCD risk. While there are reports of genotypes associated with a high risk of SCD, particularly troponin T and several beta myosin-heavy chain mutations, the available data are derived from a small number of families and may overestimate risk [19-24]. Moreover, most mutations that cause HCM are novel (ie, "private mutations"), and their association with SCD risk cannot be reliably determined. (See "Hypertrophic cardiomyopathy: Gene mutations and clinical genetic testing".)

●Blood pressure response to exercise was considered an important risk factor for SCD in older guidelines. One study suggested that patients with an abnormal blood pressure response (ie, failure to increase systolic blood pressure by more than 20 mmHg during exercise relative to resting blood pressure or a drop in blood pressure by at least 20 mmHg during exercise relative to peak blood pressure) had a higher risk of SCD (15 versus 3 percent) [25]. However, later studies either did not evaluate the association between blood pressure response and SCD due to incomplete data or did not find an association between blood pressure response and SCD in the presence of other risk factors [4,26].

Risk prediction models — The main risk model used to estimate a probability of life-threatening arrhythmias based on an individual patient's characteristics was shown to be reasonably accurate in derivation and validation studies [16]:

●In a retrospective cohort study involving 3675 patients from six European centers (2082 in the development cohort and 1593 in the validation cohort) with a median follow-up of 5.7 years, the primary outcome of SCD or appropriate ICD shock occurred in 198 patients (118 patients with SCD, 27 with aborted SCD, and 53 with appropriate ICD shock) [16]. The derived model included the variables age, maximal LV wall thickness, left atrial diameter, LVOT gradient, family history of SCD, NSVT, and unexplained syncope and had reasonable accuracy (C-index 0.7, calibration slope 95% CI 0.74-1.08). In the validation cohort, accuracy decreased (C-index 0.67, calibration slope 95% CI 0.67-1.24). Risk factors alone did not have improved accuracy (C-index 0.54, calibration slope 95% CI 0.73-1.3) compared with the multivariable risk model. The calculator can be found online.

●Subsequent validation studies of the HCM Risk-SCD calculator reported results similar to the derivation study [27-30]. In a 2019 meta-analysis of validation studies that included 7291 patients with HCM (including the original HCM Risk-SCD cohort and five subsequent cohorts), 70 percent of patients were identified as low risk, 15 percent as intermediate risk, and 15 percent as high risk for SCD or appropriate ICD shock [31]. In total, 184 SCD events occurred, and the incidence of SCD events was 1, 2.4, and 8.4 percent in the low, intermediate, and high-risk groups, respectively. Sixty-eight percent of events occurred in patients classified as intermediate or high risk by the HCM Risk-SCD calculator. Accuracy measures in this analysis were similar to those in the derivation study, though differences in methods between studies limit the ability to make direct comparisons.

The approach to use of this model in the process of SCD risk estimation and ICD placement is discussed elsewhere in this topic. (See 'Primary prevention' below.)

RECOMMENDATIONS FOR ICD THERAPY

Secondary prevention — Similar to other patients who survive an episode of sustained VT, ventricular fibrillation, or SCA, patients with HCM who survive one of these episodes should receive an ICD.

This approach is consistent with professional guidelines [32,33].

This approach is primarily based on indirect evidence from randomized trials conducted in patients who predominantly had ischemic cardiomyopathy as well as retrospective studies of patients with HCM. The evidence on secondary prevention of SCD is presented elsewhere. (See "Secondary prevention of sudden cardiac death in heart failure and cardiomyopathy".)

In patients with HCM who underwent ICD placement for secondary prevention of SCD risk, the rate of appropriate ICD therapy is approximately 10 percent per year [34,35].

Primary prevention — In patients with HCM who have an increased risk of SCD based on the presence of one or more risk factors for SCD, we suggest placement of an ICD. There are two methods for risk stratification:

In one method, the presence of a single risk factor confers increased risk (algorithm 1). The risk factors include:

●Family history of SCD, premature death, or sustained ventricular arrhythmia in a first-degree or close relative attributable to HCM.

●LV hypertrophy ≥30 mm by echocardiography or CMR.

●Personal history of syncope attributable to ventricular arrythmia based on clinical history.

●LV apical aneurysm with transmural scar.

●LVEF <50 percent.

●Nonsustained VT (NSVT) on ambulatory monitoring. For HCM risk stratification, NSVT is defined as ≥3 consecutive beats at ≥120 beats per minute with greater weight placed on runs of NSVT that are frequent (eg, ≥3 episodes per 48 hours), longer (eg, ≥10 beats), or faster (eg, ≥200 beats per minute).

●Late gadolinium enhancement (LGE) involving ≥15 percent of the LV myocardium.

This approach is most appropriate for patients <60 years old. Patients with HCM ≥60 years are at low risk for disease-related adverse events, including SCD, even in the presence of the risk factors noted above [5].

Alternatively, the risk of SCD can be estimated with the HCM Risk-SCD risk model; patients with a five-year HCM Risk-SCD estimate ≥4 percent or who have an estimate <4 percent and at least one risk factor for SCD have an increased risk of SCD (algorithm 2). In this method, the risk factors are:

●LVEF <50 percent.

●Presence of an apical aneurysm with transmural scar.

●Extensive LGE, typically defined as ≥15 percent of LV mass.

●One contributor to this topic uses an abnormal upright exercise blood pressure response and genetic variants strongly associated with premature sudden death as markers of increased risk. Abnormal blood pressure response is defined as failure to increase systolic blood pressure by more than 20 mmHg during exercise relative to resting blood pressure or a drop in blood pressure by at least 20 mmHg during exercise relative to peak blood pressure.

With this method, the higher the five-year estimate of SCD, the more likely the patient will benefit from ICD placement.

These approaches are consistent with North American and European guidelines, respectively. However, the guidelines suggest a higher certainty of evidence [32,33]. The European Society of Cardiology (ESC) guidelines recommend calculation of risk and use of specific thresholds to guide ICD decisions, while North American guidelines emphasize the use of a risk score as a shared decision-making tool. The ESC guidelines use genotype and blood pressure response as well as different NSVT criteria for estimation of SCD risk compared with the North American guidelines.

In patients with HCM, there are no prospective trials to guide selection of patients for ICD placement for the primary prevention of SCD. Thus, the approach to primary prevention ICD placement is based on retrospective studies that evaluate the association between individual risk factors and malignant arrhythmias and sudden death. As an example, in a single-center study that included 2094 consecutive patients evaluated over a 17-year period, the American Heart Association/American College of Cardiology risk criteria approach had a sensitivity of 87 percent (95% CI 79-83) and specificity of 78 percent (95% CI 76-80), while the European Society of Cardiology risk model approach had a sensitivity of 34 percent (95% CI 24-44) and specificity of 92 percent (95% CI 91-94) for the five-year risk of SCD or appropriate ICD shock [36].

Studies of individual risk factors and models are found elsewhere in this topic. (See 'Risk factors' above and 'Risk prediction models' above.)

Choice of device — The approach to the choice of device in patients with HCM is generally similar to the approach to the choice of device in patients with other indications for an ICD. The details on choice of ICD are described elsewhere. (See "Implantable cardioverter-defibrillators: Choosing a device and system descriptions".)

There are limited data on the relative efficacy of various types of ICD in patients with HCM, and the available data suggest that the advantages and disadvantages of subcutaneous ICDs are similar in patients with or without HCM [37-40].

Complications of device therapy — Long-term complications following ICD placement include the following [41-43]:

●Approximately 25 percent of patients experience inappropriate ICD discharge

●6 to 13 percent experience lead complications (eg, fracture, dislodgment, oversensing)

●4 to 5 percent develop device-related infection

●2 to 3 percent experience bleeding or thrombosis

The rate of inappropriate shocks and lead fractures appears to be higher in children than in adults, largely because their activity level and body growth place continual strain on the leads, which are the weakest link in the system [43]. This issue is of particular concern, given the long periods that young patients will have prophylactically implanted devices. Further information on device complications can be found separately. (See "Cardiac implantable electronic devices: Long-term complications".)

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Cardiomyopathy" and "Society guideline links: Arrhythmias in adults" and "Society guideline links: Ventricular arrhythmias" and "Society guideline links: Cardiac implantable electronic devices" and "Society guideline links: Catheter ablation of arrhythmias".)

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Hypertrophic cardiomyopathy in adults (The Basics)")

●Beyond the Basics topic (see "Patient education: Hypertrophic cardiomyopathy (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Clinical manifestations – The presentation of ventricular arrhythmias in patients with hypertrophic cardiomyopathy (HCM) is highly variable, ranging from an absence of symptoms to palpitations to sudden cardiac arrest (SCA), but, in general, the presentation of ventricular arrhythmias is similar to their presentation in other types of patients without HCM. (See 'Clinical manifestations' above.)

●Therapy for symptomatic ventricular arrhythmias

•Premature ventricular beats – Patients with HCM who have symptomatic premature ventricular complexes/contractions (PVCs) are typically treated similarly to patients without HCM. The approach to therapy is discussed separately. (See "Premature ventricular complexes: Treatment and prognosis".)

•Symptomatic NSVT – In patients with HCM and symptomatic nonsustained ventricular tachycardia (NSVT), we suggest initial therapy with a beta blocker rather than other agents (Grade 2C). In patients with NSVT and symptoms despite maximal beta blocker therapy, options for therapy include sotalol and amiodarone. The details on therapy with these agents are discussed in other topics. (See "Clinical uses of sotalol" and "Nonsustained ventricular tachycardia: Clinical manifestations, evaluation, and management", section on 'Treatment of patients with structurally abnormal hearts'.)

•Sustained ventricular tachycardia – For patients with frequent ventricular arrhythmias or multiple implantable cardioverter-defibrillator (ICD) shocks, adjunctive antiarrhythmic therapy is indicated. Antiarrhythmic therapy for such patients is discussed separately. (See "Secondary prevention of sudden cardiac death in heart failure and cardiomyopathy", section on 'Other treatment options'.)

Ablation is an option for treatment of ventricular arrhythmias refractory to medical therapy. Patients with an apical aneurysm or well-defined target for ablation are more likely to have success with catheter ablation.

•Restriction of physical activity – In patients with HCM, activity restriction may be appropriate. This issue is covered in detail separately. (See "Athletes: Overview of sudden cardiac death risk and sport participation", section on 'Hypertrophic cardiomyopathy' and "Athletes with arrhythmias: Treatment and returning to athletic participation", section on 'Athletes with ICDs'.)

•Outflow tract obstruction and arrhythmias – While left ventricular outflow tract (LVOT) obstruction may increase the risk of sudden cardiac death (SCD), septal reduction therapies are only indicated for relief of symptoms related to LVOT obstruction; such therapies are not expected to decrease the risk of SCD in patients with HCM. (See 'Ouflow tract obstruction and arrhythmias' above.)

●ICD recommendations

•Secondary prevention ICD placement – Similar to other patients who survive an episode of sustained VT, ventricular fibrillation (VF), or SCA, patients with HCM who survive one of these episodes should receive an ICD. (See 'Secondary prevention' above.)

•Primary prevention ICD placement – In patients with HCM who have one or more risk factors for SCD, we suggest placement of an ICD (Grade 2C). The risk factors include (algorithm 1) (see 'Primary prevention' above):

-Family history of SCD, premature death, or sustained ventricular arrhythmia in a first-degree or second-degree relative attributable to HCM.

-LV hypertrophy ≥30 mm by echocardiography or cardiovascular magnetic resonance (CMR) imaging.

-Personal history of syncope attributable to ventricular arrythmia based on clinical history.

-LV apical aneurysm with transmural scar.

-LV ejection fraction (LVEF) <50 percent.

-NSVT on ambulatory monitoring.

-Extensive late gadolinium enhancement (LGE), typically defined as ≥15 percent of LV mass.

This approach is most appropriate for patients <60 years old. Patients with HCM ≥60 years are at low risk for disease-related adverse events, including SCD, even in the presence of the risk factors noted above.

Alternatively, the risk of SCD can be estimated with the HCM Risk-SCD risk model. In patients with HCM who have a five-year HCM Risk-SCD estimate ≥4 percent or who have an estimate <4 percent and at least one risk factor, the risk of SCD is sufficiently high to warrant ICD placement (algorithm 2). In this method, the risk factors are:

-LVEF <50 percent.

-Present of an apical aneurysm with transmural scar.

-Extensive LGE, typically defined as ≥15 percent of LV mass.

-Presence of one or more pathogenic sarcomeric mutations.

-One contributor to this topic uses an abnormal upright exercise blood pressure response and genetic variants strongly associated with premature sudden death as markers of increased risk.

With this method, the higher the five-year estimate of SCD, the more likely the patient will benefit from ICD placement.

•Choice of device and device complications – In general, the choice of ICD type (eg, transvenous, subcutaneous) and the type and rate of ICD complications in patients with HCM are similar to those in other patients. (See 'Choice of device' above and 'Complications of device therapy' above.)

ACKNOWLEDGMENT — 

The UpToDate editorial staff acknowledges Perry Elliott, MD, who contributed to an earlier version of this topic review.