American College of Cardiology

MARON AND MCKENNA et al., ACC/ESC Expert Consensus Document on Hypertrophic Cardiomyopathy
JACC 2003; 42: 000-000

American College of Cardiology/European Society of Cardiology Clinical Expert Consensus Document on Hypertrophic Cardiomyopathy

A Report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents and the European Society of Cardiology Committee for Practice Guidelines

Additional Approaches to Relieve Outflow Obstruction and Symptoms
Ventricular septal myectomy has generally been confined to selected major centers having substantial experience with this procedure. However, some patients may not have ready access to such specialized surgical care because of geographical factors; or they may not be favorable operative candidates, because of concomitant medical conditions—particularly advanced age, prior cardiac surgery, or insufficient personal motivation. Two techniques can be considered as potential alternatives to surgery for selected patients who otherwise meet the same clinical criteria as candidates for surgery.

Dual-chamber pacing. Several groups had investigated the effects of permanent dual-chamber pacing on severe outflow obstruction and refractory symptoms within observational and uncontrolled study designs (80,100,222). Data in these studies were necessarily based on the subjective perception of symptom level by patients over relatively short periods of time. Such investigations reported dual-chamber pacing to be associated with a substantial decrease in outflow gradient, as well as amelioration of symptoms in most patients. These observations inferred that a reduction of gradient with pacing in turn consistently relieved symptoms. However, other catheterization laboratory studies showed that a decrease in the outflow gradient produced by temporary A-V sequential pacing could be associated with detrimental effects on ventricular filling and cardiac output (97,223).

Subsequently, dual-chamber pacing in HCM was subjected to scrutiny in three randomized, cross-over studies (double-blind in two) in which patients received 2 to 3 months each of pacing and also back-up AAI mode (no pacing) as a control, by activating and deactivating the pacemaker accordingly (43,47,98,99). Two randomized, cross-over, double-blind studies (one multicenter and one from the Mayo Clinic) reported the effects of pacing in HCM patients to be less favorable than the observational data had suggested (43,98). For example, the average decrease in outflow gradient with pacing, while statistically significant, was nevertheless much more modest (about 25% to 40%) than reported in the uncontrolled studies and varied substantially among individual patients. In one study, the average subaortic gradient, even after nine months of pacing, remained in the preoperative range (e.g., average 48 mm Hg).

In these controlled studies, subjective symptomatic improvement assessed by quality-of-life score was reported with similar frequency by patients both after periods of pacing and after the same time period without pacing (AAI-backup) (43,98). Objective measures of exercise capacity (e.g., treadmill exercise time and maximum oxygen consumption) did not differ significantly during pacing and without pacing. These observations demonstrate that subjectively reported symptomatic benefit during pacing frequently occurs without objective evidence of improved exercise capacity and can be regarded in part as a placebo effect (43,89,98). Furthermore, no correlation has been demonstrated for gradient reduction between short-and long-term pacing, suggesting that testing the gradient response to short-term pacing in the catheterization laboratory has limited practical clinical value in judging long-term efficacy (43). However, the failure to achieve gradient reduction with temporary pacing suggests that permanent pacing is probably not indicated.

As part of its design, the randomized, cross-over, single-blind European multicenter HCM pacing trial, PIC (Pacing in Cardiomyopathy) (47,48,89,107), excluded from chronic pacing those patients without significant gradient reduction during temporary pacing. With data very similar to the other two randomized studies (but also with a large proportion of patients who elected to continue pacing based on their own subjective assessment of treatment), the PIC investigators concluded that pacemaker therapy was an option for most severely symptomatic patients with obstructive HCM refractory to drug treatment. Nevertheless, taken together the available data do not support dual-chamber pacing as a primary treatment for most severely symptomatic patients with obstructive HCM. In a nonrandomized study comparing pacing and the myectomy operation, hemodynamic and symptomatic outcome proved to be superior with surgery (99).

Although it is not a primary treatment for the disease, there is nevertheless evidence to support utilizing a trial of dual-chamber pacing in selected patient subgroups that may benefit in terms of gradient relief and improvement in symptoms and exercise tolerance. For example, there may be both subjective and objective symptomatic benefit with pacing in some patients of advanced age (over 65 years) (43), for whom alternatives to surgery are often desirable. Otherwise, there are few predictive data upon which to specifically target those patients who are most likely to potentially benefit from pacing therapy; for example, there is little relationship between the magnitude of gradient reduction with chronic pacing and the symptomatic benefit ultimately achieved. Pacing-induced LV remodeling with thinning of the wall was claimed in one uncontrolled study (80) but could not be confirmed in a randomized investigation (43). Furthermore, there is no evidence that pacing reduces the risk of SCD in HCM (43,80), alters or aborts underlying progression of the disease state, or conveys favorable hemodynamic or symptomatic benefit for patients with the nonobstructive form (224).

Of potential advantage, pacing therapy permits more aggressive drug treatment by obviating the concern for drug-induced bradycardia (82). Some patients receiving an implantable cardioverter defibrillator (ICD) for high-risk status, in which obstruction to LV outflow is also present, may benefit from use of the dual-chamber pacing component of the ICD to effect a reduction in outflow gradient. The ACC/AHA/NASPE 2002 guidelines have designated pacing for severely symptomatic and medically refractory HCM patients with LV outflow obstruction as a class IIB indication (225).

However, it should be underscored that maintenance of pacing therapy (directed toward alleviating obstruction and symptoms) may be substantially more complex in HCM than in other cardiac conditions; therefore, for optimal results this procedure should be performed in centers highly experienced in both pacemaker therapy and HCM. Producing and maintaining a reduction in gradient (and presumably in symptoms) requires that pre-excitation of the right ventricular apex and distal septum be established and that complete ventricular capture— both at rest and during exercise—without compromising ventricular filling and cardiac output. Hence, ascertaining the optimal A-V interval for dual-chamber pacing is a crucial element of pacing management in HCM. Programming of the pacemaker A-V interval to ensure complete ventricular capture may require slowing of intrinsic A-V nodal conduction with a beta-blocker or verapamil, or possibly ablation of the A-V node in selected cases (thereby rendering the patient pacemaker dependent), has been suggested. It is also understood that no other treatment modality in HCM (including surgery and alcohol septal ablation) has undergone such rigorous randomized testing in order to validate its efficacy. At present, there are no data concerning the role of biventricular pacing in HCM patients with severe heart disease.

Percutaneous alcohol septal ablation. A second alternative to surgery is the more recently developed alcohol septal ablation technique (Table 1) (44 – 46,49,79,83,86 – 88,101, 226 –234). First reported in 1995, this catheter interventional treatment involves the introduction of absolute alcohol into a target septal perforator branch of the left anterior descending coronary artery for the purpose of producing a myocardial infarction within the proximal ventricular septum. Septal ablation mimics the hemodynamic consequences of myectomy by reducing the basal septal thickness and excursion (producing akinetic or hypokinetic septal motion), enlarging the LV outflow tract and, thereby, lessening the SAM of the mitral valve and mitral regurgitation (44 – 46,49,88,227).

This technique utilizes conventional methods and technology currently available for atherosclerotic CAD. After standard coronary arteriograms are performed, a coronary balloon is placed into a proximal major septal perforator artery with the aid of flexible coronary guide wires. A temporary pacing catheter is positioned in the right ventricular apex in the event that high-grade A-V block occurs. After the balloon is inflated, an arteriogram is performed through the lumen to verify that the balloon is located in the desired anatomic position and to ensure that leakage of alcohol into the left anterior descending coronary artery or coronary venous system does not occur.

Myocardial contrast echocardiography guidance (with injection of echo contrast or radio-opaque medium) is important in selecting the appropriate septal perforator branch. This technique is useful for determining the precise area of septum targeted for alcohol and infarction and whether the selected septal perforator also perfuses other distant and unwanted areas of LV or right ventricular myocardium or papillary muscles (79,230). Some groups prefer a pressure-angiographic and fluoroscopy-guided technique (226,227,233). The targeted septal perforator and area for infarction are identified by an immediate fall in outflow gradient following balloon occlusion and/or contrast injection.

The amount of ethanol to be injected is estimated by the angiographic visualization of septal anatomy and whether contrast wash-out is slow or rapid (46,79,86,226,233). Usually, about 1 to 3 cc (average 1.5 to 2 cc) of desiccated ethanol (of at least 95% concentration) is slowly infused into the septal perforator and septum via the balloon catheter, inducing a myocardial infarction demonstrable by 400 to 2,500 units of creatinine phosphokinase release, equivalent to an area of necrosis estimated to be 3% to 10% of the LV mass (20% of the septum). However, centers performing a large number of alcohol septal ablation procedures today are using smaller amounts of ethanol, leading to less creatinine phosphokinase release and smaller septal infarcts, and also reducing the incidence of complete heart block (44,233).

Successful alcohol septal ablation may trigger a rapid reduction in resting outflow gradient evident in the catheterization laboratory. More frequently, a progressive decrease in the gradient occurs after 6 to 12 months, usually achieving levels in a range equivalent to that with myectomy, and resulting from remodeling of the septum without significant impairment in global LV ejection (46,49,229 – 234). This has been reported for patients with large resting gradients at baseline as well as those with outflow obstruction present only under provocable conditions (234). Often a biphasic response of the gradient is observed with alcohol septal ablation in which an acute response with striking reduction (probably due to stunning of the myocardium) is followed by a rise to about 50% of its pre-procedure level the next day, but within several months may reach greatly reduced levels. Results of myectomy and alcohol ablation compared at two institutions showed similar gradient reductions with the two techniques (96). Another comparative analysis from a single institution showed both surgery and ablation to substantially reduce resting and provocable gradients, but to a significantly greater degree with surgery (101).

A number of other favorable structural and functional effects following ablation have been reported (231), representing the expected consequences of reduced outflow gradient, normalization of LV pressures, and reduced systolic overload. Echocardiographic analyses from two groups have reported ablation to be associated with widespread regression of LVH beyond the alcohol target area (229,233), but the extent to which remodeling occurs with time secondary to this procedure is unpredictable and not fully understood. Also, there is concern that extensive wall thinning could lead to arrhythmogenic susceptibility or even the end-stage phase.

A large proportion of ablation patients from several centers have been reported to demonstrate subjective improvement in limiting symptoms and in quality of life in observational studies over relatively short-term follow-up periods of 2 to 5 years. As with surgery, the decrease in symptoms associated with ablation is often dramatic (44,46,49,88,101,108,232,234). In addition, improved exercise performance has been shown objectively in terms of total treadmill exercise time and peak oxygen consumption in some studies (46,96,101,232). However, alcohol septal ablation has yet to be subjected to the scrutiny of randomized or controlled studies or long-term follow-up. A recent study found that both septal myectomy and ablation led to improved exercise testing parameters, but surgery was superior in this regard (232).

The mortality and morbidity associated with alcohol ablation in experienced centers have proved to be relatively low, although they are similar in surgical myectomy. Procedure-related mortality has been reported to be from 1% to 4% but is probably reduced in the more recent cases. Reports of permanent pacemaker implantation for induced high-grade A-V block have ranged from 5% to as high as 30% (46,88,101,228), but this complication appears to be decreasing substantially with the use of smaller amounts of alcohol. In contrast to septal myectomy, which usually produces left bundle branch block, alcohol ablation commonly results in right bundle branch block (45,46). It is also possible for coronary artery dissection to occur, as well as backward extravasation of alcohol, producing occlusion or abrupt coronary no-flow (87) and a large anteroseptal myocardial infarction.

Proper selection of patients for alcohol septal ablation remains a crucial issue (228). Similar to patients recommended for septal myectomy (5,7,8,11,13,41), all candidates for alcohol septal ablation should have severe heart failure symptoms (NYHA classes III or IV) refractory to all medications utilized in HCM as well as a subaortic gradient of 50 mm Hg or more measured with Doppler echocardiography either under basal conditions and/or with physiologic provocative maneuvers during exercise (228). Caution should be exercised so that in patients selected for alcohol septal ablation, outflow gradients are documented to be due to SAM and proximal mitral valve-septal contact (119), exclusive of congenital abnormalities of the mitral apparatus such as anomalous papillary muscle insertion into mitral valve, which produces more distal muscular obstruction in the mid-cavity (91,115).

Nevertheless, the number of alcohol ablations performed world-wide now approaches an estimated 3,000 over only about a six-year period, exceeding the number of surgical myectomies performed over the 40 years since this operation was introduced (228). In some instances, the frequency with which myectomy surgery has been performed for obstructive HCM has now been reduced by more than 90% (103,228) due to the recent accelerated enthusiasm for ablation.

Disproportionality in the frequency with which alcohol septal ablation is performed relative to myectomy (ablations are estimated to be at least 15 to 20 times more common than surgery at present) has raised concerns that there may have been an insidious and unjustifiable lowering of the symptom and gradient-level threshold in the selection of patients for ablation, with less symptomatic (in NYHA class II), less obstructed, and younger patients now undergoing the procedure (228). This circumstance has evolved in part because of the relative ease with which ablation can be performed (compared to surgery), with substantially less discomfort during a much shorter postoperative hospitalization and recovery period in the absence of a sternotomy. However, this fact does not justify less strict criteria for alcohol septal ablation.

Another factor that has affected patient selection for alcohol septal ablation is the practice of determining eligibility based solely on a subaortic gradient provoked by non-physiologic interventions such as dobutamine infusion (rather than exercise, for example) (88,230). Dobutamine is an inotropic and catecholamine-inducing drug that is a powerful stimulant of subaortic gradients in normal hearts or in cardiac diseases other than HCM (130,131,235) of questionable physiologic and clinical significance (235), and occasionally results in adverse consequences to patients with obstruction; dependence on dobutamine to induce gradients can expose some patients to septal ablation in the absence of true impedance to LV outflow. Therefore, dobutamine is generally not recommended for the purpose of provoking outflow gradients in severely symptomatic HCM patients who are regarded as possible candidates for major interventions.

A predominate concern raised with respect to alcohol septal ablation is the potential long-term risk for arrhythmia-related cardiac events (including SCD) directly attributable to the procedure. Unlike myectomy, alcohol septal ablation potentially creates a permanent arrhythmogenic substrate in the form of a healed intramyocardial septal scar that could increase the risk of lethal re-entrant arrhythmias (226). This is particularly relevant because many patients with HCM already possess an unstable electrophysiologic substrate as part of their underlying disease (2,208,236,237). However, since HCM patients are at increased risk for SCD over particularly long periods, possibly through much of their lifetimes, it will require many years (and probably decades) to determine the likelihood that risk for arrhythmia-related events and SCD is increased as a consequence of the healed intramyocardial scar produced by alcohol septal ablation. Indeed, this is particularly relevant for young patients in whom even a modest annual increase in the risk of SCD would have the likelihood of shortening life considerably. Reports of the non-inducibility of reentrant ventricular tachyarrhythmia in small numbers of patients in the short term after septal ablation (46) do not appear sufficient at this juncture to exclude the possibility of late-onset ventricular tachyarrhythmias and SCD over the long risk period characteristic of HCM (26,208).

Therefore, at present, the impact of alcohol ablation on the incidence of SCD is unresolved. Until more is known regarding the natural history of patients undergoing alcohol septal ablation and there is less uncertainty regarding the consequences of the intramyocardial scar, particularly careful selection of patients seems advisable and prudent (by largely confining the procedure to older adults), particularly when the option for surgical myectomy is feasible. There would not appear to be a primary role for alcohol ablation in children, and such procedures are not advised.

Due to morphologic heterogeneity, not all HCM patients with obstruction are ideal candidates for septal ablation. This therapy relies on the fixed anatomic distribution and size of the septal perforator coronary arteries. Therefore, the ablation technique cannot make adjustments for variability in the distribution and size of these arterial vessels in relation to the distribution of septal hypertrophy, or for other complexities of LV outflow tract morphology such as greatly elongated mitral leaflets and anomalous papillary muscle. The direct operative approach provides greater flexibility for relieving obstruction and also allows surgical treatment for associated cardiac abnormalities such as primary valvular disease (e.g., myxomatous mitral valve prolapse or aortic stenosis) (124), atherosclerotic CAD, or segmental myocardial bridging of the left anterior descending coronary artery (238), as well as anomalies of the mitral valve and apparatus. Also, relief of obstruction with surgery is immediate (but is often delayed with alcohol septal ablation), which may be crucial in some patients with particularly severe symptoms of heart failure.

The “learning curve” for expertise with the alcohol septal ablation technique is steep (due, in part, to the relatively small number of eligible HCM patients), particularly regarding selection of the optimal septal perforator branch; therefore, ablation should not be regarded as a routine technique to be employed by any expert interventional cardiologist. It is advisable that alcohol ablation (as well as myectomy) be largely confined to centers having substantial and specific experience with HCM and the procedure in order to assure proper patient selection, the lowest possible rates of morbidity and mortality, and the greatest likelihood of achieving benefits.

While alcohol ablation represents an option available to HCM patients and a selective alternative to surgery, it is not at this time regarded as the standard and primary therapeutic strategy for all severely symptomatic patients refractory to maximal medical management with marked obstruction to LV outflow (Table 1). Septal myectomy remains the gold standard for this HCM patient subset (7,11,14,41,232).