The management of patients with hypertrophic cardiomyopathy (HCM) requires expertise in heart failure, cardiac imaging, electrophysiology, genetic testing and counselling, and in techniques that reduce left ventricular outflow tract obstruction (LVOTO). Following the initial description of alcohol septal ablation (ASA) in 1995,1 it has gained rapid acceptance and is now the most commonly chosen treatment option for LVOTO reduction. More ASA procedures than operations have been performed, despite the passage of more than 40 years since surgical LV septal myectomy (LVM) was developed.2,3 The less invasive ASA has evolved so that its procedural mortality and efficacy are similar to those of LVM.4
The debate concerning the relative merits of ASA and LVM is not informed by any randomised studies; the only randomised trials addressing drug-refractory symptomatic LVOTO treatment options examine dual chamber (DDD) pacing. DDD, ASA and LVMM have all been subject to the contrary opinions of enthusiasts and detractors. A fourth option, mitral valve replacement, is considered only in the presence of other indications for mitral valve replacement.5 In this article, we discuss indications for invasive LVOTO reduction with an emphasis on ASA. We first describe the genetic and phenotypical heterogeneity of HCM to emphasise that HCM is a complex syndrome in which mechanisms are responsible for symptoms. A patient with successful LVOTO reduction still has HCM and is subject to symptom limitation, disease progression and sudden death (SD) risk.
Genetic and Phenotypical Spectrum
HCM has an autosomal dominant inheritance with incomplete penetrance.6 Familial LV hypertrophy (LVH) can also follow other patterns of heredity, including X-linked, recessive and mitochondrial, and can comprise a cardiac component of multisystem disorders. Family history, cardiac imaging and other clinical investigations cannot reliably distinguish between ‘classic’ HCM and other causes of LVH. Hundreds of disease-causing mutations in more than 10 genes have now been described. In just over half of HCM patients, sequencing will detect a mutation in a gene encoding a component of the sarcomere, the basic contractile unit of myocytes.7 These include beta-myosin, cardiac myosin-binding protein C, cardiac actin, alpha-tropomyosin, troponins T and I and myosin’s regulatory light chains. The remaining percentage, in whom no mutation is detected, is likely to be a heterogeneous group in which non-sarcomeric mutations result in LVH, and in which HCM ‘phenocopies’ result from hypertension, athleticism, muscle storage disease and myocardial infiltration.
Family studies show considerable variability in the magnitude and extent of hypertrophy and other aspects of the disease. In some individuals the mutation is associated with severe LVH, while other gene-positive family members have no detectable cardiac abnormalities. The presence of LVOTO and other disease features also varies extensively between individuals. The unidentified factors responsible for modifying the effects of a disease-causing mutation have powerful influences. Genotype–phenotype studies indicate that some mutations may influence the distribution of LVH, but contributions made by any mutation to the development of LVOTO remain obscure.
Currently, the clinical value of mutation detection is largely limited to screening of at-risk family members.8 A handful of mutations associated with a particularly malignant prognosis prompt closer clinical surveillance and reduce thresholds for implantable cardiac defibrillator (ICD) therapy.
Outflow Obstruction, Mid-cavity Obstruction and Mitral Regurgitation
Invasive reduction of LVOTO is indicated if symptoms in patients with LVOTO resulting from systolic anterior motion of the mitral valve (SAM) persist despite maximum pharmacological treatment (see Figure 1). Mitral–septal contact in-systole can result in a dynamic outflow tract obstruction, and SAM also interrupts leaflet coaptation, producing a variable degree of mitral regurgitation (MR). The premise of all LVOTO therapies is that outflow obstruction and MR contribute to symptoms. The relatively poor correlation between both LVOTO magnitude and MR and symptom status probably reflects the additional and variable importance of myocardial ischaemia, impaired LV filling, small stroke volumes and arrhythmias. The determinants of SAM continue to be debated, and investigators argue for either a ‘pull’ or a ‘push’ force on the mitral apparatus. In the former, the anterior leaflet is pulled towards the LVOTO as the result of the Venturi effect produced as systolic flow accelerates across the convexity of the ventricular surface of the anterior mitral leaflet. A negative pressure pulls the anterior leaflet away from the posterior leaflet and towards the septum. Alternatively, as the result of distorted LV geometry, the systolic effort pushes the mitral leaflets towards the LVOT.9 These considerations are not just academic, as modified surgical approaches allow more definitive reconfiguration of LV anatomy.
Treatments for Symptomatic Left Ventricular Outflow Tract Obstruction
Surgical Myectomy
In the 1960s, Morrow developed a surgical approach to resecting the sub-aortic ventricular septum through the retracted aortic valve at aortotomy.3 The Morrow procedure or LVM is still considered the ‘gold standard’ treatment by several influential commentators.10 Modern surgical results are impressive, with low operative mortality, substantial symptom improvement, increased exercise tolerance and, perhaps, an improved prognosis.10-13 While few doubt the potential benefits of LVM, its designation as the gold standard is supported by its earlier development, but not by randomised trial data or contemporary real-world practice. Additionally, published results are from the few centres with considerable experience, where complication rates are low. Complications include aortic incompetence, heart block, ventricular septal defects and all the complications of open-heart surgery. Surgical techniques are invaluable when more complex anatomical problems are to be addressed.14
Right Ventricle Apical Pre-excitation
DDD with right ventricular pre-excitation reduces LVOTO acutely and chronically.15,16 This effect is likely to be the result of (paradoxical) septal movement away from the direction of SAM, but reductions in cardiac output may also contribute. Apical right ventricle (RV) lead placement and carefully chosen paced atrioventricular (AV) intervals are important for successful therapy. Uniquely in HCM research, pacing has been examined in prospective placebo-controlled randomised trials. The cross-over Multicenter Pacing Therapy (M-PATHY) study compared three months of active pacing with placebo pacing.17 No benefits of active pacing were demonstrated over an apparently large placebo effect. The haemodynamic and symptomatic benefits of pacing are reported to be progressive throughout the first six months following implantation and persist after cessation of pacing.18
Potential bias from cross-over treatment effects and type 2 error due to the abbreviated duration of therapy may have contributed to a failure to demonstrate treatment benefit and the large placebo effects. Significant haemodynamic and symptomatic improvement after DDD pacing were demonstrated in the larger randomised double-blinded Pacing In Cardiomyopathy (PIC) study.19
Potential adverse effects on LV systolic and diastolic function of chronic RV pacing,20 equivocal trial data and the emergence of ASA have resulted in the relegation of pacing to a poor third choice. Nonetheless, a pacemaker has low complication rates and the rate of implantable cardioverter–defibrillator (ICD) implantation in HCM patients is increasing, most commonly using dual-chamber systems. A therapeutic trial of pacing therapy in ICD patients with symptomatic LVOTO may prove reasonable before the use of ASA or LVM. Furthermore, pacing ICD therapy will increase the safety of both ASA and LVM.
Non-surgical Septal Reduction
ASA’s rapid acceptance has undoubtedly been influenced by the convictions of proponents for and a greater willingness of patients to submit to less invasive procedures. In the non-randomised and uncontrolled studies now available, ASA substantially reduces LVOT gradient, improves symptoms, increases exercise tolerance and leads to desirable changes in cardiac anatomy and physiology. These include reduced LV filling pressure, abolition or attenuation of mitral regurgitation (MR), smaller LA size and reduction of LVH.4 Early concerns regarding arrhythmogenicity of the ‘infarct’ scar are not supported by data, but the potential for long-term complications cannot be dismissed. However, it should be noted that, despite ASA’s comparative youth, the number of patient treatment years must now exceed that of LVM.
Additional Indications for Invasive Therapy
Symptomatic patients with resting LVOTO resulting from SAM are candidates for LVOTO reduction. Patients with evidence of obstruction who do not fulfill these criteria are considered here.
Provoked Left Ventricular Outflow Tract Obstruction
LVOTO worsens if end diastolic LV volumes are reduced (postural change, dehydration, vasodilatation, post-prandially, etc.) or if inotropy is increased (after ventricular extrasystoles, exertion). A patient without LVOTO during echocardiography may develop an obstruction when rising from the exam couch or during exertion. A systolic murmur apparent on standing from a squat is strongly suggestive of provoked LVOTO, as are symptoms of postural pre-syncope and an intolerance of extrasystoles.
Labile LVOTO may be missed if echocardiographic or invasive studies do not include techniques to provoke SAM. Most studies (LVM, DDD and ASA) include patients with provocable LVOTO, and symptomatic benefits are demonstrated. The Valsalva manoeuvre and exercise can be used to demonstrate provocable LVOTO, although the Valsalva may underestimate the proportion of patients with provoked LVOTO.21 Patients may have difficulty understanding the Valsalva and an inflated lung impairs echo imaging: we ask patients to attempt to expel the stopper of a 20cc syringe placed in their mouth after taking a small breath. Imaging during exercise has obvious difficulties, but we have obtained satisfactory apical images on a supine bicycle. Pharmacological agents include amyl nitrate, sublingual glycerin trinitrate and intravenous (IV) isoproterenol. IV dobutamine is not recommended as it can provoke LVOTO in normal hearts.22 Following ventricular extra-systoles, the Brokenborough effect can demonstrate provocable LVOTO in the catheter laboratory, either through the irritant effects of manipulating a catheter in the ventricle or by paced extrasystoles. The latter technique is particularly useful during ASA, when temporary pacing can introduce identically timed extrasystoles every few cardiac cycles.
Mitral Valve Abnormalities and Mid-cavity Obstruction
Benign structural abnormalities of the mitral valve are common in HCM.23 More important abnormalities include anomalous insertions of the papillary muscles/chordae onto the mitral leaflets or septum, and displacement of the papillary muscles.23,24 Surgery is strongly favoured if significant structural mitral abnormalities co-exist with SAM, but the MR of SAM should improve following successful ASA. Pronounced papillary muscle hypertrophy may result in the uncommon and poorly understood mid-cavity (mid-ventricular) variant.25 Mid-cavity obstruction can co-exist with LVOTO, or can become apparent after successful LVOTO treatment.26 Potential adverse consequences include reduced contribution of the distal LV segment reduction to cardiac output, effects of elevated afterload on the distal LV segment (including metabolic consequences, reduction of myocardial perfusion pressures, aneurysm formation) and any (poorly defined) consequences of muscle contracting onto muscle. Monomorphic ventricular tachycardia (VT) is uncommon in HCM, but when present is frequently associated with a distal LV aneurysm resulting from mid-cavity obstruction. The detection of monomorphic VT should prompt a search for apical abnormalities that may not be apparent on standard echo views. Several groups report that pacing,27 ASA28 and surgery25 can reduce mid-cavity gradients. However, although mid-cavity disease is often associated with drug-refractory symptoms, it is not clear why symptoms develop or if gradient reduction is effective.
Asymptomatic Left Ventricular Outflow Tract Obstruction
Structural valve disease is associated with increased risk of heart failure and sudden death in asymptomatic patients. Are LVOTO and MR independently associated with a worse prognosis? Retrospective studies report that LVOTO is independently associated with a greater risk of death.29–31 These reports include important differences concerning mode of death (heart failure or arrhythmic), the importance of LVOTO as a binary or continuous risk factor and the significance of symptom status. Maron et al. analysed records from 1,101 patients followed for a mean of 6.3 years.29 Patients with LVOTO (gradient >30mmHg at rest) had a relative risk (RR) of death of 1.6 (2.7 for symptom progression) compared with those without LVOTO. Death was most often related to heart failure or stroke, but risks did not increase with increments of LVOTO above 30mmHg. In half as many patients followed for 4.5 years, Autore et al. reported that LVOTO predicted cardiovascular death in New York Heart Association (NYHA) class I or II patients (RR 2.4), but was less helpful in symptomatic cases where functional class III or IV was a powerful predictor (RR 7.9).31 After a median follow-up of 5.1 years, Elliott et al. also reported reduced survival in patients with LVOTO.30 They detect an incremental risk, with an RR for all-cause death and transplantation of 1.24 per 20mmHg of LVOTO. A very low SD risk in the least symptomatic patients leads the authors to conclude that septal reduction therapy in the asymptomatic patient is not indicated for the prevention of SD. At present, the desire to relieve symptoms is the only indication for invasive LVOTO reduction. Long-term consequences on cardiac structure and function of chronic LVOTO and MR may include progressive atrial enlargement and atrial fibrillation, pulmonary hypertension and LV systolic impairment. Furthermore, LVOTO reduction leads to modest LVH regression remote from the ablated myocardium, suggesting that some of the LVH develops in response to afterload, in addition to the ‘primary’ hypertrophy.32 Future approaches to LVOTO may include prognostic indications, and treatment protocols may resemble those for the management of asymptomatic and symptomatic structural valve disease.
Alcohol Septal Ablation
Patient Selection
Appropriate patient selection maximises the likelihood of benefit from LVOTO reduction. A convention is to regard a resting LVOT gradient of >30mmHg or a provoked gradient of >60mmHg as being evidence for LVOTO severity of sufficient magnitude to warrant consideration of invasive LVOTO reduction.
While these thresholds do not account for MR and have no experimental basis, they emphasise an appreciation that symptoms are multifactorial and the abolition of mild obstruction is unlikely to be of substantial benefit. Risks of a ventricular septal defect (complicating either ASA or LVM) mean that relatively mild septal thickness (<15mm) is a relative contraindication. Mitral valve abnormalities, co-existing mid-cavity obstruction, unsuitable septal perforator anatomy and significant coronary disease favour LVM.
Alcohol Septal Ablation Procedure
The technique aims to selectively ablate basal septal myocardium adjacent to the location of SAM–septal contact by injection of absolute alcohol into an isolated septal perforator artery. Heart block during or following the procedure is common, and a temporary pacing wire is routinely inserted. If monitoring of the LVOT gradient at the time of ablation is desired, catheters are introduced into the LV and left main coronary. For uncomplicated procedures, a single femoral arterial sheath may suffice. A guiding catheter is introduced into the left main coronary artery, and a 0.014 guidewire advanced into a proximal septal branch of the left anterior descending artery. The first septal artery is usually selected (see Figure 2a), and a short (±10mm) ‘over-the-wire’ (OTW) balloon introduced into the branch. The balloon is positioned under flouroscopy to ensure it is entirely within the septal branch and does not encroach on the lumen of the LAD (see Figure 2b).
The OTW balloon is inflated and the guidewire removed. A small volume of angiographic contrast is injected through the lumen of the OTW balloon to ensure that the inflated balloon effectively isolates the (distal) septal artery from the LAD (see Figure 2c). Echocardiographic contrast is then injected through the OTW balloon to identify the myocardial territory perfused by the septal branch. Transthoracic echocardiography is usually sufficient (transoesophageal may be necessary) to demonstrate that the contrast agent reaches only the target myocardium. The target lies adjacent to the point of mitral–septal contact and is best seen from the apical 4-chamber view. The right ventricular free wall, LV apex and papillary muscles may on occasion be supplied by proximal septal vessels, and several echo views should be used to ensure contrast enhancement is confined to the target area. Several injections of contrast may be needed, and ablation should not proceed if the desired target area is not identified or if other areas are. An alternative septal artery may be investigated. Ablation may proceed if echocardiographic localisation is supportive. Intravenous analgesia is administered and 1–2ml of absolute alcohol is administered slowly through the lumen of the OTW balloon, with the balloon remaining inflated for a further five minutes.
The echocardiographic gradient is usually markedly reduced, although this acute effect is likely to reflect myocardial stunning and may not accurately reflect the long-term outcome. The LVOT gradient will commonly increase in the days following the procedure, then fall again over a period of weeks as septal remodelling occurs.33
Alcohol Septal Ablation Complications
Serious complications were uncommon in a systematic review.4 Early (30-day) mortality was 1.5%, with a late mortality of 0.5%. Ventricular fibrillation in the peri-operative period occurred in 2.2%, LAD dissection in 1.8% and pericardial effusion in 0.6%. As the review considers trials early in the evolution of the technique and studies that included patients too frail for surgery, these data may reflect an overestimation of risk and an underestimation of benefit. Minor damage to the conducting system is common with right bundle branch block and first-degree AV block each developing in about half of the patients.4 Temporary wires should be left in place for at least 24 hours in uncomplicated patients and beta-blockers are prescribed to test AV conduction prior to removal of the temporary pacing wire in some centres. Complete heart block can occur up to several days following ASA and inpatient monitoring for five to seven days is recommended. Permanent complete heart block requiring device implantation is required in as many as 30% of patients in some series and in 10% of the patients included in the systemic review.4 The need for permanent pacing is more likely in female patients if more than one septal artery is treated, with more rapid injections of alcohol and when there is pre-existing conducting system disease (especially left bundle branch block, LBBB).34
Alcohol Septal Ablation Failure
Failure can be considered in terms of either insufficient LVOTO reduction or the persistence of symptoms despite LVOTO abolition. In the former, additional intervention (of any modality) might be considered. In the systematic analysis, ASA was repeated in 6.6% of patients and 1.9% went on to septal myectomy.4 The persistence of significant symptoms despite LVOTO abolition (by ASA, DDD or LVM) is possibly important in about 10–20% of patients, attests to the view that symptoms are multifactorial and emphasises the importance of appropriate patient selection.
Concluding Remarks and Future Directions
The therapeutic options available for LVOTO reduction each offer unique advantages and disadvantages. ASA has a principal role in the contemporary management of drug-refractory HCM patients with LVOTO. ASA has become the treatment most often chosen for these patients, even though many still consider surgical approaches ‘the treatment of choice’. While ASA is undoubtedly efficacious, relatively safe and minimally invasive, randomised trial data comparing its use with that of LVM are not available. Generating this evidence should be a principal concern of all physicians involved in the management of HCM patients. Research will also determine the value of LVOTO reduction in asymptomatic patients and in improving prognosis. Pacing may prove a pragmatic initial therapy in patients with additional indications for ICD implantation. For the present, patients must be made aware that the choice between ASA and surgery is made on the basis of considerations other than proven comparative outcomes. Surgery has a clear advantage in patients with complex LV anatomy, and ASA is preferential from the perspective of patient discomfort and desire for minimal invasiveness.