|Home | About | Journals | Submit | Contact Us | Français|
In hypertrophic obstructive cardiomyopathy, selective and asymmetric hypertrophy results in a stenotic subaortic channel, which is further narrowed by a Venturi effect (suctioning of the anterior leaflet, manifested by systolic anterior motion of the mitral valve). Better understanding of these essential pathophysiologic mechanisms has led to the definition of a new anatomo-functional entity, the 1st septal unit, which consists of the basal interventricular septal hypertrophy and its related septal arterial branches. As an alternative to surgical myomectomy, alcohol septal ablation is an effective method of reducing subaortic stenosis and improving mitral valve function. After alcohol ablation, global negative remodeling of the hypertrophied left ventricle eventually ensues.
This review presents specific anatomic and functional features of a newly identified pathophysiologic entity (the 1st septal unit) in relation to the clinical manifestations and natural history of hypertrophic obstructive cardiomyopathy. This relationship is also relevant during the performance of alcohol septal ablation interventions: related operative suggestions are provided for optimizing subaortic stenosis relief during septal ablation and for preventing complications.
Since 1994, when Ulrich Sigwart1 introduced alcohol septal ablation (ASA) into clinical practice, this method has become an important option for treating symptomatic hypertrophic obstructive cardiomyopathy (HOCM).2–14 Alcohol injection into a septal branch had previously been used by Brugada and coworkers15 for ablating ventricular tachycardia. In accordance with the natural history of HOCM,16,17 the main indications for ASA are substantial functional impairment, as characterized by New York Heart Association functional class III or IV status (dyspnea, angina, or syncopal equivalents), and a subaortic gradient of more than 50 mmHg.2–14 In selected patients with these criteria, ASA results in a level of clinical improvement similar to that seen after a surgical septal myomectomy18–27; moreover, ASA effectively relieves subaortic stenosis and mitral insufficiency.
The medical literature commonly states that the largest proximal, upper (also called “the 1st”) septal branch should be the target of ASA,1–14 that 5% to 40% of candidates present technical difficulties (a “nonsuitable” 1st septal branch),1–14,28,29 and that 5% to 10% of patients require a 2nd ASA procedure to obtain adequate relief.1–14 On the basis of our recent experience with ASA, we present some observations that posit the existence of a newly recognized (as defined here) anatomo-functional entity that seems to determine the severity and prognosis of HOCM: the 1st septal unit (FSU), which consists of the 1st septal coronary branch and its dependent (asymmetric) septal hypertrophy, at the level of mitral-septal contact (Fig. 1). We also discuss procedural considerations that may further the understanding of HOCM and improve the results of ASA.
It has long been debated whether asymmetric septal hypertrophy is a critical component or an irrelevant epiphenomenon of HOCM.30–38 In this complex condition, asymmetric septal hypertrophy has been observed on echocardiography to be accompanied by systolic anterior displacement of the mitral valve apparatus (systolic anterior motion or SAM), which constitutes the posterior portion of the obstructed left ventricular outflow tract.30–38 The high systolic flow velocity created at the subaortic level by SAM further worsens the stenosis, because the anterior leaflet of the mitral valve is suctioned into the outflow tract by a Venturi effect.38,39 For many years, the induction of negative remodeling of the upper (basal) septum was surgically pursued as a way to relieve subaortic obstruction. Diffuse left ventricular negative remodeling (reabsorption of hypertrophic myocardium) and even improvement of secondary mitral insufficiency were noted as consequences of the intervention.1–14,18–27,40 Using alcohol septal ablation, we aim at producing similar results by the subselective administration of pure or absolute alcohol into the FSU-related septal branch(es) via a catheter procedure. Pure alcohol causes immediate coagulative necrosis of the targeted territory, as discussed below. The results of both surgical septal myomectomy and ASA suggest that the FSU is indeed a critical factor that establishes both the severity and the prognosis of HOCM in individual patients16,17,36; moreover, substantial removal (negative remodeling) of the FSU, by any means, leads to dramatic limitations in the clinical repercussions and the progression of HOCM, which confirms the central role of the FSU (Fig. 2).
In human beings, the ventricular septum is normally sup-plied with blood by anterior and posterior septal pene-trating coronary branches. The left anterior descending artery (LAD) supplies approximately two thirds of the septum (the anterior portion), and the posterior descend-ing artery (which arises from the right coronary or circumflex artery, depending on the dominance pattern) provides branches that supply the posterior portion of the septum.41–45 Neighboring anterior and posterior septal branches are richly interconnected by a network of (usually small) collateral vessels, as shown in detailed classic anatomic studies41 (Fig. 3). After occlusion of the LAD or posterior descending branch, collateral vessels rapidly become dilated. Typically, only in this context are they angiographically visible in human beings.
Selective obliteration of one or more septal branches by means of temporary balloon inflation, coils, or covered stents46–48 has occasionally been performed to induce septal ischemia, hibernation, or necrosis. These approaches produce much smaller infarctions and less negative remodeling than does ASA.46–48 For example, in 1 case reported by Fifer and coworkers,47 implantation of a covered stent led to initial resolution of a large subaortic gradient associated with a modestly elevated creatine kinase level (peak, 363 U/L). After 10 months, however, the gradient and the patient's symptoms recurred, when collateral circulation developed from the right coronary artery to the 1st septal branch. The authors hypothesized that the initial hibernation state was reversed with the onset of collateral blood flow.47
During ASA, the operator not only can observe the detailed anatomy of the septal branches but also can perform both subselective septal contrast angiography (with the same balloon catheter) and contrast echocardiography.14,40,49 Nevertheless, precision is suboptimal with both of these techniques, because they indicate only the core of the eventual alcohol accumulation (alcohol diffuses through the tissues much more extensively and rapidly than do high-viscosity contrast agents). In addition, the targeted septal branch becomes occluded by clot shortly after alcohol administration begins, which changes the pattern of fluid progression. In the presence of visible collateral vessels, especially in the case of critical obstruction or occlusion of the proximal right coronary artery, alcohol injected at high pressures or in large volumes into an anterior septal vessel inevitably spills over into the posterior circulation and causes remote damage, which has occasionally been reported in the literature.1–14,40
Until advanced experience was gained with ASA, the concept of the 1st septal branch in the FSU was poorly defined in clinical practice and in the literature. Over time, it has become evident that, in treating HOCM, the operator must identify the septal branch(es) that supply the basal third of the interventricular septum, at the site of mitral-septal contact (kissing) or SAM of the mitral valve. Indeed, frequently, the anatomic 1st penetrating branch of the LAD (which technically can be negotiated by a balloon catheter) is not related to the FSU or is not the only or most favorable vessel related to it.11,14,40 The anatomic 1st septal branch of the LAD can be a small one (<1 mm in diameter) that supplies only a trivial upper portion of the septum (and possibly the bundle of His or the proximal left bundle branch) (Fig. 4) but not the portion related to SAM in the FSU. In other instances, the 1st septal branch originates from the diagonal or circumflex arteries or even from the left main trunk, right anterior sinus, or right coronary artery (Fig. 5).45 Sometimes, the 1st septal branch is a large vessel that continues subepicardially, like a duplicate LAD, and provides most of the septal branches. Frequently, close to its origin, the 1st septal branch splits into 2 vessels: 1 that typically extends to the right side of the septum (and the moderator band) and another that supplies the left side of the upper septum at the level of SAM (Fig. 6). Incidentally, the moderator band (the only structure to connect the upper ventricular septum with the free wall of the right ventricle) usually carries the right bundle branch.50 The consistent origination of the moderator-band coronary branch from the 1st septal branch is a likely reason for the almost inevitable appearance of some degree of right bundle-branch block after successful ASA.51–53 Occasionally, right ventricular branches (not septal branches) originate from the mid portion of the LAD (Fig. 7), and they should be avoided during ASA.
In a minority of cases (10% to 30%), successful ASA necessitates selective administration of alcohol into more than 1 septal branch, because the anatomic 1st septal branch is relatively small; to achieve adequate septal remodeling, a 2nd branch must be treated during the initial procedure or a later one.1–14
Like any other septal branch, the 1st septal has a rich anastomotic network41 that extends distally along the septum, toward the apex (as shown in Fig. 3). A relatively high pressure, speed, or volume of injection of alcohol during ASA can result in distal leakage into the neighboring septal branches and even into the epicardial LAD (as demonstrated by the occurrence of infarcts in areas remote from the target, in the lateral wall of the left ventricle).11,14,40
In cases of HOCM, the 1st septal branch frequently shows angiographic systolic blanching or a milking effect, which is also an angiographic expression of myocardial disarray and high (suprasystemic) intramural pressures. Because of proximal balloon inflation during ASA, the infused alcohol must be subjected to a high systolic pressure in the presence of distal vessel occlusion by blood clot. This situation may encourage the progression of the injected, low-density fluid toward anastomotic connections. At the same time, this peculiar hemodynamic regimen can promote the favorable effect of alcohol filtration through the neighboring tissues, if the injection is at a slow rate. Alternatively, if the balloon does not properly seal the instrumented septal branch, backward flow into the epicardial LAD can occur, posing an inherent risk of remote-site myocardial alcoholization. This potential is confirmed by Doppler flow-velocity studies, which have shown that retrograde systolic flow frequently occurs in epicardial vessels in cases of HOCM.54 Soon after alcoholization starts, however, local akinesia may mitigate intramural systolic suprasystemic hypertension.
In the literature and in practice, there has been much discussion concerning the ideal volume and flow rate of alcohol infusion.1–14,40 The initial use of a 2- to 5-cc infusion over a period of about 30 seconds1,3 has recently been challenged.11,14,40 The total quantity and the speed of alcohol injection appear to be fundamental determinants of adequate results.11,40 The operator tries to prevent atrioventricular (AV) block and overspills from the target territory—the 2 major complications of excessive alcoholization. The dependent, targeted territory of the functional 1st septal unit is actually predetermined by subselective angiography and contrast echocardiography. The infusion rate should be fairly slow (about 1 cc over 30 to 60 seconds) and continuous; intermittent, rapid bolus administration would likely increase the risk of overspilling, because the low-viscosity fluid could tend to progress into more distal collateral coronary branches and might not diffuse into the neighboring tissues as intended. Although some operators recommend that the alcohol infusion be followed by a saline bolus to flush the alcohol away from the catheter,40 this practice may promote unwanted alcoholi-zation of remote, critical tissues such as the AV node. The balloon is commonly kept inflated for at least 5 minutes after the end of the alcohol infusion,14,40 mainly to allow clot organization and prevent vascular runoff of alcohol, while encouraging the highly diffusible alcohol to seep through the tissues, preventing dilution.
Potentially, any variation in the coronary artery anatomy is treatable by means of ASA, although the technique must be tailored to the specific anatomy of each patient and may require special operative skills. In some cases, the balloon catheter should be advanced subselectively to include only part of a large septal branch; in other cases, the quantity of alcohol injected into 1 septal branch should be limited to prevent overspilling in 1 direction (a visible collateral), or more than 1 septal branch may need to be treated (creative sculpting is frequently required).
Avoiding the onset of complete AV block (while keeping a temporary pacemaker in place during and for 2–3 days after ASA) is of paramount importance, not only because it makes ASA easier and less expensive, but also because it reduces the risk of AV block after hospital discharge—a potential cause of death.55–57 In fact, after ASA, the onset of AV block can be quite insidious, and sometimes arises after the 48-hour period during which a temporary pacing catheter is routinely kept in place and the patient is closely monitored. In addition, after ASA, complete AV block tends to present suddenly, and frequently leads to a poor escape fascicular or ventricular rhythm (even ventricular asystole), likely because these alternative pacemakers also may be abolished by ASA.55–57 Delayed AV block (observed after 10%–25% of procedures11,14,40,51–53) is more likely due to edema or to hemorrhage-related impairment55–57 than to alcohol-related direct necrotic injury, and often disappears spontaneously within a month. To decrease the risk of AV block, we recently tested the hypothesis that contralateral infusion of lactated Ringer's solution (subselectively infused at the AV-node artery, a branch of the distal right coronary artery) simultaneously with normal anteroseptal administration of alcohol may protect the AV node (Fig. 8). Our initial results with this investigative technique have been favorable yet inconclusive.
The basic mechanism of ASA is usually assumed to be shrinkage (in fact, ablation) of the infarcted area of myocardium. This assumption, however, fails to acknowledge that the favorable effects of ASA progress over time,58,59 typically in 3 phases:
During the initial, perioperative phase of flaccid aki-nesia, the behavior of subaortic stenosis is surprising and hard to explain in the absence of real ablation of the FSU. Akinesia, resulting from myonecrosis and secondary flaccidity of the treated upper portion of the septum, is the most likely reason for the immediate abolition of, or substantial decrease in, the outflow gradient at a stage in which there is no physical loss of tissue. During this early phase, left ventricular systolic pressure seems to cause temporary reshaping and widening of the outflow tract: the consequent diminution of the outflow blood velocity disrupts the Venturi effect, reducing the tendency of the anterior mitral valve leaflet to undergo SAM.39 Only by chance does early diminution of the subaortic gradient correspond so closely to that expected in the late phase, when true absorption of the scarred myocardium is complete (Fig. 9).
In reality, the determination and interpretation of both the preoperative baseline gradient and the postoperative final result is the topic of an evolving discussion in HOCM treatment. To determine a “true” baseline gradient, it has become standard practice to 1) withhold negative inotropic medications that would decrease the obstruction (especially disopyramide and β-blockers) for at least 2 days before ASA and 2) avoid diuretics, vasodilators, and other agents that might increase the gradient. As hemodynamicists and echocardiographers know, the HOCM gradient varies widely during any given observation period, even with minor changes in levels of respiratory effort, hydration, sedation, or anxiety. To “normalize the gradient” and improve the reliability of this marker of HOCM severity, one could try to induce the worst possible scenario for a given patient, for example by the administration of nitroglycerin (which is inexpensive and immediately available in the catheterization laboratory), dobutamine, or isoproterenol, or by inducing a premature ventricular contraction by programmed pacing.11,14,40
The early postoperative (“edematous halo”) phase is not a universal feature: it occurs in 50% to 60% of patients undergoing ASA.58,59 When present, this phase involves recurrence of the gradient, the systolic murmur, the mitral regurgitation, and the SAM, as shown by echo-cardiography.59
The cause of this phenomenon is obscure. To observers who are unprepared for it, the recurrence of the gradient is disappointing and confusing; in the early practice of the procedure, several patients even underwent an early, secondary surgical myomectomy due to the apparent failure of ASA.
Our preliminary cardiac magnetic resonance imaging studies have suggested that the onset of a variable degree of edema, hemorrhage, or both, at the site of alcoholization causes the recurrence of aortic stenosis by inducing increased turgor of the damaged myocardium (Fig. 9). The few available pathologic studies of necropsy specimens obtained a few days after ASA indicate that the infarcted territory bulges at this stage and that retractive scarring ensues only during the late period.55,56 Of note, at this stage (in the presence of a recurrent gradi-ent), patients typically continue to report substantial resolution of most symptoms (especially of chest pain or pressure), likely because of the alcoholization of sensory nerve endings and infarction of the culprit territory (where pain is apparently generated), rather than actual resolution of the subaortic stenosis.
During the late postoperative (“dense scar”) phase, mature scarring is accompanied by substantial thinning of the treated area1–14 (Fig. 9) and by definitive reduction of the subaortic gradient. Usually, SAM and mitral insufficiency are also substantially reduced at this stage.11,14,40 According to initial serial observations, over the 1st few years either after surgical myomectomy or after ASA, diffuse hypertrophy of the free left ventricular wall does not progress, as would otherwise be expected; rather, the ventricular wall undergoes substantial negative remodeling, losing 10% to 30% of its mass.1–14,40,60–64
Current experience indicates that an initial infarction involving some 15 to 20 g of myocardium results in the final ablation of about 129 g of left ventricular mass.63 Interestingly, during a myomectomy, surgeons remove only 3 g of myocardium on average,65 yet eventually achieve an even greater regression of hypertrophy at 1 year (reportedly, some 150 g).63 Most likely, the myotomy (longitudinal incision) that surgeons perform simultaneously with the myectomy (removal of myocardium) is an important contributor to their final result.
All of these findings support the belief that subaortic stenosis is a major factor in the progressive nature of HOCM—a genetic condition that affects initially and primarily the individual myofibers. Moreover, ASA consistently leads to stabilization and even regression of hypertrophy in HOCM, just as a surgical myomectomy does in the same condition, or as aortic valve replacement does in severe cases of aortic valve stenosis.
In any given case of HOCM, the risk of ventricular arrhythmias or sudden death occurring is not quite clear, nor is the effect of ASA on these risks. The frequent appearance of right bundle-branch block and other conduction defects after ASA51–53 and the formation of new scar tissue might be assumed to increase the risk of arrhythmias.26,66 On the contrary, the fact that most ventricular arrhythmias in HOCM seem to originate from the subaortic septum (which, after ASA, becomes predominantly a dense, electrically silent scar15,55,56) suggests that the risk of arrhythmias might be reduced (similarly to when alcohol is used for electrical ablation15). At our institution and at others, some patients have been treated with prophylactic automatic implantable cardiac defibrillators after ASA, as recommended by Maron and colleagues,66 but follow-up observations have indicated that the rates of arrhythmias and shock are quite low.11,14,40 Larger ASA series with long-term follow-up studies are necessary to establish definitive conclusions and recommendations in this regard.
For the treatment of symptomatic HOCM, ASA is a simple, expeditious option that offers important benefits affecting multiple aspects of this complex pathologic state. Cardiologists who perform ASA should become familiar with the newly recognized anatomo-functional entity, the FSU. The extent of septal hypertrophy and the number, size, intrinsic anatomy, and collateral connections of the relevant septal branches vary in individual cases, which contributes to the complexity of ASA interventions. Nuclear magnetic resonance imaging is an ideal complement to echocardiography for establishing the mechanisms of ASA and evaluating the changes in HOCM after this procedure.
A crucial, often difficult issue that deserves further investigation is how to determine the extent of the FSU in a given case. Defining the exact FSU borders seems to elude interventional cardiologists as it does cardiovascular surgeons, who have long struggled to establish clear guidelines concerning the length of myectomies.18–26 Because variations in specific features of the FSU are the main reason for this uncertainty (unlike, for instance, the consistency in aortic valve stenosis and its treatment), a better understanding of the anatomic, functional, and technical concepts discussed above is essential for optimizing the results of ASA.
For any given case, the objectives of ASA should be as follows: 1) substantial reduction of the subaortic gradient (achieved through immediate necrosis of the treated area that leads to delayed anatomic ablation); 2) elimination of the Venturi effect at the subaortic level, which should abolish or decrease SAM and secondary mitral regurgitation; 3) progressive, delayed negative remodel-ing of ventricular hypertrophy (and possibly prevention of the evolution of HOCM into dilated cardiomyopathy38,39); and 4) avoidance of AV blockade, ventricular tachycardia, and sudden death.
Address for reprints: Paolo Angelini, MD, P.O. Box 20206, Houston, TX 77225–0206. E-mail: moc.rr.notsuoh@dminilegnap