ABSTRACT: Background. Although alcohol septal ablation (ASA) is increasingly used in hypertrophic cardiomyopathy (HC) patients who are refractory to medical therapy, the amount of alcohol that is required has not been well studied. This study sought to determine the amount of alcohol that is necessary to achieve clinical benefits of ASA. Methods. Myocardial perfusion imaging was used to determine the size of the myocardial infarction produced by ASA in 54 HC patients. Left ventricular outflow gradients (LVOTg) were determined invasively before and after ASA and by Doppler echocardiography before and at a median of 3 months after ASA. Results. LVOTg decreased at rest and after provocation in response to ASA and this was maintained on follow-up at 3 months. There was no relationship between the amount of alcohol infused and the infarct mass as determined by myocardial perfusion imaging. While the infarct mass was not correlated with the drop in the LVOTg at rest or with provocation, the quantity of alcohol infused was correlated with the drop in LVOTg at rest (r = 0.27, p = 0.05) and with provocation (r = 0.34, p = 0.02). Furthermore, infusing more than 2ml of absolute alcohol was associated with a drop in the LVOTg by more than 60 mmHg at rest (p = 0.02) and by more than 130 mmHg with provocation (p = 0.05). Conclusions. Although lower amounts of alcohol infusion are desirable to avoid side-effects, it might be prudent to infuse around 2ml of absolute alcohol in order to achieve the desirable degree of LVOTg reduction in ASA.
J INVASIVE CARDIOL 2010;22:22–26
Key words: alcohol septal ablation, left ventricular outflow gradient,
Hypertrophic cardiomyopathy (HC) is a relatively common genetic disorder that is well characterized and known to be associated with left ventricular hypertrophy, myocardial fibrosis and disarray, and diastolic dysfunction.1 Although HC constitutes a spectrum of severities, from the asymptomatic to end-stage heart failure, the condition is well associated with dyspnea on exertion, angina, syncope and the unfortunate occurrence of sudden cardiac death.1 From the early descriptions of this entity, dynamic left ventricular (LV) outflow tract (OT) obstruction was recognized as a contributing factor.2 Although we know today that LVOT obstruction is not a defining feature of HC, its presence is associated with worse symptomatic deterioration and higher mortality.3-5 When the symptoms attributed to the LVOT gradient in HC are refractory to medical therapy, amelioration of the obstruction is generally achieved by surgical septal myomectomy, or alternatively, by percutaneous infusion of alcohol in a septal perforator artery — alcohol septal ablation (ASA).6-8 It is currently suggested that only 1–1.5 ml of absolute alcohol be injected slowly into the septal artery due to concerns for safety.9 Nevertheless, the amount of alcohol that is necessary to achieve clinical benefit has not been rigorously determined. We performed a retrospective review of the patients who underwent ASA at our institution to determine the relationship of the amount of alcohol infused to the resultant septal infarct, as well as to the reduction in LVOT gradients.
The study included patients with HC who underwent ASA at the University of Alabama at Birmingham between September 2003 and April 2008. ASA is performed at our institution for HC patients with an LVOT ≥ 30 mmHg at rest or ≥ 50 mmHg with provocation who have severe symptoms that are unresponsive to maximally-tolerated medical therapy.
ASA was performed as previously described.10 Briefly, a Sprinter 1.5–3.0 mm x 6 mm balloon (Medtronic, Minneapolis, Minnesota) was inflated in the septal artery and a small amount of absolute alcohol was injected with the balloon in the inflated position to prevent leakage into the main artery. The balloon was deflated 5 minutes after infusion of alcohol. The amount of alcohol infused and the size of the balloon were recorded. All procedures were guided by contrast echocardiography. Gradients across the LVOT were measured at rest and after a premature ventricular complex induced by catheter stimulation of the LV or following the administration of amyl nitrate (provocation) before and after the infusion of alcohol.
At a median of 2 days after ASA (interquartile range 1–2 days), myocardial perfusion imaging (MPI) was performed using gated single-photon emission computed tomography 60 minutes after intravenous injection of 20–40 mCi of Tc-99m sestamibi at rest. The images were acquired with a dual-head detector (ADAC, Miltipas, California) with and without attenuation correction using emission line source. Images were obtained in 32 projections (30s/projection) using 180º anterior elliptical arc with 20% energy window centered on 140 keV. The R-R cycle was divided into 8–16 frames and gating was done with a ± 50% window. Rotating images were reviewed for motion and reconstructed with filtered back-projection using a Butterworth filter and realigned to the long axis of the heart. The results presented here are without attenuation correction. Images were interpreted quantitatively using an automated operator-independent method based on polar maps and a customized database of HC patients as previously described.10 A customized database consisting of HC patients was used to account for the increased activity in the septum. The images were interpreted blindly without knowledge of the LVOT gradients or their response to ASA.
Patients also underwent 2-dimensional echocardiograms before and after ASA. The images were interpreted by an experienced reader who was blinded to the results of the MPI studies and the invasive determination of the LVOT gradients. These studies included determination of the LVOT gradient using continuous-wave Doppler with color Doppler guidance.11
All statistical analyses are carried out using SPSS® version 11.5 for Windows® (SPSS, Inc., Chicago, Illinois). Continuous variables are presented as mean ± standard deviation and discrete variables as frequencies and percentages. Chi-squared test was used for categorical variables when appropriate. Linear regression analysis was used to determine the relationship between the reduction in the LVOT gradients and defect size as determined by MPI. LVOT gradients before and after ASA were compared using paired-sample t-test. Changes in symptoms and functional status were analyzed using analysis of variance for repeated measures (ANOVA). All tests were two-tailed, and a p value of ≤ 0.05 was considered statistically significant. All patients gave informed consent for their procedure after understanding the available alternatives. Chart review was approved by the local institutional review board.
ASA was performed on 66 HC patients at our institution, and MPI studies were available for assessment of infarct size in 54 (82%) of them, which constituted the study population. The baseline characteristics of the patients are listed in Table 1. The technical success rate for ASA was 100%. On average, the length of stay in the hospital was 4.3 ± 2.5 days (range 2–14). The procedure involved infusing 2.15 ± 1.08 ml (range 0.6–5.0 ml, median 2) of absolute alcohol per patient in a coronary artery of average diameter 2.16 ± 0.32 mm (range 1.5–3.0 mm). This produced an infarct that was detected by MPI to involve 10.5 ± 10.6% of the LV, or 19.1 ± 18.6 grams. An implantable pacemaker was required in 8 (15%) patients due to complete heart block or the development of bundle branch block. The LV function showed a change on follow-up echocardiography in 2 patients whose ejection fraction dropped from normal to 40% and 46%, respectively. All other patients had normal LV function at follow up.
The LVOT gradient was significantly reduced after ASA both at rest and after provocation (Table 2). The LVOT gradient at rest decreased by 53.06 ± 28.11 mmHg or 94.40 ± 17.01% (median 60 mmHg), while after provocation it decreased by 131.87 ± 44.85 mmHg or 92.31 ± 9.43% (median 130 mmHg). After ASA, the patients showed significant improvement in their symptom control and functional status (Table 3). The LVOT gradient continued to be suppressed when a follow-up echocardiogram was performed with Doppler measurements at a median of 3 months after the ASA (interquartile range 2–4 months). The resting LVOT gradient at rest by Doppler echocardiography decreased from 63.02 ± 33.41 mmHg before ASA to 25.31 ± 24.32 mmHg (p
There was no relationship between the amount of alcohol infused in the coronary artery and the infarct mass as determined by MPI (r = 0.03, p = 0.8). The quantity of alcohol infused was correlated with the drop in LVOT gradient at rest (r = 0.27, p = 0.05) and with provocation (r = 0.34, p = 0.02). Furthermore, infusing more than 2ml of absolute alcohol was associated with a drop in the LVOT gradient by more than 60 mmHg at rest (p = 0.02) and by more than 130 mmHg with provocation (p = 0.05) (Figure 2A). The infarct mass was not correlated with the drop in the LVOT gradient at rest (r = 0.05, p = 0.7) or with provocation (r = 0.10, p = 0.5), and infarcts that are larger than 15grams were not correlated with a reduction of the LVOT gradient by more than 60 mmHg at rest (p = 1.0) or by more than 130 mmHg with provocation (p = 1.0) (Figure 2B). An implantable pacemaker was required in 3 out of 22 patients (14%) who received more than 2ml of absolute alcohol and in 5 out of 32 patients (16%) who received 2ml or less absolute alcohol (p = 0.9). The infusion of alcohol did not result in the obliteration of the septal artery in all patients and this did not preclude benefit from the procedure (Figure 3).
The major finding of this study is that the amount of alcohol infused in the coronary artery for ASA is not correlated to the infarct size, but it is correlated to the reduction in the LVOT gradient. Indeed the infarct size (or mass) as assessed by MPI was not correlated to the reduction in gradient. The gist of the data would suggest that although lower amounts of alcohol infusion are desirable to avoid side effects, it might be prudent to infuse approximately 2 ml of absolute alcohol in order to achieve the desirable degree of LVOT gradient reduction. Furthermore, since the septal artery was not obliterated in all patients, and clinical response was evident despite the persistence of the artery, this radiological sign cannot be used to guide the operator as to the amount of alcohol that is needed in a particular patient.
Although the mainstay of treatment of HC, even in patients with significant LVOT obstruction, is pharmacological, alternative options have evolved over the years to address patients who cannot tolerate medical therapy or those who remain symptomatic despite maximum therapy.12 ASA has emerged as the most prominent substitute for the more invasive surgical myomectomy. Despite the absence of any prospective randomized comparison, recent reports indicate that it is a similarly effective and safe option.6,13,14 The basic tenet of the procedure involves infusing alcohol in one of the septal arteries to induce a small infarct in the base of the septum at the site where LVOT obstruction occurs, thereby resulting in an immediate and sustained reduction in the LVOT gradient.6,10 In our study, patients experienced benefits with regard to their symptomatic and functional status (Table 3) and a significant reduction in their LVOT gradient measured invasively immediately following ASA (Table 2) and by Doppler echocardiography measured 3 months following the procedure.
There has been healthy skepticism and concern regarding the widespread use of ASA due to the lack of long-term safety of the procedure.15 There is particular concern regarding remodeling of the LV after ASA and the presumed arrhythmogenic nature of the infarct. Despite some reports showing a slight advantage of surgical myomectomy,16 most long-term follow-up studies have by far dispelled these concerns.14,17–20 We have previously examined patients with serial MPI following ASA and demonstrated a reduction in the size of the perfusion abnormality late after ASA, despite no increase in the LVOT gradient or recurrence of symptoms.10 The patients, therefore, continued to reap the symptomatic and hemodynamic benefits of ASA, even with smaller infarcts. It has been suggested that reducing the volume of alcohol infused in the coronary artery might decrease the adverse effects without limiting the effectiveness of the procedure. Veselka et al21 randomized 34 patients with HC to either low- (1.6 ± 0.4 ml) or high- (3.4 ± 0.9 ml) dose alcohol infusion. They reported a significant correlation between the volume of injected alcohol and the size of the infarct (peak CK-MB) and no relationship between peak CK-MB and the LVOT gradient at follow up. The same group also reported that decreasing the volume of alcohol does not affect the LVOT gradient immediately after the procedure, at 3–5 days, 3 weeks and 3 months after ASA,22 but it does result in less thinning of the basal septum on follow-up echocardiography.23 They recently reported that using as low as 1 ml of alcohol resulted in similar clinical and echocardiographic benefit to higher doses.24 To our knowledge however, ours is the first study that examines the relationship between the volume of alcohol and the infarct size as determined by MPI. We found no correlation between the volume of alcohol infused and the size of the septal infarct. MPI is a more reliable method to measure the size of the infarct than single measurement of CK-MB.25 Despite this, infusing more than 2 ml of absolute alcohol resulted in a greater drop in the LVOT gradient, and there was a direct correlation between the volume of alcohol infused and the reduction in the LVOT gradient. This occurred despite similar baseline characteristics between patients who received more than and those who received less than 2 ml of absolute alcohol during their ASA (Table 1). It is interesting to note that there was no correlation between the volume of alcohol and LVOT gradient after ASA, even in our study (p = 0.7), but the reduction in LVOT gradient after ASA (taking the baseline LVOT gradient into account) was larger in the patients who received more than 2 ml alcohol. Furthermore, in our cohort, there was no detectable increase in the proportion of patients who required a permanent pacemaker due to the development of heart block or bundle-branch block in patients who received more than 2 ml of alcohol.
We are cognizant of the limitations imposed by the retrospective observational nature of our study. In particular, the amount of alcohol infused was not randomized, but determined by the interventionist performing the ablation (single operator for all procedures). There were various determinants of the volume infused, but the most important was the size of the artery injected as well as the response of the LVOT gradient on continuous monitoring. In addition, all the patients in our study, including those who received less than 2 ml of absolute alcohol, did symptomatically very well after ASA. Nevertheless, this is the only study that has examined the relationship of the amount of alcohol infused in the coronary arteries during ASA and the infarct size produced as assessed by MPI and the reduction of LVOT gradients both immediately after ASA and several months later.
From the Division of Cardiovascular Disease, University of Alabama at Birmingham, and the Division of Cardiology, Birmingham Veterans Affairs Medical Center, Birmingham, Alabama.
*The first 2 authors contributed equally to this work
The authors report no conflicts of interest regarding the content herein.
Manuscript submitted May 4, 2009, provisional acceptance given June 15, 2009, final version accepted July 30, 2009.
Address for correspondence: Raed A. Aqel, MD, BDB 383, 1530 3rd Avenue S., Birmingham, AL 35294-0007. E-mail : firstname.lastname@example.org