Regression of Left Ventricular Hypertrophy After Nonsurgical Septal Reduction Therapy for Hypertrophic Obstructive Cardiomyopathy

Patient Population

The Baylor College of Medicine Institutional Review Board approved the study protocol, and all patients provided written informed consent before participation. The first 26 consecutive patients with symptomatic HOCM and documented LVOT obstruction gradient (≥40 mm Hg at rest or ≥60 mm Hg on dobutamine provocation: mean dose 15±3 μg · kg⁻¹ · min⁻¹) who underwent NSRT, as previously described,¹⁰ and completed a 2-year follow-up composed the study cohort. The group had a mean age of 53±15 years (9 of 26 women). All patients had a septal wall thickness ≥1.5 cm with a ≥1.3 ratio of septum to posterior wall thickness. All had dyspnea, 17 had angina, and 12 suffered from presyncope/syncope.

Patients completed an NYHA classification questionnaire as well as a Bruce protocol stress test at baseline and 1 and 2 years after the NSRT procedure. Creatine kinase (CK) levels were determined before and every 6 hours during the 24-hour period after NSRT.

Echocardiographic Studies

Echocardiograms were performed with an HP or Acuson ultrasound imaging system equipped with 2.5-, 3.5-, and 5-MHz transducers. Standard parasternal and apical views were obtained. Short-axis tomograms were acquired at 3 levels: mitral valve, papillary muscle, and apex.

Color-guided continuous-wave Doppler was applied in the apical views to determine the peak LVOT gradient, with care taken to avoid contamination with the mitral regurgitation jet.¹¹ Studies were stored for later analysis.

Echocardiographic Analysis

Using a computerized reading station offline, a single observer blinded to the patients’ identity, clinical data, and study sequence performed all measurements at baseline and 1 and 2 years after NSRT.

LV minor and major dimensions at end diastole and end systole, wall thickness, end-diastolic volume (EDV), and ejection fraction (EF) were measured according to American Society of Echocardiography recommendations.¹² Left atrial volume was derived with the multiple-disks method.¹³

LVH was assessed according to previously published criteria.^{12 14} First, the wall thickness of each of the following myocardial segments was measured at both the mitral valve and the papillary muscle levels in the short-axis view¹⁴ : anterior septum, anterior lateral, inferior lateral, and inferior septum. Subsequently, the total wall thickness score was calculated at both levels.¹⁴ Second, end-diastolic myocardial areas at the mitral, papillary muscle, and apical levels recorded in the parasternal short-axis views were planimetered. Third, LV mass and mass indexed to body surface area were calculated.¹² The end-diastolic myocardial area at the papillary muscle level was used to derive mass.

Because NSRT results in an infarction limited to the anterior septum basal segment that does not extend to the papillary muscle level, this approach would tend to underestimate the overall extent of regression in LVH. A good correlation was present between total myocardial area and mass at baseline and at 1 and 2 years (r=0.8, 0.86, and 0.89, respectively, all P<0.01).

Reproducibility

For intraobserver variability (12 patients analyzed), the 95% interval of agreement was −5% to 7% for segmental wall thickness, −8% to 10% for total thickness score, −11% to 14% for myocardial area, and −12% to 14% for mass.

SPECT Myocardial Scintigraphy

Stress single photon emission CT (SPECT) imaging 6 weeks after NSRT was used to determine infarct size and was performed by previously reported methods,¹⁰ with images reconstructed and reoriented in standard views. Experienced nuclear cardiologists, blinded to all other data and using raw polar maps statistically compared with a corresponding normal data bank, determined the SPECT defect size.

Statistics

Repeated measures of ANOVA or ANOVA on ranks were applied to evaluate changes in clinical and echocardiographic parameters at the 3 time points. Bonferroni t or Student-Newman-Keuls tests were used for all paired comparisons. The relation between changes in mass, infarct size, and acute reduction in LVOT gradient was evaluated by simple linear regression analysis. The study had >80% power to detect a 20% change in LVH. Statistical significance was declared if P≤0.05.

Changes in LV Size and EF

Although the long-axis dimension was unchanged, LV end-diastolic dimensions (anteroposterior and mediolateral) increased significantly after NSRT (Table 1⇑). Likewise, the minor-axis end-systolic dimensions increased (P<0.01), leading to significant increases in EDV (Figure 1A⇓) and end-systolic volumes (P<0.01). LVEF was relatively unchanged; however, EF at year 2 was statistically lower (P<0.05).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Individual data points of LVEDV (A), wall thickness score at mitral valve level (B), wall thickness score at papillary muscle level (C), and mass (D) at baseline and 1 and 2 years after NSRT.

Regression of LVH

Septal thickness at the infarction site decreased from 20 mm before to 12 mm at 1 year (P<0.01) and 10 mm at 2 years (P<0.01 versus baseline, P>0.05 versus 1 year), and all parameters of LVH showed evidence of regression. The anterior septum thickness (distal to infarction site), inferolateral wall, inferior septum, and anterolateral wall at both the mitral valve and papillary muscle levels were significantly less both 1 and 2 years after NSRT (Table 2⇓). A number of segments became significantly thinner, and the total wall thickness score at both levels continued to decrease (Figure 1B⇑ and 1C⇑) up to 2 years after NSRT.

Similarly, the myocardial area at the mitral, papillary muscle, and apical levels was significantly smaller after NSRT (P<0.01).

LV mass also continued to decrease through the 2-year follow-up (Figure 1D⇑); likewise, mass corrected for body surface area decreased significantly (P<0.001). This latter change was greater in year 1 than in year 2 (median values 38 versus 11 g/m²; P<0.05). The incidence of complete heart block was distributed equally between patients having LVH regression ≥20% or <20%.

Determinants of LVH Regression

Significant but weak correlations were present between the reduction in LV mass at 2 years and the infarct size as assessed by both CK (r=0.48, P=0.05) and SPECT imaging (r=0.4, P=0.05). This weak correlation was due to several large infarcts and big gradient reductions when we initially performed the procedure and to our later ability to reach successful hemodynamic results despite smaller infarcts by targeting ethanol delivery to the culprit septal segments using myocardial contrast echocardiography. The strongest relation was evident with the acute reduction of LVOT gradient at the time of the procedure (r=0.63, P=0.01; Figure 2⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Relation of change at 2 years in LV mass to peak CK (left) and acute LVOT gradient reduction (right).

On multiple regression analysis, CK leak and the acute reduction in gradient accounted for 52% of the variance in LV mass reduction (r=0.72, R²=0.52, P<0.001).

Implications for the Pathogenesis of HOCM

LVH in HOCM is believed to be a compensatory process secondary to the decreased contractility induced by the mutation that leads to increased ventricular pressure and stress, which induces hypertrophy,¹⁶ which is further enhanced by obstruction.¹⁷ Results of the present study show the association of LVH regression with the elimination of obstruction, confirming the notion that hypertrophy may be a secondary phenomenon.

Implications for Treatment

In our study cohort, NSRT essentially eliminated the LVOT gradient and was associated with marked reduction in symptoms and improved exercise tolerance. These long-term beneficial results are similar to those achieved with surgery.^{5 6} NSRT also resulted in regression of LVH, which may be another beneficial effect, given that the frequency of sudden death in HOCM patients increases with increased LVH.⁴ Regression of LVH may also contribute to symptomatic improvement and is probably related to the fact that the increasing hypertrophy in patients with obstruction contributes to a decrease in LV compliance and impaired exercise tolerance. The resultant septal infarction of NSRT, however, may predispose patients to ventricular dysrhythmias and potentially offset the benefit of LVH regression. Clearly, more prospective data are needed, because the number of patients followed up, despite no occurrence of sudden death over 2 years, is too small to determine the impact of NSRT on ventricular dysrhythmias. Also, when counseling HOCM patients regarding this procedure, serious complications of NSRT noted by others and ourselves should be considered. Side effects observed in our later experience, but not in the present group, include left anterior descending coronary artery dissection (6 patients, 3.2%), death (3 patients, 1.6%), and sustained ventricular tachycardia (1 patient, 0.5%, not on β-blockers).

Although complete heart block developed in 7 patients (27%), including 4 with CK >2500 U/L, after we modified our technique, the incidence of heart block in the subsequent 162 HOCM patients decreased to 8.6%. Our modifications included injecting ethanol at a slower rate (1 to 1.5 mL/min instead of bolus) and using intracoronary myocardial contrast echocardiography to help target ethanol to the culprit septal segments. This latter incidence (8.6%) of complete heart block after NSRT is close to the 5% to 7% rate reported after surgery.