Methods

The serum NT-proBNP levels were determined in 218 consecutive outpatients with HC who had been referred to the Referral Center for Cardiomyopathies in Florence (Italy) before January 2011 as a part of a comprehensive noninvasive cardiovascular evaluation, including clinical history, physical examination, 12-lead electrocardiography, and echocardiography. The diagnosis of HC was determined by 2-dimensional echocardiographic identification of left ventricular (LV) hypertrophy associated with nondilated ventricular chambers in the absence of any other cardiac or systemic disease capable of producing the magnitude of hypertrophy evident. Of the 218 patients, 35 with known coronary artery disease (n = 15) or evidence of systolic dysfunction at the first evaluation (defined as a LV ejection fraction <50%; n = 20) were excluded. The remaining 183 patients with HC constituted the study group enrolled in the present project. All had had a stable clinical course, without episodes of acute HF or worsening of symptoms, in the previous 3 months.

The local ethics committee approved the study, and all patients gave informed consent to participate. Peripheral blood samples, collected after 30 minutes of complete rest, were measured using an electrochemiluminescent immunoassay (Elecsys proBNP II assay, Roche Diagnostics, Indianapolis, Indiana) on a Cobas E601 analyzer (Roche Diagnostics).

Transthoracic echocardiography was performed using commercially available equipment. The standard evaluation included M-mode, 2-dimensional, and Doppler study, according to the recommendations of the American Society of Echocardiography. LV systolic dysfunction, characterizing end-stage HC, was defined as an LV ejection fraction <50%. A restrictive LV filling pattern was defined as a ratio of early to late velocity of mitral inflow ≥2, associated with a deceleration time of ≤150 ms (or a deceleration time <120 ms for patients with atrial fibrillation). LV outflow tract obstruction was considered present when a peak instantaneous outflow gradient of ≥30 mm Hg was measured using continuous wave Doppler echocardiography under basal (at rest) conditions or with physiologic provocation (Valsalva maneuver or exercise).

After enrollment, the patients were evaluated annually or more often if dictated by the clinical status, using 12-lead electrocardiography, echocardiography, and 24-hour electrocardiographic Holter monitoring. Pharmacologic therapy was prescribed as appropriate, according to the current HC guidelines. Patients with refractory symptoms related to LV outflow tract obstruction despite optimal medical therapy were referred for surgical myectomy; alcohol septal ablation was reserved for patients at high operative risk owing to advanced age and co-morbidities. An implantable cardioverter-defibrillator was offered to survivors of cardiac arrest or sustained ventricular arrhythmias, for primary prevention in high-risk patients evaluated using established risk markers for sudden cardiac death, and in patients with end-stage disease.

Genetic counseling and mutational analysis were offered to all patients as a part of the standard policy at our center. After providing informed consent, 133 of the study patients (73%) were screened for mutations in the protein-coding exons and splice sites of 8 myofilament genes, including myosin binding protein C, β-myosin heavy chain, the regulatory and essential light chains (MYL2 and MYL3), troponin-T, troponin-I, α-tropomyosin, and α-actin. Direct deoxyribonucleic acid sequencing was performed using ABI-Prism 3730 (Applied Biosystems, Foster City, California), as previously described.

The primary end point was defined as a composite of cardiovascular death, heart transplantation, and resuscitated cardiac arrest. Death was considered to be of cardiovascular origin if due to worsening HF, sudden death, or related to cardiac surgery. Sudden cardiac death was defined as unexpected sudden collapse or death occurring during sleep in patients who had previously experienced a stable clinical course. The secondary end point (SE) was a composite of HF-related events, including HF death, pulmonary edema requiring hospitalization, progression to advanced symptomatic status (New York Heart Association class III or IV), end-stage disease, onset of atrial fibrillation, and stroke.

The differences between groups were analyzed using Pearson’s chi-square test for discrete variables and the unpaired Student t test or 1-way analysis of variance, as appropriate, for continuous variables. Because the NT-proBNP values were not normally distributed in our population, logarithmic transformation was applied. For the same reason, the patients were divided into tertiles according to their NT-proBNP levels. Nonparametric tests were used for the comparisons among the groups (Mann-Whitney U test) and for correlation analysis (Spearman’s test). Survival curves were estimated using the Kaplan-Meier method, and the log-rank test was used to compare the survival distributions of the groups. The performance of NT-proBNP in predicting the primary end point and SE was assessed using receiver operating characteristic analysis. Calculations were performed using the Statistical Package for Social Sciences, version 19.0.0, software (SPSS, Chicago, Illinois).

Results

The mean age of the 183 study outpatients was 50 ± 17 years. Most were men (64%) and had no or only mild symptoms at their initial evaluation (New York Heart Association class I or II for 82%); 26% had a history of atrial fibrillation ( Table 1 ). On echocardiographic examination, the maximum LV wall thickness was 22 ± 6 mm, 36% had at rest or inducible LV outflow tract obstruction, and 12 patients (6.5%) had a restrictive LV filling pattern. Of the study population, 2/3 were receiving long-term therapy with β blockers, and 12% were taking amiodarone. Also, 32 patients (17%) received an implantable cardioverter-defibrillator for primary prophylaxis of sudden cardiac death because of their risk profile ( Table 1 ).

The median NT-ProBNP value for the study group was 615 pg/ml (range 22 to 16,013). Because the NT-ProBNP distribution was not normally distributed in our population, the patients were grouped according to the tertiles, with the 33rd and 67th percentile values used as cutoffs (i.e., 310 and 1,025 pg/ml, respectively). Patients in the upper tertile had a larger left atrial diameter and maximum LV wall thickness and were more likely to have advanced symptoms (New York Heart Association class III), atrial fibrillation, and dynamic LV outflow tract obstruction ( Table 1 ). The women had greater NT-proBNP values than the men (median 1,074 vs 401 pg/ml, respectively; p = 0.001) and represented most patients in the upper tertile ( Table 1 ). The NT-proBNP values were comparable in patients taking and not taking β blockers (1,097 ± 1,314 vs 1,178 ± 2,362 pg/ml, respectively; p = 0.76), verapamil (1,021 ± 1,328 vs 1,147 ± 1,844 pg/ml, p = 0.69), and amiodarone (1,107 ± 1,828 vs 1,231 ± 1,250 pg/ml, p = 0.27). Of the 119 patients who underwent cardiac magnetic resonance imaging, we found no differences in the NT-proBNP values between the 73 patients with late gadolinium enhancement and the 46 without (1,098 ± 1,485 vs 1,143 ± 1,903 pg/ml, respectively; p = 0.86).

Of the 134 patients who had underwent genetic screening, 98 (73%) were genotype positive and 36 (27%) had negative test results. The relative prevalence of genes involved was 45% for myosin binding protein C, 35% for β-myosin heavy chain, 5% for both regulatory light chain and troponin-T, and 1% for troponin-I. Of genotype-positive patients, 8% had multiple mutations (including 7% with double and 1% with triple mutations). The NT-proBNP values were similar among the genotype-positive and genotype-negative patients (median value 1,297 vs 1,041 pg/ml; p = 0.44). Furthermore, no differences were found among the genotype-positive patients regarding the specific gene affected (overall p [analysis of variance] = 0.97).

During 47 ± 34 months of follow-up, 8 patients (4%) reached the primary end point. Of these, 4 died suddenly—2 patients died from HF-related causes, 1 patient died perioperatively after surgical myectomy, and 1 patient underwent transplantation. None of the patients experienced appropriate implantable cardioverter-defibrillator intervention. The SE occurred in 67 patients (37%) during the follow-up period, at a rate of 9.4% annually; 33 patients (18%) experienced >1 event. The cumulative rates of HF-related events were 21% (5.3%/yr) for acute HF requiring hospitalization, 11% (2.8%/yr) for worsening of symptomatic status to New York Heart Association class III or IV; and 8% (1.9%/yr) for new-onset atrial fibrillation. Eight patients (4%; 1.1%/yr) experienced cerebrovascular events judged to be cardioembolic. The global rate of progression to end-stage disease was 11.5% (n = 21, 2.8%/yr).

When the events were analyzed with regard to the NT-proBNP distribution, the cumulative rate of the primary end point at the end of the follow-up period was 5% (1.3%/yr) and 8% (2.1%/yr) in the middle and upper tertile, respectively. No event occurred in the lower tertile. As shown in Figure 1 , the patients in the lower tertile (<310 pg/ml) had a favorable outcome compared with those in the other tertiles (overall p = 0.01). The patients in the middle and upper tertiles, however, had comparable survival rates. On multivariate Cox regression analysis, only NT-proBNP and a restrictive LV filling pattern were independently associated with the primary end point (hazard ratio [HR] for log-transformed NT-proBNP 5.8, 95% confidence interval [CI] 1.07 to 31.6, p = 0.04; HR for a restrictive LV filling pattern 5.19, 95% CI 1.23 to 21.9, p = 0.02; Table 2 ). Of the patients in the middle NT-proBNP tertile, the presence of a restrictive LV filling pattern conferred additional prognostic stratification, although it had no additional value for patients in the upper tertile ( Figure 1 ). Receiver operating characteristic curve analysis demonstrated moderate power for NT-proBNP in the prediction of the primary end point (area under the curve 0.73, p = 0.02). The NT-proBNP value of 810 pg/ml proved the best threshold for the identification of patients at risk of cardiovascular death, with high sensitivity (88%) but low specificity (61%; Figure 2 ).

(A) Survival free from the primary end point (PE; a composite of cardiovascular death, heart transplantation, and resuscitated cardiac arrest) stratified by NT-proBNP tertiles (overall log-rank p = 0.01). After pairwise comparison with Bonferroni correction, only the lower to upper tertile comparison reached statistical significance (p = 0.01; HR 9.26, 95% CI 1.58 to 53.9). (B) Freedom from PE stratified by NT-proBNP tertiles and presence or absence of mitral pulsed wave restrictive filling pattern (RFP).

Stay updated, free articles. Join our Telegram channel

Join