Detectable Serum Cardiac Troponin T as a Marker of Poor Prognosis Among Patients With Chronic Precapillary Pulmonary Hypertension

Differential Diagnosis

This trial reports on baseline cardiac troponin T (cTnT) assessed in serum samples of patients ultimately diagnosed as having pulmonary arterial hypertension (PAH, n=51) or distal chronic thromboembolic pulmonary hypertension (CTEPH) not suitable for surgical treatment (n=5). The diagnosis required adherence to a local algorithm, including in all patients lung function tests and arterial blood gases, perfusion lung scintigraphy, spiral computed tomography with both contrast enhancement and high resolution assessment, transthoracic echocardiography, and, finally, right heart catheterization with direct pulmonary hemodynamic measurements including pulmonary flow. In addition, standard contrast pulmonary angiography and transesophageal echocardiography were performed in selected patients (5 and 49, respectively) whenever considered appropriate. CTEPH was considered nonoperable only after consultation with a surgical team experienced in pulmonary thromboendarterectomy. Pulmonary arterial hypertension was diagnosed according to WHO Evian classification.⁶ It was associated with collagen vascular disease in 10 patients, Eisenmenger syndrome in 4 patients (ventricular septal defect, patent ductus arteriosus recanalized after closure, aortopulmonary window with PH developing despite surgical closure, and atrial septal defect with partial anomalous pulmonary venous drainage, respectively), and portal hypertension in 1 patient. Primary pulmonary hypertension was eventually diagnosed in 36 patients.

Troponin T

Peripheral venous blood sampling for cTnT was performed after informed consent in fasting state on the morning of the day of the 6-minute walk test (6mWT). This was a part of a trial evaluating safety of exercise testing of patients with severe chronic PH, which was approved by a local ethical committee. Serum was centrifuged and stored frozen for later serial analysis of cTnT levels, which did not influence management. cTnT was assessed with highly sensitive third-generation quantitative test (electrochemiluminescence method ECLIA, Roche Diagnostics). Detection limit for cTnT was 0.01 ng/mL. According to the manufacturer, 99% of the healthy population has cTnT <0.01 ng/mL, as assessed with this test. Thus, detectable concentrations of cTnT ≥0.01 ng/mL were considered abnormal.

In addition to initial evaluation, 52 of 56 patients had at least 1 reassessment of cTnT during the follow-up period. In 31 patients, serum cTnT level was reassessed within 24 hours to check for the reproducibility of the results. In 43 of 56 patients, a total of 141 serum samples were evaluated for cTnT beyond the 2nd day but within the 2nd year of the follow-up. In 26 patients, results of at least 4 serial cTnT tests were available for analysis.

Prognostic Evaluation

All patients underwent baseline evaluation including assignment of WHO functional class and unencouraged 6mWT with self-evaluation of dyspnea using Borg dyspnea score. Serum for assessment of N-terminal pro-BNP (NT-proBNP) with electrochemiluminescence method (ECLIA, Roche) was secured. At right heart catheterization, mean pulmonary arterial, mean right atrial pressures, pulmonary artery occlusion pressure, and mixed venous oxygen saturation were assessed. Pulmonary vascular resistance and cardiac index were calculated based on thermodilution or Fick measurements of pulmonary flow, whichever was appropriate. Oxygen consumption was directly measured in all patients with Eisenmenger syndrome. Pulmonary vasoreactivity was assessed with inhaled nitric oxide administered as 20-ppm mixture over 5 minutes. In addition, echocardiographic studies were performed with patients in left lateral position using the methodology previously reported as useful for evaluation of patients with primary PH. The ratio of diastolic areas of both ventricles, RV systolic area change, and left ventricular eccentricity index were assessed.⁷ Doppler index of myocardial function was calculated.⁸ Right atrial area was measured.⁹ The amount of pericardial effusion, if present, was assessed using a previously described score.⁷

Treatment and Follow-Up

Fifteen of 56 patients were taking chronic low-dose calcium channel blockers already at presentation. According to local management strategy, the patients who decreased mean pulmonary arterial pressure by ≥20% and pulmonary vascular resistance by ≥30% (n=5 among 51 patients with PAH) received chronic oral treatment with calcium channel blockers in maximal tolerated doses. All patients received chronic anticoagulation, unless contraindicated. Patients with signs of fluid retention received diuretics. Ten patients were undergoing chronic digoxin treatment, and this treatment was maintained in the absence of side effects. Chronic oxygen therapy was introduced in a few patients with resting hypoxemia. Neither chronic intravenous epoprostenol therapy nor lung or heart and lung transplantation were available therapeutic options during the time of this study. However, most nonresponders in functional class III/IV attributable to PAH after initial evaluation received either oral (n=21) or subcutaneous (n=12) prostacyclin analogues or endothelin receptor blocker (n=7). Patients in the CTEPH group were treated with conventional therapy and subcutaneus (n=3) or inhaled (n=1) prostacyclin analogues. Two patients underwent balloon atrial septostomy. Patients were followed for a mean of 17±8.5 months (range, 0.5 to 24 months).

Statistical Analysis

Patients with cTnT(+) and cTnT(−) were compared with respect to clinical characteristics, functional reserve, hemodynamics, and outcome. Data are expressed as mean±SD. For categorical variables, the differences between the groups were compared using Yates corrected χ² test. For continuous variables, Student’s t test or Wilcoxon test was used depending on the character of distribution. P<0.05 was considered statistically significant. The proportion of patients surviving at each time point was estimated by the Kaplan-Meier method. Survival for the groups with and without detectable cTnT was compared using the log-rank test. Univariate and multivariate Cox proportional hazard analysis of selected variables was performed to identify factors independently related to mortality and providing best-fitted model related to outcome. Results are expressed as hazard ratios with 95% CIs. All analyses were performed using STATISTICA 5.5 (StatSoft) computer software.

Repeated cTnT Assessment

In clinically stable patients with PH, cTnT levels were reproducible. When rechecked after 24 hours, both tests were negative in 27 patients, whereas in 3 patients both tests were positive and revealed almost identical cTnT levels (0.049 and 0.046, 0.032 and 0.034, and 0.019 and 0.017 ng/mL, respectively) In 1 patient, TnT was negative on first evaluation and borderline (0.010 ng/mL) after 24 hours. This patient clinically improved as a result of combined treatment with bosentan, beraprost, and atrial septostomy and remained cTnT(−) at days 338 and 656 of follow-up.

Among 8 initially cTnT(+) patients, 7 had at least 1 reassessment of cTnT status. In 3 patients, cTnT remained increased whereas in 1 it was no longer present on day 210 of beraprost treatment but reappeared at day 337. All 4 of those patients died. In contrast, 3 patients in whom repeated tests revealed disappearance of initially elevated cTnT (0.029, 0.015, and 0.021 ng/mL, respectively) survived. They were treated with continuous subcutaneous treprostinil infusion, oral bosentan, and septostomy on top of combined prostanoid treatment (beraprost and treprostinil), respectively.

Among initially cTnT(−) patients, 3 became cTnT(+) during the first and an additional 2 during the second year of follow-up. Two of the 5 patients died. In both cases, cTnT was detected during the first year of follow-up.

Limitation of the Trial

The study population was not entirely homogeneous, because it included patients representing various classes of PAH as well as a small group with nonoperable CTEPH. One might expect CTEPH patients, usually older and more likely to have concomitant coronary atherosclerosis, to suffer myocardial damage and release cTnT more easily when faced with severely altered pulmonary hemodynamics than patients with PAH.²⁴ However, otherwise, CTEPH and PAH share similar mechanisms leading to RV ischemia in the setting of severe RV pressure overload. We noticed neither clear differences in the prevalence nor in the prognostic implication of detectable serum cTnT between those 2 groups.

Although patients in our trial were treated in several different ways, this did not significantly affect their outcome, whereas cTnT retained its prognostic significance. This may be attributable either to the small therapeutic subgroups or indicate that patients with cTnT(+) require therapy that was not available (intravenous prostanoids or lung transplantation) or was rarely used (atrial septostomy) in our study population. Only some of the previously identified prognostic markers were found significant in our study population (functional class, 6mWT, cardiac index, and pulmonary vascular resistance), whereas others were not (right atrial and pulmonary arterial pressures and mixed venous oxygen saturation). This could be attributable to the relatively small study group.

In summary, detectable cTnT is a so-far ignored independent marker of increased mortality risk in patients with chronic severe pulmonary arterial or thromboembolic hypertension, supporting the role of progressive RV myocyte injury in the vicious circle leading to hemodynamic destabilization. Identification of effective methods protecting RV myocytes against cTnT leakage and impact of such protection on outcome of patients with precapillary PH represents an interesting direction for future research.