Impaired right ventricular systolic function (RVSF) may complicate the treatment of constrictive pericarditis (CP) by pericardiectomy, which is a procedure that remains with significant morbidity and mortality. We evaluated RVSF in patients with CP who underwent pericardiectomy to determine the prognostic value of RVSF. RVSF was assessed by measuring Tricuspid Annular Plane Systolic Excursion (TAPSE) in 35 patients (mean age 52 ± 15.4 years) who underwent pericardiectomy. Thirty-one patients (88.6%) had reduced RVSF (TAPSE ≤1.8 cm). Eight patients (23%) had postoperative events (heart failure 3 and hospital mortality 5). Logistic regression showed that concomitant coronary artery bypass grafting (CABG) (p = 0.052), left ventricular ejection fraction (p = 0.059), left atrial diameter (p = 0.028), and TAPSE (p = 0.016) were borderline or significant univariate predictors of events. TAPSE (p = 0.018, odds ratio = 0.605 [0.40 to 0.92]) and CABG (p = 0.033, odds ratio = 20 [1.26 to 315]) were independent predictors of events on multivariate analysis. Stepwise analysis showed that TAPSE provided incremental prognostic value (p = 0.029, chi-square increase 11.6 to 16.3) to the combination of CABG, ejection fraction, and left atrial diameter. Receiver-operating characteristic curve analysis showed an area under the curve of 0.815 for TAPSE. TAPSE of 1.38 cm had a sensitivity of 88% and specificity of 67% for identifying patients with events. TAPSE was also inversely related to the length of hospital stay after pericardiectomy (p = 0.02, R = −0.424). Hence, our study showed that RVSF is frequently reduced in patients with CP who underwent pericardiectomy. In conclusion, TAPSE is an independent predictor of events and provides incremental prognostic value to other clinical and echocardiographic variables.
Constrictive pericarditis (CP) is the end result of an inflammatory process that leads to a thickened, fibrotic, and sometimes calcified pericardium that limits diastolic filling of the ventricles. Surgical removal of the pericardium generally results in improvement of patient’s symptoms and survival in those with progressive constriction. However, there is substantial perioperative morbidity and mortality associated with pericardiectomy. Reported post-pericardiectomy mortality varies from 6% to 14% with most modern studies reporting perioperative mortality in the range of 6% to 8%. Low cardiac output may be the most common cause of perioperative morbidity and mortality. Epicardial involvement with associated myocardial contractile dysfunction or longstanding, irreversible ventricular remodeling have been proposed as possible causes of low cardiac output. We have previously reported in several patients with CP that reduced longitudinal right ventricular systolic function (RVSF) is associated with epicardial involvement, more difficult removal of the pericardium, and persistent RV dysfunction after pericardiectomy. However, the clinical value of assessment of RVSF in patients with CP has not been established. In this retrospective study, we used Tricuspid Annular Plane Systolic Excursion (TAPSE) to determine the prevalence of reduced RVSF in patients with CP who underwent pericardiectomy, and we investigated the value of assessment of RVSF in predicting postoperative outcome.
Methods
This study was approved by the Indiana University Institutional Review Board. A total of 35 patients with CP who underwent pericardiectomy from 1999 to October 2010 at the Indiana University Health hospitals were identified through surgical records. The diagnosis of CP was based on clinical features, cardiac catheterization, echocardiography, and/or magnetic resonance imaging in accordance with the already published guidelines. All patients had proven pericardial fibrosis and calcification on pathologic examination.
The echocardiography database was used to identify the echocardiogram performed closest to the time of pericardiectomy in each patient. Left-sided cardiac 2-dimensional echocardiography and Doppler measurements performed at the time of each examination by trained sonographers were used for this study. An investigator, blinded to the outcome data, measured right atrial minor axis dimension, RV basal dimension (RVD1), and fractional area change (FAC) according to previously published guidelines from digitally stored images archived on a commercially available system (Fuji Synapse 4.0, Indianapolis, Indiana). A blinded investigator measured TAPSE from the 2-dimensional echocardiograms. In the 4-chamber view, the lateral tricuspid annulus at end-diastole was identified, and the full extent of excursion was measured using an off-line analysis program that enabled measurement of distance along the exact direction of excursion of the annulus toward the apex. In our laboratory, the 2-dimensional method of measuring TAPSE has been shown to be highly correlated with time velocity integral (TVI) of the S velocity of the lateral tricuspid annulus obtained by tissue Doppler (r = 0.91 and mean % difference of 10% [difference between TAPSE and TVI measurement/mean value of TAPSE + TVI]). For example, for TAPSE = 2, and TVI = 2.2, the mean % difference = 2.2 − 2.0/2.1 = 9.5%. In this study, TVI was also measured in 17 subjects who had tissue Doppler assessment of RV systolic function. In our laboratory, the normal mean value for TAPSE in subjects with no cardiac disease is 2.54 ± 0.34 cm. TAPSE of ≤1.8 cm (2 SD less than the mean) was considered significantly reduced.
Follow-up was obtained by review of hospital medical records. Postoperative congestive heart failure and inhospital mortality were considered as significant events. One patient who had a second major surgical procedure during the same admission for pericardiectomy was censored at the time of the second procedure. Heart failure was defined by the presence of pulmonary edema on chest X-ray and the requirement for diuretic therapy for treatment. Heart failure was also defined as hypotension because of low cardiac output and requiring intravenous inotropic or vasopressor support.
Two-tailed t tests were used to compare continuous variables, and the chi-square test was used for comparison of categorical variables. A p value ≤0.05 was considered significant. Binary logistic regression was used to identify univariate predictors of cardiac events. Multivariate analysis (forward likelihood method) was performed using variables with p ≤0.10 on univariate analysis. Stepwise multivariate analysis was performed, using variables with p ≤0.10 on univariate analysis in the first block and TAPSE was added in a second step to determine if RVSF provided any incremental prognostic information. Receiver-operating characteristic curve analysis was conducted to determine the sensitivity, specificity, and optimal threshold value of TAPSE for prediction of adverse cardiovascular events. Linear regression analysis was performed to identify the relation of TAPSE and length of hospital stay after pericardiectomy in those who survived to discharge. SPSS, version 22.0, for Windows was used for data analysis (SPSS Inc., Chicago, Illinois).
Results
The mean age was 52 ± 15.4 years. The mean left ventricular ejection fraction (LVEF) was 55 ± 9%. Five of 35 patients (14%) were women. Hypertension was present in 14 (40%), diabetes mellitus in 10 (29%), and chronic kidney disease in 8 (23%). A history of congestive heart failure was present in 3 (8%) and coronary artery disease in 11 (31%). The possible causes of CP in our patient population are listed in Table 1 . Five patients (14%) had concomitant coronary artery bypass grafting (CABG) at the time of pericardiectomy. Cardiopulmonary bypass was used in 10 patients (29%). All patients underwent complete pericardiectomy.
| Causes | Number of patients (n=35) |
|---|---|
| Idiopathic/Viral | 19 (54%) |
| Uremia | 5 (14%) |
| Post Cardiac Surgery | 4 (11%) |
| Post Radiation | 3 (9%) |
| Trauma | 2 (6%) |
| Post Infectious | 2 (6%) |
Preoperatively, TAPSE was successfully measured in all patients. The mean TAPSE was 1.4 ± 0.4 cm (range 0.6 to 2.3 cm). Mean TVI was 1.42 ± 0.44 cm in 17 subjects with tissue Doppler assessment. The mean % difference between TAPSE and TVI in the subjects with both measurements was 9%. Thirty-one patients (88.6%) had reduced RV systolic function (TAPSE of ≤1.8 cm).
Eight of 35 patients (23%) had postoperative events. Three had severe but nonfatal heart failure. All 3 required prolonged inotropic/vasopressor support because of hypotension and low cardiac output and 1 also had respiratory failure requiring temporary re-intubation. The 5 remaining patients with events died. One died intraoperatively from persistent shock after pericardiectomy and the remaining 4 patients had refractory heart failure postoperatively and died despite aggressive hemodynamic and ventilatory support. In 7 of 8 patients with events, preoperative TAPSE was severely reduced (<1.4 cm, 3 SDs less than the mean). In the remaining patient, preoperative TAPSE was mildly reduced (1.8 cm) and worsened postoperatively to 1.17 cm before death occurred. LVEF was normal in 6 patients with events, mildly reduced (45%) in 1, and severely reduced in the remaining patient. Two patients with events had concomitant CABG.
The clinical and echocardiographic parameters of patients who did or did not have cardiovascular events after pericardiectomy are compared in Table 2 . Variables that were borderline or significant univariate predictors of events included concomitant CABG (p = 0.052), left atrial (LA) diameter (p = 0.028), LVEF (p = 0.059), and TAPSE (p = 0.016). RV FAC (p = 0.068) tended to be lower in patients with events. RV FAC could be measured in only 26 patients, and because of missing data, this variable was excluded from multivariate analysis. On multivariate analysis, TAPSE (p = 0.018) and the need for CABG (p = 0.033) were shown to be independent predictors of postoperative cardiovascular outcomes. The odds ratio for TAPSE expressed in millimeters was 0.605 (0.40 to 0.92) and for CABG was 20 (1.26 to 315). Hence, for each millimeter increase in TAPSE, the odds of an event decrease by almost 40%. Stepwise multivariate analysis showed that TAPSE added incremental prognostic value to the combination of CABG, LVEF, and LA diameter (p = 0.029, chi-square increase 11.6 to 16.3, Figure 1 ). The addition of TAPSE to the other variables improved the classification of those with and without events from 82% to 94%.
| Variables | No CV Events (n=27) | CV Events (n=8) | p value | OR (95% CI) |
|---|---|---|---|---|
| Age (years) | 54.1 ± 14.8 | 47.1± 17.2 | 0.262 | |
| Male gender | 22(81%) | 8(100%) | 0.999 | |
| Hypertension | 9(33%) | 5(62%) | 0.150 | |
| Diabetes mellitus | 2(7%) | 2(25%) | 0.799 | |
| Chronic kidney disease | 6(22%) | 2(25%) | 0.870 | |
| Prior heart failure | 3(11%) | 0(0%) | 0.999 | |
| Prior coronary artery disease | 8(30%) | 3(37%) | 0.939 | |
| Prior therapeutic radiation | 3(11%) | 0(0%) | 0.999 | |
| Cirrhosis | 4(15%) | 2(25%) | 0.507 | |
| Coronary artery bypass graft | 2(7%) | 3(37%) | 0.052 | 7.5(0.98-57.1) |
| Cardiac valve surgery | 1(4%) | 0(0%) | 1.000 | |
| Cardiopulmonary bypass | 7(26%) | 3(37%) | 0.527 | |
| Left ventricle ejection fraction (%) | 57±4.13 | 48±16.1 | 0.059 | 0.89(0.8-1.0) |
| Left ventricle end-diastolic dimension (mm) | 4.5±1.57 | 4.3±0.66 | 0.802 | |
| Left ventricle end-systolic dimension (mm) | 2.8±0.7 | 3.4±0.73 | 0.091 | |
| Left atrial diameter (cm) | 3.9±0.64 | 4.6±0.86 | 0.028 | 4.69(1.18-18.6) |
| Deceleration Time (sec) | 182±57 | 152±51 | 0.201 | |
| Tricuspid annular plane systolic excursion (cm) | 1.51±0.35 | 1.08±0.37 | 0.016 | 0.01(0.00-0.46) |
| Right ventricle fractional area change | 0.44±0.1 | 0.37±0.1 | 0.068 | 0.00(0.00-2.1) |
| Right ventricle Diameter base (cm) | 3.4±0.7 | 3.5±0.5 | 0.679 | |
| Right atrial Diameter (cm) | 4.19±0.72 | 4.15±0.68 | 0.874 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
