© 2015, Wiley Periodicals, Inc. DOI: 10.1111/echo.12909
Echocardiography
Prognostic Value of Pericardial Effusion on Serial Echocardiograms in Pulmonary Arterial Hypertension Omar Batal, M.D.,* Zeina Dardari, M.S.,† Chelcie Costabile, D.O.,* John Gorcsan, M.D.,* Vincent C. Arena, Ph.D.,† and Michael A. Mathier, M.D.* *Division of Cardiology, UPMC, Heart and Vascular Institute, Pittsburgh, Pennsylvania; and †Department of Statistics, University of Pittsburgh, Pittsburgh, Pennsylvania
Background: Pericardial effusion in pulmonary arterial hypertension (PAH) is an indicator of right heart failure and poor prognosis; its significance on serial transthoracic echocardiograms (TTE) is not clear. Methods: Baseline and follow-up TTE (1.0 0.5 years), clinical parameters, and outcomes were studied (N = 200) in consecutive patients with PAH who underwent TTE at our center between October 1999 and November 2007. Study baseline TTE was 2.8 4.0 years from initial PAH diagnosis. Results: Over median follow-up of 3.6 2.6 years from baseline TTE, 106 patients (53%) died. Pericardial effusion was present in 20% at baseline, and at any time during the study in 29%. Patients with any pericardial effusion during follow-up were more likely to have underlying connective tissue disease. They also had significantly higher mean right atrial pressure and pulmonary vascular resistance, had lower cardiac output by invasive hemodynamic studies, had higher serum creatinine, and were more likely to be treated with prostanoids. Patients were also significantly likely to have more echocardiographic right atrial dilation and right ventricular dilation and dysfunction, and worse tricuspid regurgitation with higher peak velocity. During follow-up, there was significantly increased use of prostanoids (58% vs. 28%) and combination therapy (8% vs. 2%) compared to baseline. Persistence of pericardial effusion on both baseline and follow-up TTE was associated with worse outcome, and an independent predictor of survival after adjusting for age, creatinine, functional class, and hemodynamics (P < 0.01). Conclusion: Persistence of pericardial effusion in PAH despite vasoactive therapy predicts worse outcomes; absence or resolution of pericardial effusion with therapy suggests better prognosis. (Echocardiography 2015;32:1471–1476) Key words: pulmonary arterial hypertension, pericardial effusion, survival, echocardiography
Pulmonary arterial hypertension (PAH) is a disease caused by a pathologic increase in pulmonary vascular resistance leading to progressive right ventricular (RV) failure and death.1 Pericardial effusion in the setting of PAH has been described as associated with worse right-sided cardiac hemodynamics and RV decompensation, and is a predictor of worse outcomes.2–12 Data about the significance of pericardial effusion during follow-up is limited.8 Furthermore, the significance of resolution versus persistence of pericardial effusion on serial echocardiograms has not been described. We hypothesized that the presence or persistence of pericardial effusion during follow-up is associated with RV dysfunction and decreased survival. Address for correspondence and reprint requests: Michael A. Mathier, M.D., Division of Cardiology, UPMC, Heart and Vascular Institute, 200 Lothrop Street, Scaife Hall, Room S-559 Pittsburgh, PA 15213. Fax: 412-647-0595; E-mail:
[email protected] Methods: We tested the hypothesis that pericardial effusion on serial transthoracic echocardiogram (TTE) was associated with poor outcome in PAH. We studied consecutive patients with PAH who underwent a TTE at our center between October 1999 and November 2007. The database was then constructed by a comprehensive review of patient medical records. Of 359 patients, 89 were excluded due to not having PAH13 and 73 were excluded due to not having a follow-up TTE in our system. Data analyzed includes demographics, PAH etiology, medical comorbidities, PAH medications, New York Heart Association functional class (NYHA), TTE variables, hemodynamics, and serum creatinine and sodium. The initial TTE in this study was performed an average of 2.8 4.0 years from the time of diagnosis of PAH (approximately 33% were at the time of diagnosis) and was considered to be the study baseline TTE (TTE1). The follow-up TTE (TTE2) was performed 1.0 0.5 years from 1471
Batal, et al.
TTE1. A pericardial effusion was considered significant if it persists throughout the cardiac cycle and was graded as small if it measures 2 cm. The observer who determined pericardial effusion size was blinded to the clinical variables. A trivial pericardial effusion was defined as a visible separation of pericardial layers that persisted during the cardiac cycle.3 Patients were grouped according to size of pericardial effusion as no pericardial effusion (or only trivial pericardial effusion) versus presence of a pericardial effusion (≥small in size). As a trivial pericardial effusion may occasionally be seen in otherwise normal pericardium,14,15 in this study, a pericardial effusion refers to one that is ≥small in size unless stated. Right atrial and ventricular size by echocardiography was estimated qualitatively and also measured quantitatively using ASE guidelines.16 Right heart hemodynamics were obtained 0.5 0.7 years from TTE1 (available for 85% of patients). Categorical variables were summarized as frequency and percentage and compared for statistical significance using Fisher’s exact test or chisquare as appropriate. Continuous variables were summarized as mean SD, and comparison for statistical significance among groups was performed using the 2-sample t-test or Kruskal– Wallis test as appropriate. The prognostic significance of pericardial effusion was analyzed by separating patients into 4 groups: group 1 had
no (or only trivial) pericardial effusion at TTE1 and TTE2; group 2 had a pericardial effusion (≥small in size) at TTE1 that resolves at the time of TTE2; group 3 did not have a pericardial effusion at TTE1 but developed an effusion at TTE2; and group 4 had a pericardial effusion at TTE1 that persisted at TTE2. For time-based analysis, time is from the initial study TTE rather than time from initial diagnosis. Kaplan–Meier survival analysis was performed for groups 1–4, and a Cox proportional hazard model was used to adjust for confounder variables (predefined as age, right heart hemodynamics, NYHA, PAH etiology, and serum creatinine).1 A P-value < 0.05 was considered statistically significant. Results: Patient characteristics are summarized in Table 1. The etiology of PAH was idiopathic in 47%, connective tissue disease in 33%, congenital heart disease in 12%, cirrhosis in 7%, and related to anorexigen use in 1%. Patients with PAH due to connective tissue disease were significantly more likely to have a pericardial effusion during the study period (at time of TTE1 or TTE2). The following comorbidities were present: coronary artery disease in 14%, diabetes in 15%, hyperlipidemia in 25%, hypertension in 15%, and a history of smoking in 46%; these comorbidities were not significantly associated with pericardial effusion. PAH medications at the time of TTE2 included prostanoids (58%), phosphodiesterase
TABLE 1 Patient Characteristics
Variable
All Patients (N = 200)
Age (years) Female (%) Caucasian (%) Body mass index (kg/m2) Creatinine (mg/dL) NYHA at TTE1 NYHA at TTE2 Idiopathic pulmonary arterial hypertension (%) Connective tissue disease (%) Mean right atrial pressure (mmHg) Mean pulmonary artery pressure (mmHg) Fick cardiac output (L/min) Pulmonary vascular resistance (Wood units)
54 71 91 29 1.1 2.5 2.4 47 33 10 48 5.1 8
15
9 0.6 0.7 0.8
8 16 1.9 6
No Pericardial Effusion at TTE1 or TTE2 (N = 142) 54 68 92 29 1.0 2.4 2.4 49 26 9.2 48 5.4 7
15
9 0.5 0.7 0.7
8.8 17 1.9 6
Pericardial Effusion Present at TTE1 or TTE2 (N = 58) 55 76 89 30 1.3 2.5 2.5 43 50 11 51 4.6 11
13
10 0.7 0.8 0.9
6 14 1.8 7
P-Value* 0.7 0.3 0.4 0.6 0.0003 0.8 0.2 0.2 0.001 0.007 0.2 0.006 0.0003
*When the analysis was repeated at time point of TTE1 for the presence or absence of pericardial effusion, the results were similar to above except that the difference in mean right atrial pressure did not achieve statistical significance. At the time point of TTE2, the results were again similar but the difference in creatinine, mean right atrial pressure, and Fick cardiac output did not achieve statistical significance; however, idiopathic pulmonary arterial hypertension was associated with less frequent pericardial effusion (P = 0.04). NYHA = New York Heart Association; TTE = transthoracic echocardiogram.
1472
Pericardial Effusion in PAH
V inhibitor (20%), endothelin receptor antagonist (52%), and combination therapy (8%). Pericardial effusion was noted in 20% of patients at the time of TTE1, 22% of patients at TTE2, and 29% of patients at TTE1 and/ or TTE2. The frequency of pericardial effusions at TTE1 and/or TTE2 is summarized in Table 2. Patients with a pericardial effusion at any time during the study (i.e., at TTE1 and/or TTE2) were more likely to be treated with a prostanoid than patients with absent pericardial effusion at both time points (60% vs. 34%, P = 0.001). Patients with a pericardial effusion during the study were found to have significantly worse right heart hemodynamics (Table 1). Patients who had a pericardial effusion at TTE2 were significantly more likely to have moderately severe or severe right atrial and ventricular dilation and RV dysfunction and were also more likely to have more severe tricuspid regurgitation and a higher
tricuspid valve regurgitation peak velocity on TTE2 (Table 3). In our study population, only 2 patients underwent pericardiocentesis. One patient had severe PAH secondary to systemic sclerosis, on epoprostenol and presented electively for pericardiocentesis of a large effusion. After 670 mL of fluid was removed, she developed respiratory distress and hypotension requiring brief intubation and inotropic support (dobutamine). She was discharged after a week-long hospitalization, and repeat TTE2 weeks later showed no pericardial effusion. She was alive at the last point of contact within our hospital system 5 years later. The second patient also had scleroderma and PAH. After admission to the hospital with rapid atrial fibrillation, a large pericardial effusion with evidence of increased pericardial pressures was detected, and consequently, 1100 mL of fluid was removed and another 830 mL was drained
TABLE 2 Size of Pericardial Effusion Group 1 (N = 142)
Group 2 (N = 15)
Group 3 (N = 18)
Group 4 (N = 25)
Transthoracic Echocardiogram Size of Pericardial Effusion None (N)* Small (N) Moderate (N) Large (N)
1
2
1
2
1
2
1
2
142 0 0 0
142 0 0 0
0 13 1 1
15 0 0 0
18 0 0 0
0 16 0 2
0 19 2 4
0 17 1 7
*In our study analysis, we did not distinguish between patients with a trivial versus no pericardial effusion. To note, at TTE1, there were 15 patients with a trivial pericardial effusion. At TTE2, there were 12 patients with a trivial pericardial effusion. TTE = transthoracic echocardiogram.
TABLE 3 Pericardial Effusion and Right Heart Size and Function Parameters on Follow-Up TTE Pericardial Effusion Absent (N = 157) % >Moderate right atrial dilation Right atrial area (cm2) % >Moderate right ventricular dilation Mid-right ventricular diameter (cm) Mid-right ventricular to left ventricular ratio % >Moderate right ventricular dysfunction Tricuspid annular plane systolic excursion (cm) Tei Index Right ventricular fractional area change (%) % >Moderate tricuspid regurgitation % Plethoric inferior vena cava* Tricuspid regurgitation velocity (m/sec)
20% 21 21% 4.0 1.0 14% 1.6 0.38 30 27% 27% 3.9
Pericardial Effusion Present (N = 43)
9 1.1 0.5 0.6 0.30 13
0.8
40% 25 60% 4.8 1.4 45% 1.4 0.45 27 67% 41% 4.5
8 1.2 0.6 0.4 0.28 12
0.7
P-Value