Author + information
- Luciano Babuin, MD,
- Jorge R. Alegria, MD,
- Jae K. Oh, MD,
- Rick A. Nishimura, MD and
- Allan S. Jaffe, MD⁎ ()
- ↵⁎Cardiovascular Division, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905
To the Editor:Constrictive pericarditis (CP) often is a challenging diagnosis despite sophisticated modalities (1–4). If a simple test could alert physicians to the possibility of CP in patients with evidence of heart failure, it would be valuable.
Recent data (5) suggest that brain natriuretic peptide (BNP), released in response to cardiac stretch (6), is normal or only minimally elevated in patients with CP and more elevated in patients with restrictive cardiomyopathy (RCMP). However, there are a variety of etiologies for CP. Many patients at our institution present after either mantle chest radiation and/or open heart surgical procedures (7). These patients may have additional myocardial damage that could raise BNP even when concomitant CP is present. To determine whether BNP levels in all patients with CP are different from those with restriction, we studied patients with well-documented CP and RCMP. We separated patients with CP into those with idiopathic disease and with CP due to secondary causes.
We reviewed 22 patients with surgically confirmed CP where BNP was measured pre-operatively. Patients were divided into those with idiopathic CP or secondary CP if there had been previous cardiac surgery or chest radiation. We identified patients with RCMP who had appropriately timed BNP values as a comparison group. The diagnosis of RCMP was made according to criteria by Hurrell et al. (4). These patients also had characteristic echocardiographic findings. The diagnosis of CP was confirmed at surgery.
Two-dimensional echocardiographic examinations were performed as previous reported (8); BNP measurements were performed with the Biosite assay (La Jolla, California) close to the time of the echocardiographic examinations and before surgical pericardiectomy. The interval between echocardiogram and BNP measurement was 1.3 ± 7.4 days, and in all patients except 6, the interval was 1 day.
Summary statistics are presented as frequencies (percentages) for categorical variables, as mean ± SD for the normal variables, and as median and first and third quartile (Q1, Q3) for skewed variables. The normal distribution of the variables was verified performing the Shapiro-Wilk W test. Only BNP was very right skewed. Log10-transformed values of BNP (Log10BNP) were used in the analysis to obtain a normal distribution of the values. Categorical variables were compared by Fisher exact test, continuous normal variables by analysis of variance (followed by Tukey’s HDS post-test). All tests were two-sided, and for all analyses p < 0.05 was considered statistically significant. Data were analyzed with JMP Version 5 (SAS Institute Inc., Cary, North Carolina).
Twenty-two patients had surgically confirmed CP, 11 patients had idiopathic CP, and 11 secondary CP (8 with previous cardiac surgery and 3 with radiation therapy). They were compared to 11 patients with RCMP. Patient characteristics are reported in Table 1.There were no significant differences among groups. Median BNP was 80 (44 to 193) ng/l for idiopathic CP, 278 (118 to 526) ng/l for secondary CP, and 499 (361 to 606) ng/l for RCMP. The Log10-BNP was 1.9 ± 0.3 ng/l for idiopathic CP, 2.4 ± 0.3 ng/l for secondary CP, and 2.7 ± 0.2 ng/l for RCMP. There was a statistically significant difference between the BNP values with idiopathic CP and RCMP (p < 0.05) and between patients with idiopathic CP and those with secondary CP (p < 0.05). There were no significant differences in BNP between patients with secondary CP and RCMP (Fig. 1).
All patients with BNP <150 ng/l had CP. Higher BNPs were seen in patients with secondary CP and RCMP. The only differentiating finding with RCMP was a BNP >650 ng/l, which did not occur with CP.
Our data confirm and extend those recently published (5). Brain natriuretic peptide levels were significantly lower in patients with idiopathic CP than those with RCMP. However, BNP is not significantly different in patients with CP secondary to previous cardiac surgery or radiation compared to those with RCMP. Thus, BNP is a useful non-invasive marker to distinguish CP from RCMP when CP is not due to a secondary cause. Relatively normal BNP values in the setting of increased jugular venous pressure should point to a diagnostic possibility of CP.
The findings of higher BNP values in secondary constriction are important given in the modern era, previous cardiac surgery and previous chest radiation therapy are more prevalent causes (7). In these situations, higher values cannot be used to distinguish CP from RCMP.
These data are consistent with the hypothesis proposed by Leya et al. (5) that in patients with CP the myocardium may be “intrinsically normal and myocardial stretch may be prevented by the constraining pericardium.” This assumption is less valid in patients with a history of additional cardiac disease. In these patients, myocardial abnormalities may result in a higher release of BNP than those with idiopathic CP. This distinction may have been obscured in the prior study by the small number of patients (six with CP and five with RCMP). Only three of the six patients with the diagnosis of CP had a history of prior surgery, and those patients with “history of radiation-induced cardiomyopathy” were excluded.
The results of our study are important in clinical practice. The differentiation of CP from RCMP can be more readily made in cases of idiopathic CP versus pure RCMP. However, in those cases of mixed myocardial and pericardial disease, the differential diagnosis becomes more difficult.
Our study was a retrospective study with a modest-sized cohort of non-consecutive patients. Hemodynamic data taken proximate to the BNP values were not uniformly available.
In conclusion, BNP levels are significantly lower in patients with idiopathic CP, compared to those with CP after surgery and/or radiation and RCMP. Relatively normal BNP levels in patients with right-sided heart failure should raise the suspicion of CP. Higher levels cannot distinguish between CP and RCMP, especially in patients with secondary CP.
Please note: Dr. Jaffe consults for many companies that make BNP assays and has both research grants and consultantships with them, including Roche, Dade, and Beckman. In addition, he is a consultant for Abbott, Ortho, DiaDexus, and Sensera, all of whom make BNP assays. Of note, the BNP assay used in this study was not made by any of those companies. Drs. Babuin and Alegria contributed equally to this report.
- American College of Cardiology Foundation
- Nishimura R.A.
- Hurrell D.G.,
- Nishimura R.A.,
- Higano S.T.,
- et al.
- Leya F.S.,
- Arab D.,
- Joyal D.,
- et al.
- Ling L.H.,
- Oh J.K.,
- Schaff H.V.,
- et al.
- Oh J.,
- Seward J.,
- Tajik A.