Elevated B-type natriuretic peptide (BNP) is a marker of poor outcomes in heart failure, acute coronary syndromes, and sepsis. Elevated cardiac troponin I (cTnI) is associated with adverse outcomes in infective endocarditis. It was hypothesized that elevated BNP would be associated with increased rates of morbidity and mortality in patients with infective endocarditis, particularly when combined with elevated cTnI. Consecutively enrolled patients in the International Collaboration on Endocarditis Prospective Cohort Study (ICE-PCS) were evaluated at a single center. The association between elevated BNP and a composite outcome of death, intracardiac abscess, and central nervous system event and the individual components of the composite was determined. Similar analyses were performed in patients who had BNP and cTnI measured. Of 103 patients, 45 had BNP measured for clinical indications. The median BNP level was higher in patients with the composite outcome (1,498 vs 433 pg/ml, p = 0.03) and in those who died (2,150 vs 628 pg/ml, p = 0.04). Elevated BNP was significantly associated with the composite outcome (p <0.01) and intracardiac abscess (p = 0.02). Patients with elevation of BNP and cTnI had a significantly higher probability of the composite outcome (69%) than patients with either BNP or cTnI elevated (29%) or neither BNP nor troponin elevated (0%) (p for trend <0.01). In conclusion, these data demonstrate a significant association between elevated BNP alone and in combination with cTnI for serious outcomes in infective endocarditis and warrant prospective evaluation.
Despite modern medical and surgical therapies, infective endocarditis (IE) remains a disease with considerable morbidity and mortality. Identifying patients at increased risk for adverse outcomes is complicated by the broad spectrum of cardiac pathology and infecting organisms. Cardiac biomarkers may facilitate diagnosis and risk stratification in this heterogenous disease. For example, several studies have demonstrated that elevated cardiac troponins are associated with patient morbidity and mortality in IE. In contrast, less is known about the prognostic value of B-type natriuretic peptide (BNP) in IE. BNP is a hormone synthesized and secreted primarily by ventricular cardiomyocytes in response to increased myocardial stretch and volume overload. Circulating BNP levels appear to serve as an integrative marker of multiple pathologic insults, and elevated levels are predictive of poor outcomes in multiple disease states, including heart failure, acute coronary syndromes, and septic shock. We hypothesized that elevated BNP levels in patients with IE would also be associated with poor clinical outcomes, particularly when accompanied by elevation in cardiac troponin I (cTnI).
Methods
Patients were consecutively enrolled at a single center as a part of the International Collaboration on Endocarditis Prospective Cohort Study (ICE-PCS) from December 2002 to June 2007. The study protocol was approved by the institutional review board at the University of Texas Southwestern Medical Center, and informed consent was obtained. Patient data were retrospectively examined, including BNP and cTnI measurements if performed. All BNP and cTnI laboratories were drawn for clinical indications before cardiac surgery without study investigator input. BNP elevation was prespecified as ≥400 pg/ml and cTnI elevation was prespecified as ≥0.1 ng/ml on the basis of the laboratory reporting criteria at our institution. Echocardiograms were read by a level III trained physician, and echocardiographic characteristics were based on the clinical interpretation. Left ventricular systolic dysfunction was defined as an ejection fraction <50% during the index hospitalization. Coronary artery disease was defined as a history of myocardial infarction or documentation of coronary artery disease in the medical record. End-stage renal disease was defined as the need for hemodialysis. The predefined primary outcome was a composite of in-hospital mortality, presence of an annular or myocardial abscess reported on echocardiographic interpretation or by surgical findings, and the occurrence of a central nervous system (CNS) event, defined by new, focal deficits and the presence of abnormal results on computed tomography or magnetic resonance imaging. The individual components of the composite outcome constituted secondary outcomes.
Characteristics were compared between patients with IE in whom BNP was and was not measured. In patients with BNP measurements, characteristics and outcomes were compared between those with and without elevated levels. Stratified analyses were performed on the basis of the presence or absence of left ventricular systolic dysfunction and severe aortic and/or mitral valve regurgitation. Clinical outcomes based on the absolute BNP level were also assessed. In patients with BNP and cTnI measured, outcomes were compared between patients who had neither, 1, or both biomarkers elevated. Fisher’s exact, chi-square, and Mann-Whitney tests were used as appropriate. Two-sided p values <0.05 were considered statistically significant. All analyses were performed using GraphPad Prism 5 software (GraphPad Software Inc., La Jolla, California).
Results
Of 103 IE patients enrolled, 45 had BNP measured for clinical indications. Baseline characteristics of enrolled patients are presented in Table 1 . Patients who had BNP measured were of similar ages and had similar underlying medical conditions, including a history of heart failure and end-stage renal disease, compared to patients who did not have BNP measured. Patients were less likely to have BNP measured if they were intravenous drug users. Of the 45 patients with BNP measured, 32 (71%) had levels greater than or equal to the prespecified threshold of 400 pg/ml ( Table 2 ). Patients with BNP ≥400 pg/ml were less likely to have isolated right-sided IE (p <0.01) and more likely to have severe aortic and/or mitral valve regurgitation (p = 0.02). There was nonsignificant a trend toward depressed left ventricular systolic function in patients with elevated BNP (p = 0.07).
| Characteristic | BNP Measurement | ||
|---|---|---|---|
| Yes (n = 45) | No (n = 58) | p Value | |
| Mean age (years) | 49 | 45.2 | 0.12 |
| Men | 31 (69%) | 41 (71%) | 1.00 |
| Nonwhite | 32 (71%) | 38 (66%) | 0.67 |
| History of heart failure | 11 (25%) | 11 (20%) | 0.63 |
| Coronary artery disease | 5 (11%) | 1 (2%) | 0.08 |
| Diabetes mellitus | 15 (33%) | 12 (21%) | 0.18 |
| End-stage renal disease | 4 (9%) | 7 (12%) | 0.75 |
| Intravenous drug use | 8 (19%) | 28 (48%) | 0.00 |
| Right-sided infective endocarditis only | 6 (13%) | 11 (19%) | 0.59 |
| Prosthetic valve infective endocarditis | 5 (11%) | 11 (19%) | 0.41 |
| Staphylococcus aureus infection | 18 (40%) | 28 (49%) | 0.43 |
| Depressed left ventricular systolic function | 13 (29%) | 17 (29%) | 1.00 |
| Severe aortic or mitral valve regurgitation | 20 (44%) | 15 (26%) | 0.06 |
| Mean creatinine ⁎ (mg/dl) | 3.05 | 3.16 | 0.87 |
| Characteristic | BNP (pg/ml) | ||
|---|---|---|---|
| <400 (n = 13) | ≥400 (n = 32) | p Value | |
| Mean age (years) | 46 | 50 | 0.33 |
| Men | 8 (62%) | 23 (72%) | 0.50 |
| Nonwhite | 9 (69%) | 23 (72%) | 1.00 |
| History of heart failure | 4 (33%) | 7 (22%) | 0.46 |
| Coronary artery disease | 0 (0%) | 5 (16%) | 0.30 |
| Diabetes mellitus | 3 (23%) | 12 (38%) | 0.49 |
| End-stage renal disease | 0 (0%) | 4 (13%) | 0.31 |
| Intravenous drug use | 3 (23%) | 5 (17%) | 0.68 |
| Right-sided infective endocarditis only | 5 (38%) | 1 (3%) | <0.01 |
| Prosthetic valve infective endocarditis | 2 (15%) | 3 (9%) | 0.62 |
| Staphylococcus aureus infection | 6 (46%) | 12 (38%) | 0.74 |
| Depressed left ventricular systolic function | 1 (8%) | 12 (38%) | 0.07 |
| Severe aortic or mitral valve regurgitation | 2 (15%) | 18 (56%) | 0.02 |
| Mean creatinine ⁎ (mg/dl) | 1.37 | 2.19 | 0.21 |
The median BNP level was 1,498 pg/ml (interquartile range [IQR] 574 to 2,294) in patients who developed the composite outcome of death, intracardiac abscess, or CNS event and 433 pg/ml (IQR 116 to 1,573) in those who did not (p = 0.03; Figure 1 ). The median BNP level in patients who died in the hospital was 2,150 pg/ml (IQR 673 to 3,164) and 628 pg/ml (IQR 184 to 1,689) in patients who survived to discharge (p = 0.04; Figure 1 ).
A BNP level ≥400 pg/ml was associated with 4.1-fold increase in the composite outcome (63% vs 15%, p = 0.01; Figure 2 ) and was significantly associated with the individual outcome of intracardiac abscess (p = 0.02). When examining death and CNS events, there was a nonsignificant trend between the outcomes and elevated BNP ( Figure 2 ). A significant association of BNP elevation with the composite outcome was maintained when the analyses were restricted to patients with normal left ventricular systolic function (n = 32; 65% vs 17%, p = 0.01; Figure 3 ) and patients without severe aortic or mitral regurgitation (n = 25; 64% vs 18%, p = 0.04).
Of the 45 patients with IE who had BNP measured, 40 (89%) also had cTnI measured. Twenty-six patients (65%) had BNP and cTnI elevated, 7 patients (17.5%) had either BNP or cTnI elevated, and 7 patients (17.5%) had neither BNP nor cTnI elevated. Patients with BNP and cTnI elevated had a significantly higher probability of meeting the composite outcome (69%) than patients with either BNP or cTnI elevated (29%) or neither BNP nor cTnI elevated (0%) (p for trend <0.01; Figure 4 ). Patients with BNP and cTnI elevated were more likely to reach the individual end points of death (p for trend <0.05), intracardiac abscess (p for trend <0.01), and CNS event (p for trend <0.05) than patients with only BNP or cTnI elevated or neither elevated ( Figure 4 ).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
