Prognostic value of cardiac troponin levels in infective endocarditis

Recent analyses suggest that early surgery is associated with improved 6-month mortality in infective endocarditis (IE), emphasizing the need to identify patients who would benefit from early intervention. Although well established as a marker for myocardial injury and increased mortality in the setting of acute coronary syndromes [Heidenreich et al. 2001], cardiac-specific troponins (Tn) have also been found to associate with clinical outcomes in other noncardiac diseases [Dispenzieri et al. 2004]. Isolated reports suggest that elevated cardiac troponin I (TnI) and/or T (TnT) serum concentrations associate with poor prognosis in patients with IE. We performed a systematic review of this subject. An electronic literature search of the National Library of Medicine database using PubMed (all listings through 30 September 2012) was performed using the search terms ‘endocarditis’ and ‘troponin’, limited to human studies and English language articles. After excluding case reports (four) and letters to the editor (one), six original reports remained for inclusion in this analysis (Table 1).

Table 1.

Summary of studies evaluating the relationship between cardiac markers and infective endocarditis.

	Study design	Marker(s)	Mean age (years)	Sample size eligible for analysis	Deaths in sample (number) (%)	Number in cohort (%)with elevated Tn	Excluded predisposing conditions *
Watkin et al. 2004	Retrospective	TnI	51	15	2(13)	11(73)	Yes
Purcell et al. 2008	Retrospective	TnI	47	51	(26)	33(64)	No
Tsenovoy et al. 2009	Prospective	TnI	60	62	15(24)	35(57)	Yes
Stancoven et al. 2011	Retrospective	TnT, hs TnT	51	42	4(10)	25(56), 39(93)	No
Shiue et al. 2010	Retrospective	TnI, BNP	49	40, 45	NR	NR, 32(71)	No
Kahveci et al. 2007	Prospective	TnI, NT-pro-BNP	41 $	38, 45	9(20)	NR	Yes

*

Conditions that may result in cardiac troponin release.

$

Reported median age of study population, range 9–75 years.

BNP, brain natriuretic peptide; hs-TnT, highly sensitive cardiac troponin T; NR- not reported; NT-pro-BNP, N-terminal pro-B-type natriuretic peptide; TnI, cardiac troponin I; TnT, cardiac troponin T

The reviewed reports suggest Tn elevation commonly occurs in the setting of IE. Although inclusion criteria and assays differed between studies, the majority of patients (57–73%) with IE demonstrated elevated concentrations of cardiac markers. The initial investigation by Watkin and colleagues reported that 2/15 patients in his sample died of septic shock, both of whom had TnI ≥ 0.03 ng/ml [Watkin et al. 2004]. The five studies that followed demonstrated that elevated levels of cardiac biomarkers (TnI, highly sensitivity cardiac troponin T, brain natriuretic peptide, N-terminal pro-B-type natriuretic peptide), associated with mortality, either as an individual primary endpoint and/or as a component of a composite outcome. However, the lack of Tn elevation in IE might correlate better with outcomes than detection of elevated serum levels. In the studies that evaluated negative predictive value (NPV), normal Tn levels had an NPV as high as 94% for mortality, myocardial abscess, and central nervous system events [Purcell et al. 2008].

Studies in this analysis were uniformly limited by small sample sizes and single-center designs. Three studies originated from the same institution [Purcell et al. 2008; Shiue et al. 2010; Stancoven et al. 2011], with potential overlap in patient population for two of the reports [Purcell et al. 2008; Shiue et al. 2010]. Prospective multicenter studies would be needed to determine if a clinically relevant, sufficiently robust relationship exists between cardiac biomarkers and major clinical outcomes in patients with IE.