Pediatric Infective Endocarditis Over the Last 30 Years: Thromboembolic Complications and Associations of Valve Vegetation and Heart Failure with Increased Mortality

Pediatric infective endocarditis (IE) continues to be a condition associated with high morbidity and mortality. Contributing factors for IE-related morbidity include thromboembolic complications (TEC) (cerebrovascular events (CVE): acute ischemic stroke (AIS); non-CVEs: arterial/venous TEC outside of the central nervous system). Conversely, reported mortality risk factors include age, gender, vegetation size, and TEC.

To describe the experience of a single center tertiary pediatric hospital including patients diagnosed and treated with IE over the last three decades.

A retrospective chart review of pediatric patients with definite IE by the Dukes Criteria admitted to the Hospital for Sick Children between January 1978 and January 2008 was conducted. Data distribution was divided into periods I (PI) (1978-1989), PII (1990-1999), and PIII (2000-2008). Thrombotic events were included only if radiologically proven, concomitantly and/or after IE confirmation; infectious etiology was accepted only by positive blood cultures; and heart failure was defined by clinical criteria. Descriptive statistics including percentages and means were calculated; categorical correlations used Chi-square statistics; survival analysis included Kaplan Meier analysis; p values<0.05 was considered statistically significant.

113 patients (pts) aged 0 to 18 years were included [mean: 3.6 yrs; females: 46 pts (41%)]. Congenital heart defects (CHD) were present in 95 pts (84%); cyanotic CHD in 33/95 (35%). In 64 pts (57%), 67 cardiac vegetations were found by echocardiogram [mitral (19%), aortic (18%), tricuspid valve (16%); multiple vegetations: 24/67 (36%)]; aneurysms occurred in 6 pts (5%). Large vegetations (≥1 cm) developed in 32 pts (27%). Only CHD (p=0.04), but not age, gender, or organism correlated with vegetation development. Overall, community acquired IE (CA-IE) was seen in 69/113 pts (61%), caused mostly by S. viridans; nosocomial and postoperative IE was caused mostly by S. aureus (54% vs. 27%; p<0.05), showing significantly higher mortality (27% vs. 7%; p<0.05). Period III showed the highest rate of older pts (p<0.001), CA-IE (p=0.04), vegetations (p=0.01), and TEC (p=0.01), but not of surgical interventions; yet, mortality was comparable in all periods (p=0.98).

Fourty-one pts developed 44 TEC (40%), the presenting feature in 14 pts (12%). 22 pts (19%) were diagnosed with CVE (18 AIS; 4 cerebral mycotic aneurysms); 23 pts (20%) had non-CVE. Only older age (>3yrs) correlated with either CVE (p=0.04), or non-CVE (p=0.004). Conversely, TEC (p=0.06), AIS, or non-CVE did not predict mortality (p=0.5). Anticoagulation usage included: a) heparin/low molecular weight (3 pts), warfarin (8 pts), and antiplatelet therapy (22 pts); no major bleeding occurred. Overall, 5-yr mortality rate was 15%. Risk factors for mortality by univariate analysis included vegetations (p=0.001), vegetation size (p=0.01), and heart failure (p<0.001), of which vegetations (OR: 11.1, 95%CI: 1.5–88; p=0.005) and heart failure (OR: 145.0, 95%CI: 18–1221; p<0.0001) were confirmed by multivariate logistic regression. Overall survival at 1 month from IE diagnosis in patients without or with heart failure was approximately 92% and 75%, respectively (p<0.0001).

Our single tertiary centre study reports a growing, high incidence rate of TEC in children with confirmed IE over the last three decades. Children >3yrs had a higher risk to develop any type of TEC; however, neither CVE nor non-CVE were associated with an increased mortality. While no major bleeding events occurred in pts treated with anticoagulants, there are limitations to comment on its safety in this condition. Mortality rate remained stable over time; heart failure and the presence of vegetations significantly increased the risk for death.

No relevant conflicts of interest to declare.