Culture-positive and culture-negative endocarditis in patients with cancer: a retrospective observational study, 1994-2004.

Yusuf SW, Ali SS, Swafford J, Durand JB, Bodey GP, Chemaly RF, Kontoyiannis DP, Tarrand J, Rolston KV, Yeh E, Raad II, Safdar A.

Medicine (Baltimore). 2006 Mar;85(2):86-94.

Abstract

Endocarditis is uncommon in patients with cancer. The characteristics of culture-positive (CPE) and culture-negative endocarditis (CNE) in high-risk cancer patients are not known; therefore we sought to evaluate the disease characteristics in patients with endocarditis at a comprehensive cancer center. We retrospectively reviewed the transthoracic (TTE) and transesophageal (TEE) echocardiograms obtained from 654 consecutive cancer patients in whom endocarditis was suspected between 1994 and 2004. Endocarditis was confirmed in 45 (7%) of 654 patients using modified Duke University criteria based on information obtained from hospital records and computerized data systems. In 21 (95%) of 22 cases, TEE examinations were diagnostic, and 16 (42%) of 38 patients with initially nondiagnostic TTE studies had the diagnosis confirmed by TEE study; this difference between diagnostic TEE and initial nondiagnostic TTE was significant (p < 0.0001). Among the 26 (58%) patients with CPE, Staphylococcus aureus (35%) was the most common organism isolated, followed by coagulase-negative Staphylococcus species (23%). Eighteen (78%) of 23 patients with a central venous catheter had CPE, whereas only 8 (36%) of 22 patients without a central venous catheter had CPE (odds ratio [OR], 6.3; 95% confidence interval [CI], 1.69-23.53; p < 0.006). Vegetations were larger in patients with CPE than in patients with CNE (median +/- standard deviation, 10 +/- 8.8 vs. 8.7 +/- 3.9 mm). Fifteen patients (58%) with CPE and 10 (53%) with CNE had embolic complications. We note that cutaneous and septic pulmonary emboli were more common in patients with CPE than in patients with CNE (31% vs. 11% and 15% vs. 0%, respectively), whereas embolic cerebrovascular and fatal embolic coronary events were more common in patients with CNE than in those with CPE (37% vs. 12% and 21% vs. 0%, respectively; p = 0.026). The 4-week endocarditis-attributable death rate did not differ significantly between the groups (CPE, 15% vs. CNE, 32%; p = 0.28). On stepwise multivariate regression analysis, patients with neutropenia (OR, 22.52; 95% CI, 2.25-225.48; p < 0.008) and those with embolic cerebrovascular events (OR, 17.07; 95% CI, 1.63-178.45; p < 0.01) had an increased probability of death due to endocarditis. The clinical spectrums of CPE and CNE differed in these patients with cancer. In patients with CNE, embolic cerebrovascular and fatal myocardial infarction were relatively common.