Mortality after bloodstream infections in allogeneic haematopoietic stem cell transplant (HSCT) recipients

Abstract

Bloodstream infections (BSIs) are frequent after allogeneic haematopoietic stem cell transplantation (HSCT). The aim of this study was to identify predictors of mortality after BSI in patients who undergo HSCT. Patients who underwent HSCT between 1 January 2004 and 31 January 2008 and developed BSI during the first year post-transplantation were included. Variables influencing overall mortality at 7 and 30 days after BSI were analysed. BSIs developed in 149 patients, within a median of 9 days after undergoing HSCT. Early and late mortality were 15 and 27%, respectively. Of the BSI, 54% were due to Gram-positive microorganisms, 33% were due to Gram-negative microogranisms, 10% were polymicrobial and 3% were fungal. The associated 7-and 30-day mortality was respectively 10 and 24% (Gram positive), 22 and 31% (Gram negative; Pseudomonas aeruginosa mortality 67%, all within 7 days), 13 and 27% (polymicrobial) and 40% (fungal, all within 7 days). Early mortality was higher in relapsed disease at HSCT (25.9%, p = 0.01), but lower in early (i.e. within 20 days of HSCT) BSI (11.7%, p = 0.03) and BSI due to Gram-positive infective agents (10%, p = 0.05). Multivariate analysis confirmed a higher mortality in late BSI [odds ratio (OR) 3.29, p = 0.03] and relapsed disease at HSCT (OR 2.2, p = 0.04). Late mortality was associated with the type of underlying disease (OR 0.44 for diseases other than acute leukaemia, p = 0.05) and its status (OR 6.04 for relapse at HSCT, p = 0.001). Appropriate empirical therapy was associated with lower early and late mortality in single Gram-negative BSI (16 vs. 45% for 7-day mortality, p = 0.09; 21 vs. 64% for 30-day mortality, p = 0.02). BSIs are frequent during the first year after HSCT and are associated with a high mortality rate. The aetiology influenced early mortality, while the type and phase of the underlying disease played a pivotal role in late mortality. Appropriate empirical therapy is crucial in BSI due to Gram-negative infective agents.