Reduced mortality from KPC-K.pneumoniae bloodstream infection in high-risk patients with hematological malignancies colonized by KPC-K.pneumoniae

BACKGROUND: KPC-K.pneumoniae bloodstream infection (KPC-KpBSI) mortality rate in patients with hematological malignancies is reported about 60%. The initial treatment active against KPC-K.pneumoniae is crucial for survival and KPC-K.pneumoniae rectal colonization usually precedes KPC-KpBSI. We evaluated the impact on KPC-KpBSI mortality of the preemptive use of antibiotics active against KPC-K.pneumoniae, as opposed to inactive or standard empiric antibiotics, for the empiric treatment of febrile neutropenia episodes in patients with hematological malignancy identified as KPC-K.pneumoniae intestinal carriers. METHODS: We compared the outcomes of KPC-KpBSIs occurring in high-risk hematological patients known to be colonized with KPC-K.pneumoniae, during two time periods: March2012-December2013 (Period 1, initial approach to KPC-K.pneumoniae spread) and January2017-October2018 (Period 2, full application of the preemptive strategy). The relative importance of the various prognostic factors that could influence death rates were assessed by forward stepwise logistic regression models. RESULTS: KPC-KpBSI-related mortality in hematological patients identified as KPC-K.pneumoniae carriers dropped from 50% in Period 1 to 6% in Period 2 (p < 0.01), from 58 to 9% in acute myeloid leukemia carriers(p < 0.01). KPC-KpBSIs developed in patients identified as KPC-K.pneumoniae carriers were initially treated with active therapy in 56% and 100% of cases in Period 1 and Period 2, respectively (p < 0.01), in particular with an active antibiotic combination in 39 and 94% of cases, respectively(p < 0.01). The 61% of KPC-KpBSI observed in Period 1 developed during inactive systemic antibiotic treatment (none in Period 2, p < 0.01), fatal in the 73% of cases. Overall, KPC-KpBSI-related mortality was 88% with no initial active treatment, 11.5% with at least one initial active antibiotic (p < 0.01), 9% with initial active combination. Only the initial active treatment resulted independently associated with survival. CONCLUSIONS: In high-risk hematological patients colonized by KPC-K.pneumoniae, the empiric treatment of febrile neutropenia active against KPC-K.pneumoniae reduced KPC-KpBSI-related mortality to 6% and prevented fatal KPC-KpBSI occurrence during inactive systemic antibiotic treatment.