Objectives: Clostridioides difficile infections (CDI) are common in autologous (auto-HSCT) or allogenic hematopoietic stem cell transplant (allo-HSCT) recipients. However, the impact of CDI on patient outcomes is controversial. We conducted this study to examine the impact of CDI on patient outcomes. Methods: We performed a retrospective single-center study, including 191 lymphoma patients receiving an auto-HSCT and 276 acute myeloid leukemia (AML) patients receiving an allo-HSCT. The primary endpoint was overall survival (OS). Secondary endpoints were causes of death and, for the allo-HSCT cohort, GvHD-and relapse-free survival (GRFS). Results: The prevalence of CDI was 17.6% in the AML allo-HSCT and 7.3% in the lymphoma auto-HSCT cohort. A higher prevalence of bloodstream infections, but no differences concerning OS or cause of death were found for patients with CDI in the auto-HSCT cohort. [AU] In the allo-HSCT cohort, OS and GRFS were similar between CDI and non-CDI patients. However, the leading cause of death was relapse among non-CDI patients, but it was infectious diseases in the CDI group with fewer deaths due to relapse. Conclusions: CDI was not associated with worse survival in patients receiving a hematopoietic stem cell transplantation, and there were even fewer relapse-related deaths in the AML allo-HSCT cohort.

Bloodstream infections (BSI) are a frequent complication in patients with hematological and oncological diseases. However, the impact of different bacterial species causing BSI and of multiple BSI remains incompletely understood. We performed a retrospective study profiling 637 bacterial BSI episodes in hematological and oncological patients. Based on the 30-day (30d) overall survival (OS), we analyzed different types of multiple BSI and grouped BSI-associated bacteria into clusters followed by further assessment of clinical and infection-related characteristics. We discovered that polymicrobial BSI (different organisms on the first day of a BSI episode) and sequential BSI (another BSI before the respective BSI episode) were associated with a worse 30d OS. Different bacterial groups could be classified into three BSI outcome clusters based on 30d OS: favorable (FAV) including mainly common skin contaminants, Escherichia spp. and Streptococcus spp.; intermediate (INT) including mainly Enterococcus spp., vancomycin-resistant Enterococcus spp., and multidrug-resistant gram-negative bacteria (MDRGN); and adverse (ADV) including MDRGN with an additional carbapenem-resistance (MDRGN+CR). A polymicrobial or sequential BSI especially influenced the outcome in the combination of two INT cluster BSI. The presence of a polymicrobial BSI and the assignment into the BSI outcome clusters were identified as independent risk factors for 30d mortality in a Cox multivariate regression analysis. The assignment to a BSI outcome cluster and the differentiated perspective of multiple BSI open new insights into the prognosis of patients with BSI and should be further validated in other patient cohorts.