Evaluation of Pre-Transplant Risk Factors of Bacterial Bloodstream Infection (BSI) in Patients with Hematopoietic Stem Cell Transplantation (HSCT) and Comparison of BSI Between Allogenic and Autologous HSCT

Backgrounds

Although improvement of supportive cares including antimicrobial prophylaxis has significantly ameliorated treatment outcomes in patients with hematopoietic stem cell transplantation (HSCT), bacterial bloodstream infection (BSI) is still a major cause of morbidity and mortality of HSCT. The purpose of this study is to evaluate pre-transplant risk factors of BSI after HSCT and compare the nature of BSIs between allogenic and autologous HSCT.

Methods

Adult patients (age ≥ 18 years) who received either allogenic or autologous HSCT in a single institution between November 2002 and June 2012, were analyzed. Report of blood cultures from the start time of conditioning therapy to hospital discharge, and clinical relevance of the BSIs, were reviewed. Evaluated potential risk factors of BSI were type of HSCT (allogenic vs. autologous), age, gender, disease (AML vs. others), time from diagnosis to HSCT, amount of infused CD34⁺ stem cell, HCT-CI score, and use of antibiotics from conditioning for prophylactic or therapeutic intent. Blood level of C-reactive protein (CRP), albumin, and ferritin within 14 days from HSCT were also included in the analysis. Among patients with allogenic HSCT, modified EBMT score, concomitant acute GVHD, type of donor and intensity of conditioning were additionally evaluated.

Results

One hundred and thirty-four patients were analyzed (59 for allograft and 75 for autograft). Thirty-eight bacterial isolates from blood of 34 patients (25.3%; 20 patients for allograft and 14 for autograft) were reported. Patients with allogenic HSCT had more frequent BSI (odds ratio 2.23, p = 0.047) compared to those with autologous HSCT. BSI of autologous HSCT were reported earlier compared to those of allogenic HSCT (mean 12.1±3.4 vs. 26.5±18.5 days, p = 0.007). HSCT in patients with AML (p = 0.029), no use of antibiotics from conditioning (p < 0.001), and elevated CRP (p = 0.001) were independently associated with higher incidence of BSI in whole patient population. Among 59 patients with allogenic HSCT, HSCT > 180 days after diagnosis (p = 0.035), elevated CRP (p = 0.018), lower serum albumin (p = 0.033), and acute GVHD ≥ G2 (p = 0.015) showed relation to BSI in multivariate analysis. As for 75 patients with autologous HCT, only no antibiotics use (p = 0.007) and elevated CRP (p = 0.031) were independent risk factors of BSI. BSIs after allograft was more fatal: 7 of 20 patients with allograft (35%) died of BSI whereas only 1 of 14 patients who underwent autologous HSCT expired. 

Conclusion

Except pre-transplant serum CRP elevation, allogenic and autologous HSCT have different risk factors. BSI with autologous HSCT occurred earlier and showed better clinical outcomes compared to BSI with allograft. Distinctive natures of bacterial infection after HSCT between allogenic and autologous HST should be considered to establish the best defense strategy against BSI.