Breakthrough viridans streptococcal bacteremia in allogeneic hematopoietic stem cell transplant recipients receiving levofloxacin prophylaxis in a Japanese hospital

BACKGROUND Breakthrough viridans streptococcal bacteremia (VSB) in patients with hematological malignancy receiving levofloxacin prophylaxis is a major blood stream infection (BSI) occurring during febrile neutropenia. However, clinical data focused on VSB in allogeneic hematopoietic stem cell transplant (allo-HSCT) recipients are lacking. METHODS The medical records of allo-HSCT recipients who received oral levofloxacin prophylaxis between January 2011 and August 2013 at Toranomon Hospital were reviewed to evaluate breakthrough VSB. Stored viridans streptococcal (VGS) species were identified by using sodA gene sequencing, and were assessed for drug susceptibility. RESULTS Among the 184 allo-HSCT recipients on levofloxacin prophylaxis, 28 (15.2 %) experienced breakthrough VSB. All of the 28 recipients with VSB were treated with a cefepime-based or piperacillin/tazobactam-based regimen. The susceptibility rates of the VGS strains for levofloxacin, cefepime, piperacillin/tazobactam, meropenem, and vancomycin were 0 %, 95 %, 100 %, 100 %, and 100 %, respectively. Both the MIC50 (minimum inhibitory concentration) and the MIC90 of ceftazidim (0.5 μg/mL and 2 μg/mL, respectively) were higher than the MIC90 of all the other anti-pseudomonal beta-lactams (APBLs). Only 1 VGS strain had a penicillin MIC ≥ 2 μg/mL by the Etest (3.6 %). There were no cases with acute respiratory distress syndrome (ARDS) that was associated with VSB, although the rate of viridans group streptococcal shock syndrome was high (26 %). The crude 30-day mortality rate in the VSB group (10.7 %) did not differ significantly from that in the BSI without VSB group (9.3 %) or non-BSI group (7.0 %) (P = 0.77). Also, VSB was not a risk factor for all-cause mortality up to 60 days following allo-HSCT (P = 0.43). CONCLUSIONS APBL with increased anti-VGS activity (APBL-VA) monotherapy would typically be optimal for treating the VGS strains in this setting. Indication of adding an empiric anti-gram-positive agent to APBL-VA for treating VSB should depend on local factors, such as the susceptibility results. In addition, breakthrough VSB is probably not a major cause of death in allo-HSCT settings, where beta-lactam non-susceptible VGS and the ARDS are rare.