Abstract

Abstract 4141

Bacteremia is a common complication after alloBMT due to neutropenia, mucositis, GvHD, immunosuppressive agents, and slower immune reconstitution and full chimerism after reduced intensity conditioning. Since symptoms and signs of bacteremia (e.g., fever) may be blunted or absent in BMT recipients receiving intensive immunosuppression, our clinical practice applies a low threshold for obtaining blood cultures. Our goals were to define risk factors for bacteremia at pre-specified periods after BMT and to compare rates of bacteremia using microbiological criteria vs. CDC/NHSN criteria which require symptoms and signs such as fever, chills or hypotension. We performed a retrospective cohort study of 168 adult (>18 yrs) consecutive first alloBMT patients from January 2008 to May 2012. A bacteremia episode (BE) was defined according to the EBMT Infectious Disease Working Party criteria which excludes skin flora contaminants. A second set of positive blood cultures of the same bacterial strain >30 days apart was considered a recurrent BE. Blood cultures which grew different bacterial strains were defined as a single episode of polymicrobial bacteremia if drawn < 48 hrs apart and as unique BEs if drawn > 48 hrs apart. Patient characteristics include: median (range) age 54 (19–73) yrs, 57% male, 49% AML, 14% ALL, 11% MDS, 7% NHL, 20% other diseases, 77% KPS 50–80, 81% peripheral blood, 64% unrelated donor, 23% <10/10 HLA match, 73% reduced intensity conditioning regimens. After 8/2011, all patients received levofloxacin prophylaxis during neutropenia, whereas before 8/2011, no routine antibacterial prophylaxis was used. At discharge, penicillin prophylaxis was initiated for all patients until at least 1 yr post-alloBMT. All patients received voriconazole or micafungin as anti-fungal prophylaxis from day -1 until off immunosuppression and acyclovir as anti-viral prophylaxis for at least 1 yr post-alloBMT. Overall, there were 238 BEs in 168 patients for a rate of 1.42 BE/pt. The rate of BEs was highest in the early peri-BMT period and dropped significantly after day+30 (Table), and 80% of BEs before day+30 occurred during neutropenia. The proportions of patients with >1 BE in each time period were 49%, 31%, and 28%, respectively. While 67% of BEs (defined by microbiological criteria), from admission (first day of conditioning regimen) to day+30 met CDC/NHSN criteria, only 25% of BEs after day+30 did so. The majority of BEs did not result in a central line replacement. From admission through 1 yr post-allo BMT, 59 (35%), 49 (29%) and 60 (36%) patients had 0, 1, and >1 cumulative total BEs, respectively. We performed a univariate analysis of risk factors including gender, age, KPS, diagnosis, donor relation, HLA match, SC type, use of levofloxacin as antimicrobial prophylaxis, and conditioning regimen intensity, for >1 BE for each of the 3 time periods, and cumulatively. Lack of levofloxacin prophylaxis was a significant risk factor for BEs in all time periods post-alloBMT. In addition, grade II-IV acute GvHD was a significant risk factor for >1 or >2 BEs in the day+31 to 100, day+101 to 365 and admission to day+365 time periods. Levofloxacin prophylaxis decreased the rate of patients with >1 BE from 54% to 14% in the day+30 time period (P<0.0001) and from 69% to 36% from admission to 1 yr post-alloBMT (P=0.003). BEs resulted in death within 7 days of first positive culture in 4 patients due to MRSA, E faecium, K pneumonia, and E faecalis/Candida. We observed the greatest discordance between applying microbiological criteria vs. CDC criteria to define BEs from days+31 to 365. This discordance reflects the lack of sensitivity of typical clinical signs/symptoms of infection in the setting of GvHD post-alloBMT, and points to the need for modified criteria to define BEs for this patient population. Development of improved prophylaxis strategies that focus on patients with GvHD may diminish the incidence and severity of these infections.

Time Period# of BEs# pts with >=1 BE/# pts evaluable (%)# (%) of BEs meeting CDC criteria# (%) of BEs resulting in a CVC line changeTotal # of Observation Days# BEs/Observation DaysRate of BE/1000 pt-Days
Admission to Day +30 92 82 (49) 62 (67) 7 (8) 5,970 0.0154 15 
Day +31 to 100 70 51 (31) 18 (26) 10 (14) 10,693 0.0065 
Day +101 to 365 76 40 (28) 18 (24) 5 (7) 29,836 0.0025 
Time Period# of BEs# pts with >=1 BE/# pts evaluable (%)# (%) of BEs meeting CDC criteria# (%) of BEs resulting in a CVC line changeTotal # of Observation Days# BEs/Observation DaysRate of BE/1000 pt-Days
Admission to Day +30 92 82 (49) 62 (67) 7 (8) 5,970 0.0154 15 
Day +31 to 100 70 51 (31) 18 (26) 10 (14) 10,693 0.0065 
Day +101 to 365 76 40 (28) 18 (24) 5 (7) 29,836 0.0025 

Disclosures:

Off Label Use: Levofloxacin is approved for treatment of infections. We will discuss its use as infection prophylaxis.

Author notes

*

Asterisk with author names denotes non-ASH members.