The Impact of Colonization with Multidrug-Resistant Bacterial Pathogens (MRP) in Allogeneic Hematopoietic Cell Transplantation - Clearance of the MRP Is Pivotal for Survival

Multidrug-resistant bacterial pathogens (MRP) such as extended-spectrum beta-lactamase producing enterobacteriaceae (ESBL), vancomycin-resistant enterococci (VRE), methicillin-resistant Staphylococcus aureus (MRSA) and multi-resistant Pseudomonas aeruginosa (P. aeruginosa) are an emerging challenge in allogeneic hematopoietic cell transplantation (HCT). However, no comprehensive data are available on the prevalence of MRP, their impact on the outcome after HCT and on the probability to clear a MRP. It was the purpose of this study to systematically analyze the issue of MRP in HCT.

PATIENTS AND METHODS: From 07/2010 to 12/2015 a total number of 121 (43 F; 78 M) consecutive patients who received the first allogeneic HCT at our institution were analyzed retrospectively. As baseline investigation before conditioning all patients underwent a comprehensive screening for MRP. Swabs from nose, throat, axilla, urethra and anus as well as samples from stool and urine were collected. During the course of transplantation surveillance cultures were performed weekly. In addition, routine microbiological investigations were done from blood, urine, swabs, stool or central venous catheters whenever clinically needed. In MRP colonized patients surveillance stool specimen were taken until the MRP was repeatedly non-detectable. Multidrug-resistant gram neg. bacteria were categorized as 4MRGN (resistant to cephalosporins, piperacillin, fluorochinolones and carbapenems) or as 3MRGN (resistant to 3 of these 4 antimicrobial drug groups). The primary endpoint of this analysis was day 100 non relapse mortality (NRM). Secondary endpoints were NRM and overall survival (OS) after 2 years. A further endpoint in MRP+ patients was the time to non-detectability of the MRP.

RESULTS: The patient characteristics were as follows: Underlying diseases were AML (62), ALL (7), CML (8), MPN (5), lymphoma (9), MDS (25), and multiple myeloma (5). The conditioning regimen was myeloablative in 50, reduced intensity in 71 patients. Patients were transplanted with peripheral blood stem cells (105) or bone marrow (16) from matched siblings (28), matched unrelated (67), mismatched (15) or haploidentical donors (11). 33 patients (27%) were colonized by at least one MRP (MRP+ group) either at baseline (baseline MRP+ group, n=18, 15%) or at any other time point until day 100 post HCT. The 33 MRP+ group patients were colonized by 42 MRP (baseline MRP+ group: 19 MRP). Detected MRP were 3MRGN E. coli or Klebsiella pneumonia (17), 4MRGN (9) or 3MRGN (2) P. aeruginosa, multi-resistant Stenotrophomonas maltophilia (2), 3MRGN Citrobacter freundii (1), 3MRGN Acinetobacter baumanii (1), 4MRGN Enterobacter cloacae (2), VRE (7) and MRSA (1). Out of these 33 patients 12 (36%) developed an infection with an MRP after HCT: septicemia (n=9), pneumonia caused either by 3MRGN Klebsiella (n=1) or by 4MRGN P. aeruginosa (n=1) and urinary tract infection by 4MRGN Enterobacter cloacae (n=1). 5 patients died MRP related due to septicemia (4MRGN P. aeruginosa n=4, VRE n=1). However, day 100 and 2-year NRM of MRP colonized vs non-colonized patients were essentially the same: 15 and 21% vs 15 and 24%, respectively. Even for the baseline MRP+ group there was no significant difference of NRM: 17 and 29% vs 15 and 22%. Overall survival was also not impaired in the MRP+ group 2 years post HCT (median follow up 32.4 months, range 7.5 to 71.4 months): MRP colonized versus non-colonized patients: 60 vs 55% (baseline MRP+ group 54 vs 58%). Out of the 33 MRP+ group patients 21 patients were able to clear the MRP. On day 100 after HCT 36% of patients had been able to clear the MRP. Median time to non-detectability of the MRP was 6.3 months. In 12 patients the MRP did not disappear until the end of the observation period or death (median follow up 15 months). There was a highly significant (p<0.0001) survival difference between patients who cleared the MRP vs those with MRP persistence. Whereas 17 out of 21 (81%) patients who cleared the MRP survived, only 2 out of 12 patients with MRP persistence stayed alive (median survival 6.6 months). Day 100 NRM was 4 vs 42% (p=0.0023).

CONCLUSIONS: Since colonization by MRP had no neg. impact on the outcome in our cohort HCT of MRP colonized patients is feasible. However, the outcome of patients who do not clear their MRP is dismal. In order to increase the probability to clear the MRP we suggest to review the use of antibiotics in MRP colonized patients critically.