The use of fecal surveillance cultures in predicting bacteremia in patients undergoing intensive chemotherapy and stem cell transplant is an unsettled issue (Neshar et al. Transpl. Infect Dis. 2015). With the increasing incidence of multi-drug resistant (MDR) organisms and high mortality rates with these infections, we sought to describe the spectrum of MDR identified in fecal surveillance, and re-visit the use of fecal surveillance in predicting infection with MDR organisms post-allogeneic stem cell transplant (allo-SCT).


We analyzed data from all patients who underwent allogeneic stem cell transplant during a 2 year period (2014-2015). Patients with MDR strains of bacteria (defined in this study as defined as Vancomycin resistant enterococcus (VRE), Extended spectrum Beta-Lactamases (ESBL) and Carbapenem resistant enterobacteriacea (CRE)) in fecal surveillance were compared with patients who did not have MDR in fecal surveillance cultures. Baseline characteristics and post allo-SCT outcomes including MDR blood culture positivity, severe sepsis and 100-day transplant related mortality (TRM) were compared. Multivariate analysis using logistic regression model was used to determine independent predictors of outcome.


A total of 313 allogeneic stem cell transplants were performed in 299 patients, of which data on pre-transplant fecal surveillance cultures were available in 232 transplants. The incidence of MDR isolates in fecal surveillance cultures was 56% (134/232, with E.Coli 118, Klebsiella 33 and Enterococcus 13). Of these, 129 were ESBL alone (78.6%), 17 were CRE alone (10.3%), 10 were ESBL + CRE (6%), 6 (3.6%) were VRE alone. More than one drug resistant organism was isolated in fecal surveillance in 31 patients. The incidence of any MDR positivity in post-transplant blood cultures in all patients was 13.8% (32/232), with 9.4% CRE, 2.6% ESBL, and 1.7% VRE. Thirty-one patients had severe sepsis without any MDR organism isolated on blood culture, but of these 3 had CRE in sputum, 2 had Colistin resistant Acetinobacter (sputum), 2 had candidemia, 3 had NF-GNB and one had Enterococcus.

Baseline characteristics between patients who were positive and negative for MDR in fecal surveillance were similar in relation in relation to age (p=0.058), diagnosis (0.133), type of transplant - matched family/MUD/Haplo (p=0.610), stem cell source (0.370), grade 3-4 GVHD (0.834). There was a significantly higher subsequent blood MDR positivity (any MDR) (p=0.012), 100-day mortality (p=0.012) and poor outcome (severe sepsis or 100 day mortality) (p=0.006) in patients who had MDR detected in fecal surveillance cultures. However, of the 25 patients who had MDR isolated in both fecal surveillance culture as well as subsequent blood cultures, only 9 patients had the same organism and susceptibility pattern in both.

Factors influencing 100-day mortality included patient's age (p=0.001), MDR positivity in blood (p=0.0001), MDR in fecal surveillance (p=0.01), use of an alternate (Haplo/MUD/family) donor (p=0.0001), GVHD grade 3-4 (p=0.000) and severe sepsis (p=0.000). On multivariate analysis, only patient's age (p=0.015), alternate donor (0.009), severe sepsis (p=0.000) and grade 3-4 GVHD (p=0.001) retained significance in predicting 100-day mortality.


Multidrug resistant organisms are frequently seen on fecal surveillance in the pre-transplant setting. MDR in fecal surveillance is associated with a higher incidence of MDR positive blood cultures but not with the same organism, and on univariate analysis is associated with higher incidence of 100 day mortality. Newer strategies to reduce bacteremia and mortality in this group of high risk patients need to be considered.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.