Children with Gut Acute Graft Versus Host Disease (aGVHD) Following Allogeneic Stem Cell Transplant (AlloHCT) Have an Increased Incidence of Enteric Bacteria Blood Stream Infections (EB-BSI)

Bacterial septicemia is a leading cause of mortality following AlloHCT. A few studies have reported a higher incidence of BSI in patients with aGVHD, supporting the theory that aGVHD is associated with an increased risk of infection. The aim of our study was to evaluate the relationship between the presence of gut aGVHD and the subsequent development of EB-BSI.

In this retrospective study, EB-BSI (enterococci and gram-negative rods) data was collected between day 0 and day +180 following AlloHCT, The incidence rates of EB-BSI were compared before and after the onset of gut aGVHD. Two hundred sixty four children underwent AlloHCT for malignant (n=162, 61.4%) and non-malignant (n=102, 38.6%) disorders between 2000-2012.  Median age was 9 years, M/F ratio was 64/36%, and donor sources included matched family donors (n=107, 40%) and matched unrelated donors (n=157, 60%). Patients were conditioned with either myeloablative (n=122, n=46%), reduced toxicity (n=66, 25%), or reduced intensity (n=76, 29%) regimens. Additionally, a subset of the patients received alemtuzumab (n=71, 27%) or rabbit ATG (n=134, 51%).

The groups of patients with EB-BSI (n=122, 46.2%) and without EB-BSI (n=142, n=53.8%) were comparable (age, p=0.18; sex, p=0.8, disease type, p=0.13; donors, p=0.26; conditioning regimen, p=0.19; use of alemtuzumab, p=0.8 and ATG, p=0.4). The overall incidence of aGVHD was 44% (n=115), including both gut aGVHD (28%) and non-gut aGVHD (16%). The incidences of grades II, III and IV gut aGVHD were 48.7%, 48.7% and 2.7%, respectively.

If the development of gut aGVHD is a risk factor for development of EB-BSI, we reasoned that there should be a temporal relationship between onset of aGVHD and infection.  We therefore calculated the infection density both before and after the onset of gut aGVHD, by dividing the number of episodes of EB-BSI during each time period by the total number of patient-days in that time period multiplied by 1000. Prior to the onset of gut aGVHD, the infection density was 2.4 infections/1000 person-days.  After onset of gut aGVHD, it was 6.5 infections/1000 person-days (p-value=0.022).  These data support the hypothesis that gut aGVHD predisposes patients to development of EB-BSI, given the asymmetric distribution of EB-BSI after onset of aGVHD. We plan to analyze the infection densities in several other groups and risk factors associated with higher infection density in patients with gut aGVHD specifically.

To our knowledge, this is the first study to demonstrate that the development of gut aGVHD increases the risk for EB-BSI. Strategies to reduce the elevated risk of EB-BSI in patients who develop gut aGVHD, such as prophylactic use of probiotics, should be considered based on these findings.