Summary
A microbiome-focused companion study found no association between gut-microbiome diversity at engraftment and one-year non-relapse mortality. Exploratory analyses found differences in microbiome diversity, microbial load and bile-acid metabolism signals between two GVHD-prevention regimens.
A microbiome-focused companion study to the randomized phase 3 BMT CTN 1703 transplant trial found that gut-microbiome diversity at neutrophil engraftment was not associated with non-relapse mortality at one year. Exploratory analyses of 304 participants found that two graft-versus-host disease (GVHD) prevention regimens were associated with different microbiome patterns in the weeks after transplant.
Allogeneic hematopoietic cell transplantation uses donor blood-forming cells to treat some blood cancers and disorders. GVHD can occur when donor immune cells attack the recipient’s tissues. BMT CTN 1801 enrolled participants from the randomized BMT CTN 1703 trial to study their gut microbiomes. All received reduced-intensity conditioning before transplantation. The analysis included 2,575 stool samples from 304 participants: 147 received tacrolimus and methotrexate (Tac/MTX), while 157 received post-transplant cyclophosphamide (PTCy), followed by tacrolimus and mycophenolate mofetil (MMF).
The primary microbiome outcome was not associated with mortality
The study’s primary question was whether microbiome diversity around neutrophil engraftment was associated with non-relapse mortality during the following year. Engraftment was defined by the first of three consecutive days on which a participant’s neutrophil count reached at least 500 cells per microlitre of blood. Researchers selected a stool sample collected at or within 14 days after that point.
They found no significant association between diversity at engraftment and non-relapse mortality, whether diversity was analysed as a continuous measure (P = 0.314) or in tertiles (P = 0.520). Diversity at engraftment was also not significantly associated with the other clinical outcomes examined.
A separate, prespecified analysis found that lower microbiome diversity in samples collected before conditioning was associated with shorter overall survival (P = 0.007). These microbiome–outcome comparisons are associations: participants were assigned to prophylaxis regimens, but their microbiome diversity was measured rather than assigned.
Differences appeared in the weeks after transplantation
Microbiome diversity declined after transplantation in both study arms, with a larger difference between regimens from days 14 to 28. The PTCy arm also had lower estimated absolute microbial load during this period; its average load was about half that of the Tac/MTX arm. Participants in the PTCy arm had more exposure to non-prophylaxis antibiotics in some analyses, but accounting for this exposure did not remove the difference in microbial load.
The researchers also found lower prevalence of Clostridium scindens and lower abundance of microbial pathways involved in secondary bile-acid metabolism in the PTCy arm during days 14–28. The presence of those pathways was associated with lower chronic-GVHD risk, mainly among participants receiving Tac/MTX. The pathway analysis estimates microbial metabolic potential from sequencing data; it is not a measurement of bile-acid concentrations in stool.
The findings point to a microbiome response that varies with the GVHD-prevention regimen, rather than a simple relationship in which lower diversity at engraftment predicts worse outcomes for all transplant recipients. The two treatment arms were combinations of medicines, so the study cannot separate the contribution of PTCy from that of MMF or other regimen differences. Its participants all received reduced-intensity conditioning, and the results may not apply to other conditioning settings.