Early gut microbiome metabolites and diversity predict GVHD-free survival after allogeneic hematopoietic cell transplantation.

BACKGROUND: Disruption of the gut microbiome has been implicated in graft‑versus‑host disease (GVHD) and mortality after allogeneic hematopoietic cell transplantation (HCT). However, prior studies have focused on GVHD‑specific or mortality endpoints in isolation rather than integrated measures of transplant success. We evaluated whether early gut microbial metabolic injury and impaired ecological recovery are associated with GVHD‑free survival (GFS), a composite endpoint incorporating clinically significant GVHD and death. METHODS: We conducted a prospective observational cohort study of 62 patients undergoing allogeneic HCT at a single center. Stool samples were collected before conditioning and at weeks 2 and 4 after HCT. Fecal short‑chain fatty acids (SCFAs) were quantified using liquid chromatography-tandem mass spectrometry, and microbial diversity was assessed using 16S rRNA gene sequencing. Prespecified landmark analyses were performed at day +15 (week‑2 biomarkers) and day +28 (week‑4 biomarkers). The primary endpoint was GFS at 24‑months, defined as survival without grade II-IV acute GVHD, moderate-severe chronic GVHD, or death. Multivariable Cox regression models adjusted for key clinical covariates were used, with biomarkers modeled as log₂‑transformed continuous variables. RESULTS: During follow‑up, 43 patients (69%) experienced GFS failure, due to acute GVHD (n=18), chronic GVHD (n=18), or death without prior GVHD (n=7). In day +15 landmark analyses, higher week‑2 concentrations of all three SCFAs were independently associated with improved GFS: butyrate (adjusted hazard ratio [aHR] per doubling 0.81, 95% CI 0.71-0.93; p=0.002), propionate (aHR 0.83, 95% CI 0.75-0.93; p<0.001), and acetate (aHR 0.85, 95% CI 0.74-0.97; p=0.018). In day +28 landmark analyses, recovery of microbial diversity at week 4 was independently associated with GFS, with each doubling of the Simpson diversity index associated with a lower hazard of GFS failure (aHR 0.53, 95% CI 0.34-0.83; p=0.005). Similar associations were observed for Shannon diversity (aHR 0.51, 95% CI 0.30-0.84; p=0.009), whereas Chao1 richness showed a concordant but borderline association (aHR 0.64, 95% CI 0.41-1.02; p=0.061). CONCLUSIONS: Early post‑transplant depletion of microbiome‑derived metabolites and impaired recovery of gut microbial diversity are independently associated with inferior GVHD‑free survival after allogeneic HCT. These findings are consistent with temporally distinct associations between early metabolic injury and subsequent ecological recovery and support further investigation of microbiome‑directed strategies to improve transplant outcomes.