Immunomodulatory Metabolites Shape Gut Microbiome Recovery and Outcomes After Allogeneic HCT

The intestinal microbiome influences immune recovery and long-term outcomes after alloge-neic hematopoietic stem cell transplantation (allo-SCT). While reduced bacterial diversity and depletion of immunomodulatory microbial metabolites during peri-engraftment have been linked to acute graft-versus-host disease (GvHD) and mortality, it remains unclear whether microbiome recovery by day +100 is best captured by bacterial diversity or by functional mi-crobial metabolites. We aimed to define late post-transplant microbiome recovery and test whether a metabolite-based biomarker improves the prediction of clinical outcomes. In this prospective, longitudinal, observational study, we performed serial stool sampling from pre-transplant baseline to day +100 in a discovery cohort (n=20, Munich) and an independent val-idation cohort (n=100, Regensburg). Microbiome composition was assessed by 16S rRNA amplicon sequencing (with additional metagenomics for selected patients), and 57 metabolites were quantified by targeted mass spectrometry. Patients were classified as RECOVERY or NO RECOVERY based on bacterial richness from day +56 to +100 relative to baseline. To capture functional microbial output, we adapted the Immune-Modulatory Metabolite Risk Index (IMM-RI, based on butyric -, propionic -, isovaleric acid, desaminotyrosine, and indole-3-carboxaldehyde) to the late post-transplant period. Bacterial richness frequently improved by day +100, but microbiome recovery did not reliably indicate return to baseline community structure, was not mirrored by metabolite restoration, and showed limited association with survival or transplant-related mortality (TRM). In contrast, IMM-RI post-transplant low identified patients with preserved butyrate-associated biosynthetic capacity and significantly improved overall survival (OS) in both cohorts (validation p<0.0001). In the validation cohort, low IMM-RI post-transplant was associated with reduced relapse rates, but higher incidences of chronic GvHD; and stool butyric- and valeric acid concentra-tions were increased in chronic GvHD of the skin, indicating context-dependent metabolite effects. These findings indicate that functional metabolite profiling outperforms bacterial diversity for predicting outcomes after allo-SCT and support microbial metabolites as predictive bi-omarkers for risk stratification and precision microbiome interventions after allo-SCT.