Abstract
Background: The gut microbiome and microbial metabolites are key predictors and modulators of tumorigenesis, anti-tumor immunity, and treatment response. Microbial dysbiosis and loss of protective microbiota-derived metabolites are evident during stem cell engraftment in the autologous and allogeneic stem cell transplantation settings. Short chain fatty acids (SCFA), produced by anaerobic gut commensals with fermentative activities, are histone deacetylase inhibitors. Elevated fecal butyrate levels are associated with minimal residual disease negativity at 3 months and better CAR T cell responses in multiple myeloma (MM) patients. We hypothesize that loss of SCFA producers accelerate MM progression and worsens outcomes.
Methods: We profiled the gut microbiome by 16S rRNA sequencing of the V4-V5 region. We collected stool samples of 64 newly diagnosed MM patients who underwent autologous stem cell transplantation (ASCT) prior to ASCT, during engraftment and 2-4 months post-ASCT. To quantify the number of SCFA producers of each type we used the definitions from HUMAnN 3.0 and Microbial Metabolites Database (MiMeDB). We also trained a Random Forest predictor to identify the most associated taxa for late relapse and found multiple butyrate producers among the top 10 associated taxa. Correlations between taxa, clinical covariates and progression-free survival (PFS) were assessed with the Multivariate Cox regression model. The model included 172 events. We reported significant results at a false discovery rate (FDR) correction < 0.05. The impact of butyrate was tested in the immunocompetent syngeneic 5TGM1 model. The viability of mouse 5TGM1 MM cells treated with butyrate was determined by CellTiter-Glo luminescent viability assay. MM progression was monitored weekly by in vivo bioluminescence imaging and measuring serum IgG2b paraprotein levels by ELISA.
Results: We observed a significant decrease in the abundance of butyrate-producing bacteria during stem cell engraftment compared to pre-ASCT levels. In a linear regression model, loss of butyrate producers at engraftment were associated with inferior PFS. In the multivariate Cox regression model, our analysis of the microbiome revealed 89 bacterial taxa that were significantly associated with PFS after FDR correction. Among the most significant taxa, Parvimonas micra (HR,1.13, p=0.013) and Ligilactobacillus salivarius (HR,1.10, p=0.0059) were identified as risk factors, while butyrate producers Pseudobutyrivibrio xylanivorans (HR,0.94, p=0.377), Roseburia Sp. 831B (HR, 0.94, p=0.226) and Petrocella atlantisensis (HR=0.71, p=0.00015) predicted better outcomes. To determine the therapeutic potential of butyrate, we first assessed its MM-intrinsic effects on 5TGM1 cell viability. After 48 hr, we observed an 87% decrease in cell viability (p<0.0001) with an IC50=1.423 mM. We then treated 5TGM1 mice with butyrate starting on day 7 post-5TGM1 injection. After 28 days, butyrate markedly decreased MM burden by 95% (p=0.001) and reduced serum paraprotein levels by 36% (p=0.034).
Conclusions: We observed that depletion of butyrate-producing bacteria was associated with worse PFS at time of engraftment. Higher abundance of specific butyrate-producing taxa at this critical time pointwas associated with better outcomes, suggesting that the gut microbiome may represent a novel actionable biomarker and potential therapeutic target to improve post-transplant outcomes in MM. Complementing our findings, our in vitro and in vivo 5TGM1 therapeutic studies highlight the feasibility of leveraging a post-biotic as an adjuvant to improve treatment response. Importantly, our findings suggest that the gut microbiome provides prognostic information independent of traditional clinical markers for MM progression-free survival.
