Gut microbiome in DLBCL patients undergoing first‐line R‐CHOP regimen—The Oncopassport Study

Introduction: The search of biomarkers that can predict and influence treatment outcomes of diffuse large B-cell lymphoma (DLBCL) is constantly evolving, especially in the front-line setting. Preclinical studies have shown that gut microbiome (GM) diversity and specific bacterial species can have a direct impact on drug efficacy and immune-related toxicity. Although R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone) remains the gold standard as first-line treatment for DLBCL patients, about 40% of them relapsed or are refractory to this regimen. In addition, major adverse events (AEs) related to this combination, such as hematological toxicity, including febrile neutropenia (FN), and gastrointestinal (GI) AEs, do not affect all treated populations and could influence the role of GM in the tolerability and response to R-CHOP. Reconstructing GM dynamics in DLBCL patients from diagnosis to the end of R-CHOP may provide important information for patient management. Methods: Fifty DLBCL patients undergoing first-line R-CHOP regimen were enrolled (NCT03797170; RF-2016-02363730). Stools were collected at baseline, before each therapy cycle, and at response assessment (after 3 cycles and at the end of therapy), to profile GM compositional changes. Additional samples were collected upon occurrence of FN and GI AEs. GM was analyzed by 16S rRNA amplicon sequencing. Results: Of 50 enrolled patients, 45 provided at least one fecal sample. Interim and final overall response rate were 91.1% (36 complete response). Specifically, 4 patients worsened their response between cycles 3 and 6, and 4 patients improved their response. At the end of treatment, only 3 patients had progression of disease. Three patients had FN (all grade 4) and 24 patients had at least one GI AE. Beta diversity analysis showed separation between baseline and end-of-therapy GM structures (Figure 1A). In particular, the phylum Tenericutes and the genus Gemmiger increased, while the family Bifidobacteriaceae and the genus Oscillospira decreased. Additionally, patients reporting GI AE or FN segregated at baseline from those who did not report them, being enriched in Bacteroidetes and Butyricimonas, while depleted in Actinobacteria, Peptostreptococcaceae, Dorea, and Coprobacillus (Figure 1B). Conclusions: Identifying factors that may modify the response and tolerance to first line is crucial to improving lymphoma patients outcome and adherence to therapy. To date, numerous efforts have been made to recognize biologic factors that can identify patients with DLBCL at increased risk of relapse and toxicity to R-CHOP, with poor results. Interestingly, in our series, putative early GM signatures of GI AE or FN were identified, paving the way for new personalized GM-based intervention strategies to improve patient prognosis. The research was funded by: Italian Ministry of Health, RF-2016-02363730 Keywords: Aggressive B-cell non-Hodgkin lymphoma, Diagnostic and Prognostic Biomarkers