Impact of antibiotic exposure on CAR-T cell therapy outcomes: Analysis in multiple Myeloma and lymphoma

Abstract Background: The gut microbiome is known to play a significant role in cancer immunotherapy. Antibiotic use can disrupt the intestinal microbiota and potentially diminish the antitumor efficacy of CAR-T cell therapy. According to a 2023 study by Stein-Thoeringer et al. in Nature Medicine, administration of broad-spectrum antibiotics prior to CAR-T infusion was associated with poorer treatment outcomes. Objective To assess the association between pre-CAR-T antibiotic exposure and clinical outcomes (response rates, progression-free survival [PFS], overall survival [OS]) and treatment-related adverse events (cytokine release syndrome [CRS], immune effector cell-associated neurotoxicity syndrome [ICANS], and prolonged hematologic toxicity) in patients with multiple myeloma (MM) and lymphoma. Methods We retrospectively analyzed the cohorts of CAR-T therapy treated patients in the three disease categories above. Patients were grouped by antibiotic exposure within a defined pre-CAR-T window (4 weeks prior to infusion). A risk-scoring model for severe hematologic toxicity further categorized patients into high-risk (score ≥3) and low-risk (score 0–2) subgroups. Outcomes were compared between groups. Results Among 334 patients, the MM cohort (n=199) showed no significant difference in objective response rate (ORR: 70.0% vs. 81.7%) between antibiotic-exposed (n=90) and non-exposed (n=109) groups, but complete remission (CR) rates were significantly lower in the antibiotic group (40.0% vs. 57.8%). Antibiotic-exposed patients had shorter median PFS (6.7 vs. 13.9 months) and OS (21.9 vs. 36.1 months). CRS (71.1% vs. 69.7%) and ICANS (3.3% vs. 1.8%) incidences were comparable. Similarly, stratified analyses of ATB classes were conducted, yielding consistent results. In the lymphoma cohort (n = 135), 61 (45.2%) patients were exposed to antibiotics. Importantly, the difference in ORR between the two groups was not statistically significant (66.2% vs. 70.5%), nor was there a significant difference in the incidence of CRS (16.4% vs. 21.6%). However, patients who received antibiotics had a significantly shorter median PFS and OS than patients who did not receive antibiotics (P< 0.05). Conclusion Pre-CAR-T antibiotic exposure, particularly to broad-spectrum low-risk antibiotics (HR ABX), reduces complete remission rates and shortens survival across MM and lymphoma, potentially via gut microbiome disruption. While CRS/ICANS risks remain unaffected, judicious antibiotic use is critical to optimize CAR-T efficacy. Future strategies should prioritize microbiome preservation or selective use of low-risk antibiotics (LR ABX).