First results of a dose-escalation study evaluating <i>Clostridium butyricum</i> MIYAIRI 588 (CBM588) with nivolumab/ipilimumab (nivo/ipi) in metastatic renal cell carcinoma (mRCC).

Abstract

503 Background: CBM588 has shown potential to improve clinical outcomes when combined with immune checkpoint inhibitors (ICIs) in untreated mRCC (Dizman et al Nat Med 2022; Ebrahimi et al Nat Med 2024). Since no significant treatment-related adverse events were related to CBM588 in previous studies, we sought to determine if CBM588 capsules, a novel formulation of this live biotherapeutic, designated MO-03, could be delivered at higher doses with similar tolerability and greater biologic effect. Methods: Patients (pts) with treatment-naïve mRCC and any international mRCC Database Consortium (IMDC) risk were treated with nivo (3 mg/kg q3w ×4, then 480 mg q4w) and ipi (1 mg/kg q3w ×4) plus CBM588 at one of three dose levels (4 × 10⁸, 1.2 × 10⁹, or 4 × 10⁹ CFU BID) in a 3+3 dose-escalation design. The primary endpoint was safety, with secondary endpoints including response rate and progression-free survival. Stool samples collected at baseline and Week 13 were analyzed by metagenomic sequencing to assess microbial composition across CBM588 dose levels. Plasma samples collected at baseline and Cycle 3 were analyzed for soluble mucosal addressin cell adhesion molecule-1 (sMAdCAM-1), a biomarker associated with dysbiosis (Fidelle et al Science 2023). Results: Twelve pts were treated between 6/26/2024 and 9/15/2025. Median age was 69 (range 48-80). All had clear cell histology without sarcomatoid features; two had rhabdoid features. Half the cohort had ≥3 metastatic sites, most commonly involving the lung, lymph nodes, and contralateral kidney. Most patients had favorable (50%) or intermediate (42%) IMDC risk. No dose-limiting toxicities (DLTs) were observed at any dose level, and no dose-dependent increase in immune-related toxicity was noted with CBM588 escalation. Notably, Bacteroides thetaiotaomicron showed a dose-dependent increase in relative abundance across escalating CBM588 levels (P = 0.0043 for Level 3 vs Level 1 and P = 0.0047 for Level 3 vs Level 2, Mann–Whitney U), suggesting a selective modulation of gut microbial composition. Among 11 pts with evaluable disease, 3 were responders (1 complete response, 2 partial responses), 6 had stable disease, and 2 had progressive disease. Baseline sMAdCAM-1 levels were similar between responders (Rs) and non-responders (NRs) (median: 205,196 vs. 206,462 pg/mL). However, percent change from baseline to Cycle 3 demonstrated a diverging pattern, with an early increase in Rs and relative stabilization in NRs (median: +8.2% vs –0.1%). Conclusions: No DLTs were encountered with CBM588 capsules at the highest dose level (MO-03) in combination with nivo/ipi, supporting the dosing strategy in the upcoming phase III SWOG study S2419 (BioFront). Translational studies indicate a dose effect upon microbiome composition, and sMAdCAM-1 dynamics suggest a potential association with response. Clinical trial information: NCT06399419 .