Assessment of short-chain fatty acid (SCFA) and circulating cytokine trends in a clinical trial of nivolumab/ipilimumab and CBM588 in metastatic renal cell carcinoma (mRCC).

4556 Background: Recent evidence suggests that blood and stool SCFA levels (including acetic acid and butyric acid) are associated with clinical outcomes with immune checkpoint blockade in melanoma (Nomura et al. JAMA Network Open 2020). Herein, we report trends in plasma SCFA and cytokine levels in a randomized phase I clinical trial of CBM588, a live bacterial product containing a butyric acid-producing bacterium, C. butyricum, in patients with mRCC treated with immune checkpoint blockade. Methods: Treatment naïve patients with clear cell and/or sarcomatoid mRCC and intermediate- or high-risk IMDC disease were randomized to nivolumab/ipilimumab or nivolumab/ipilimumab and CBM588 in a 1:2 fashion. Blood samples were collected at baseline and on treatment (week 12 +/- 5 weeks). Quantitative assessment of plasma SCFAs and cytokines was performed by liquid chromatography-mass spectrometry and a Human Cytokine 30-plex protein assay, respectively. Changes in SCFAs and cytokine levels were assessed using paired sample T-tests. Results: A total of 29 patients were randomized to the nivolumab/ipilimumab (n=19) and nivolumab/ipilimumab with CBM588 (n=10) arms. Objective response rates (ORR) with nivolumab/ipilimumab with and without CBM588 were 58% and 20%, respectively. No statistically significant change in SCFA levels was observed between baseline and on-treatment samples in either arm. Notably, however, a numerical increase in levels of acetic acid (414.3 vs 814.0 nmol/g) and isobutyric acid (7.9 vs 9.0 nmol/g) was seen in the nivolumab/ipilimumab and CBM588 arm. In contrast, both factors decreased in the nivolumab/ipilimumab arm (563.1 vs 401.1 nmol/g and 10.0 vs 8.8 nmol/g, respectively). In patients with an uptrend in isobutyric acid levels with nivolumab/ipilimumab and CBM588, ORR was 64% compared to 50% in those without an uptrend in such levels). Similarly, patients with increases in butyric acid with nivolumab/ipilimumab and CBM588 had an ORR of 66%, compared to 43% among those without an increase in butyric acid levels. Correlations between acetic acid and chemokines, including CCL2 and CCL4, were observed with correlation coefficients of 0.83 and 0.82, respectively (p<0.001 for both). Conclusions: A substantial increase in isobutyrate levels in CBM588 responders is indicative of altered amino acid metabolism (specifically, valine and leucine) by modified gut flora in these patients. These observations, combined with alterations in serum cytokines, provide biologic rationale for the clinical activity seen with CBM588 in combination with nivolumab/ipilimumab in this trial. Clinical trial information: NCT03829111 .