Abstract
BACKGROUND: Immune checkpoint inhibitors (ICIs) have been a breakthrough in cancer immunotherapy, but secondary resistance (SR) and immune-related adverse events (irAEs) are significant clinical dilemmas. Although the gut microbiota is associated with ICI efficacy and irAEs, the knowledge of longitudinal gut microbiota dynamics during SR and irAE development is still quite limited. METHODS: This was a prospective observational cohort study of cancer patients initially receiving anti-programmed cell death-1 (PD-1) treatment between May 2020 and October 2022. Clinical information was collected to evaluate therapy response and AEs. Patients were divided into a secondary resistance (SR) group, a non-secondary resistance (NSR) group, and an irAE group. Fecal samples were longitudinally obtained from baseline across multiple timepoints and analyzed with 16S rRNA sequencing. RESULTS: Thirty-five patients were enrolled, and 29 were evaluable. After a median follow-up of 13.3 months, NSR patients had a favorable progression-free survival (PFS) compared with SR (457.9 IQR 241.0-674.0 days vs. 141.2 IQR 116.9-165.4 days, P=0.003) and irAE patients (457.9 IQR 241.0-674.0 days vs. 269.9, IQR 103.2-436.5 days, P=0.053). There were no significant differences in the microbiota between groups at baseline. Several previously reported beneficial microbiomes for ICI efficacy including Lachnospiraceae, Ruminococcaceae, Agathobacter, and Faecalibacterium showed decreasing trends as secondary resistance developed, yet not achieved significance (P>0.05). Significant changes in butyrate-producing bacteria were also presented in the SR cohort (P=0.043) with a decreasing trend upon secondary resistance occurrence (P=0.078). While the abundance of IgA-coated bacteria was stable in the SR cohort, there was a temporary decrease upon ICI treatment initiation and reestablishment after continuation of ICI treatment in the NSR cohort (primary ICI response: 0.06, IQR 0.04-0.10; durable ICI response: 0.11, IQR 0.07-0.14; P=0.042). Bacteroides contributed most to the difference between baseline and irAE occurrence, which decreased after irAE occurrence (Baseline: 0.10 IQR 0.07-0.36; irAE occurrence: 0.08 IQR 0.06-0.12) and was restored upon irAE remission to a comparable level as baseline (irAE remission: 0.10 IQR 0.09-0.18). CONCLUSIONS: The development of SR and irAEs is related to the longitudinal dynamics of the intestinal microbiota. The investigation into the preventative and protective effects of enteric microbe manipulation strategies is further required.