XiaoYan Huang, ChunHong Xiao, WeiXuan Hong, JunWei Fang, Xi Chen, ZhiRong Yang, LiFeng Cai, JiaQi Huang
Immunochemotherapy effective, well‑tolerated, improves survival and QoL. Early immune/microbiota changes associate with benefit. CD3⁺ T, CD4⁺/CD8⁺, Th1/Th2, NK may serve as correlative markers for early response discrimination, but not pre‑treatment predictors due to no baseline differences and limited bacterial data, yet they are useful for efficacy assessment.
OBJECTIVE: To evaluate efficacy/safety/survival of PD‑1 inhibitor plus chemotherapy in advanced gastric cancer, and analyze early changes in immune cells and gut bacteria (Lactobacillus, E. coli, Bifidobacterium, Enterococcus) and their association with response.
METHODS: Prospective cohort (n = 100). Responders (CR/PR/SD, n = 74) vs non‑responders (PD, n = 26) by RECIST version 1.1. Immune and bacterial levels at baseline and post‑treatment. Logistic regression and ROC.
RESULTS: ORR = 36.0%, DCR = 74.0%. Median follow‑up = 8.0 mo; PFS = 11.0 mo. Responders had longer PFS (11.73 vs 2.00 months) and OS (15.00 vs 8.00 months), both P = 0.001. AEs occurred in 42% of patients (grade 1‑3) and were tolerable; KPS and QoL improved in 57 patients. No baseline differences. Post‑treatment, responders showed increased levels of CD3⁺ T, CD4⁺ T, CD4⁺/CD8⁺, Th1/Th2, Lactobacillus, and Bifidobacterium, and decreased levels of CD8⁺ T, NK, Treg, E. coli, and Enterococcus. (all P < 0.05); non‑responders showed only mild increases in Th1/Th2, Lactobacillus, and E. coli. Between groups, responders had higher levels of CD3⁺, CD4⁺/CD8⁺, Th1/Th2, Lactobacillus, and Bifidobacterium, and lower levels of NK, E. coli, and Enterococcus. Regression identified CD3⁺ T, CD4⁺/CD8⁺, Th1/Th2, and NK as independent factors; ROC highest AUC = 0.870 (NK).
CONCLUSION: Immunochemotherapy effective, well‑tolerated, improves survival and QoL. Early immune/microbiota changes associate with benefit. CD3⁺ T, CD4⁺/CD8⁺, Th1/Th2, NK may serve as correlative markers for early response discrimination, but not pre‑treatment predictors due to no baseline differences and limited bacterial data, yet they are useful for efficacy assessment.