Ailing Cao, Yaobin Lin, Shaoxing Guan, Youhao Chen, Wenyu Zhai, Yuheng Zhou, Shoucheng Feng, Yanping Guan, Yiyu Zhang, Min Huang, Xueding Wang, H. Long
BACKGROUND: Neoadjuvant anti-programmed cell death 1 (PD-1) immunochemotherapy has shown promising efficiency in the treatment of early-stage non-small-cell lung cancer (NSCLC), but it has not consistently yielded durable responses. Biomarkers for the prediction of efficacy are warranted. METHODS: We performed shotgun metagenomic and plasma/faecal metabolomic studies in 44 NSCLC patients who underwent neoadjuvant tislelizumab plus platinum-based doublet chemotherapy. Samples were collected at baseline and before surgical resection, and the major pathologic response (MPR) was evaluated. RESULTS: MPR patients showed a significantly higher gut-microbial alpha diversity, an enrichment of Ruminococcaceae, Lachnospiraceae and Clostridiales species, and an increased plasma level of tryptophan metabolites at baseline. On the contrary, non-MPR patients were characterized by enrichment of Prevotella species in faecal samples and higher plasma levels of linoleic acid metabolites. A high predictive accuracy was achieved using a small panel of differential microbial (Clostridium sp. M62/1 and Eisenbergiella tayi) or metabolomic features (linoleic acid, oxindole-3-acetic acid and quinolinic acid) with AUCs > .85. CONCLUSIONS: The baseline characteristics of the gut microbiota and plasma metabolites could provide early predictions of the response to neoadjuvant anti-PD-1 immunochemotherapy. TRIAL REGISTRATION: NCT05244837. KEY POINTS: Baseline metagenomic and metabolomic signatures were significantly associated with the major pathologic response of neoadjuvant anti-PD-1 immunochemotherapy. Integrated microbial model (consists of Clostridium sp. M62/1 and Eisenbergiella tayi) and metabolomic model (consists of linoleic acid, oxindole-3-acetic acid and quinolinic acid) could provide early predictions of the response.