Host genetic variants in the lactose metabolism pathway associated with bifidobacterium abundance have prognostic value for immunotherapy efficacy

BACKGROUND AND PURPOSE: The potential predictive and therapeutic value of Bifidobacterium in immune checkpoint inhibitor (ICI) treatment has been widely studied. However, the value of its genetic determinants on the prognosis of ICI treatment remains unclear. EXPERIMENTAL APPROACH: We examined the associations between 11 single nucleotide polymorphisms (SNPs) located at host genes associated with Bifidobacterium abundance and the outcomes of ICI treatment in 370 eligible Chinese cancer patients. KEY RESULTS: LCT/MCM6 SNPs, rs3739020, rs3739022, rs56263017 and rs55809728, were strongly associated with overall survival (OS), and the LCT haplotype analysis showed that genetically predicted lactose poor metabolisers exhibited significantly poorer OS (P-value = 0.002, adjusted HR = 0.45, 95%CI = 0.27-0.74). In polygenic SNP analysis, the genetically inferred high galactose level carriers exhibited significantly prolonged OS (P-value = 0.007, adjusted HR = 0.32, 95%CI = 0.14-0.73) and progression-free survival (PFS, P-value = 0.001, adjusted HR = 0.45, 95%CI = 0.29-0.71). All four SNPs and the lactose metabolic phenotype were not associated with the occurrence of overall immune-related adverse events (irAEs). Two-sample mendelian randomisation analysis showed genetically predicted Bifidobacterium abundance correlated with lactose metabolism pathway abundance (P value = 0.012, Beta coefficient = 0.549). CONCLUSIONS AND IMPLICATIONS: Lactose metabolism pathway SNPs, which are linked to Bifidobacterium abundance in prior studies, have prognostic value for immunotherapy efficacy in Chinese cancer patients, and the genetically predicted lactose extensive and intermediate metabolisers exhibited better immunotherapy efficacy. These findings remain exploratory and require further validation with direct microbiome and metabolite measurements.