FAM111B, LAG3, and FGL1, A Three-Gene Signature as a Predictive Biomarker for Immunotherapy Response in Bladder Urothelial Carcinoma

<title>Abstract</title> <p>Background Immunotherapy has markedly extended survival in some patients with bladder urothelial carcinoma (BLCA). However, only a subset of patients experience sustained clinical benefit. Thus, the identification of reliable biomarkers for predicting immunotherapy response are required. Methods Transcriptomic and clinical data from the IMvigor210 cohort treated with anti–PD-L1 therapy were analyzed and validated using the TCGA cohort. FAM111B, LAG3, and FGL1 expression were evaluated in relation to immunotherapy response, prognosis, tumor immune microenvironment characteristics, genomic features, and predicted chemotherapeutic sensitivity. A three-gene nomogram was constructed to predict immunotherapy response. Additionally, multiplex immunofluorescence was performed using a BLCA tissue microarray, and immunohistochemistry (mIF) was conducted using human and murine BLCA tumor tissues to validate protein expression patterns at the tissue level. Results BLCA patients with high FAM111B and LAG3 and a low FGL1 expression exhibited a better response to PD-L1 blockade and displayed a significantly prolonged overall survival. This three-gene signature was associated with an immune-inflamed tumor microenvironment characterized by increased effector immune cell infiltration, a higher tumor mutational burden, activation of immune-related pathways, and enhanced sensitivity to multiple chemotherapeutic agents. Quantitative data from human tissue samples by mIF showed that about 8.33–10.00% of BLCA patients had high FAM111B and LAG3 expression with low FGL1 expression, potentially indicating significantly better immunotherapy response. Conclusions When examining FAM111B, LAG3, and FGL1, the detected three-gene-based signature was associated with an immune-inflamed tumor microenvironment and found to be more therapy-responsive. Thus, this three-gene profile holds potential as a combinatorial biomarker to inform personalized therapeutic approaches involving PD-L1 blockade therapy.</p>