Obesity enhances the response to neoadjuvant anti-PD1 therapy in oral tongue squamous cell carcinoma

Abstract

Objectives: Previous studies have demonstrated that obesity may impact the efficacy of anti-PD1 therapy, but the underlying mechanism remains unclear. In this study, our objective was to determine the prognostic value of obesity in patients with oral tongue squamous cell carcinoma (OTSCC) treated with pembrolizumab and establish a subtype based on fatty acid metabolism-related genes (FAMRGs) for immunotherapy.

Materials and methods: We enrolled a total of 56 patients with OTSCC who underwent neoadjuvant anti-PD1 therapy. Univariate and multivariate Cox regression analyses, Kaplan-Meier survival analysis, and immunohistochemistry staining were performed. Additionally, we acquired the gene expression profiles of pan-cancer samples and conducted GSEA and KEGG pathway analysis. Moreover, data from TCGA, MSigDB, UALCAN, GEPIA and TIMER were utilized to construct the FAMRGs subtype.

Results: Our findings indicate that high Body Mass Index (BMI) was significantly associated with improved PFS (HR = 0.015; 95% CI, 0.001 to 0.477; p = 0.015), potentially attributed to increased infiltration of PD1 + T cells. A total of 91 differentially expressed FAMRGs were identified between the response and non-response groups in pan-cancer patients treated with immunotherapy. Of these, 6 hub FAMRGs (ACSL5, PLA2G2D, PROCA1, IL4I1, UBE2L6 and PSME1) were found to affect PD-1 expression and T cell infiltration in HNSCC, which may impact the efficacy of anti-PD1 therapy.

Conclusion: This study demonstrates that obesity serves as a robust prognostic predictor for patients with OTSCC undergoing neoadjuvant anti-PD1 therapy. Furthermore, the expression of 6 hub FAMRGs (ACSL5, PLA2G2D, PROCA1, IL4I1, UBE2L6 and PSME1) plays a pivotal role in the context of anti-PD1 therapy and deserves further investigation.

Keywords: fatty acid metabolism; immunotherapy; obesity; oral tongue squamous cell carcinoma; prognosis.

FIGURE 1

The workflow of the current work.

FIGURE 2

Obese patients benefit from anti‐PD1 therapy by increasing infiltration of PD1+ T cells. (A, B) Kaplan–Meier plots of progression‐free survival (A) and overall survival (B) according to body mass index (BMI) group (BMI < 24 and BMI≥24) in patients with tongue squamous cell carcinoma. (C) Verification of BMI and immune cell infiltration in OTSCC (n = 27). immunohistochemical images show the immune cell infiltration (CD4+ T cell, CD8+ T cell, B cell, and PD1+ T cell) in OTSCC tissues. (D–G), human protein quantification analysis results of immunohistochemical staining by Qupath.

FIGURE 3

Fatty acid metabolism impacts anti‐PD1 therapy effects in pan‐cancer. (A–D) Result of Gene Set Enrichment Analysis (GSEA) between responders and non‐responder groups in melanoma (A), non‐small cell lung cancer (B), renal cell carcinoma (C) and stomach adenocarcinoma (D). (E–H) Result of Kyoto Encyclopedia of Genes and Genomes (KEGG) between responders and non‐responder groups in melanoma (E), non‐small cell lung cancer (F), renal cell carcinoma (G) and stomach adenocarcinoma (H).

FIGURE 4

Differentially expressed FAMRGs between responder and non‐responder samples. (A‐D) Volcano map of the differentially expressed genes in pan‐cancer. The red, green, and gray dots indicate high, low, and no difference in expression between responder and non‐responder samples (p < 0.05 & |log2 FC|>1). (E–H) Authentication of 91 FAMRGs in the four cancer datasets through Venn diagrams. (Melanoma (A), NSCLC (B), RCC (C), STAD (D)). (The upregulated genes were displayed in red, with downregulated FAMAGs in blue).

FIGURE 5

(A–F) Histogram of hub FAMRGs expression in 24 types of unpaired normal and normal tissues from TCGA using Wilcoxon rank‐sum test. (G–L) Histograms of hub FAMRGs in normal and HNSCC with significant differences from UALCAN portal.

FIGURE 6

Correlation of hub FAMRGs expression with anti‐PD1 therapy response and prognostic value in cancer samples. (A, B). ACSL5 and PLA2G2D expression levels between responder and non‐responder samples in melanoma. (C) IL4I1 expression levels between responder and non‐responder samples in non‐small cell cancer. (D) PROCA1 expression level between responder and non‐responder samples in renal cell carcinoma. (E, F) UBE2L6 and PSME1 expression level between responder and non‐responder samples in stomach adenocarcinoma. (G–L) The Kaplan–Meier curves of the low and high FAMRGs expression in pan‐cancer patients treated with anti‐PD1 therapy (n = 520).

FIGURE 7

FAMRGs affect anti‐PD1 therapy efficiency through T cells infiltration. Correlation analysis of hub FAMRGs (A, ACSL5, B, PLA2G2D, C, IL4I1, D, PROCA1, E, UBE2L6, F, PSME1) and 7 immune cells infiltration from TIMER in HNSCC samples, including B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils, dendritic cells. (G–L), Pearson correlation analysis between hub FAMRGs and PD‐1 level in HNSCC Patient characteristics.