FOLR2⁺ macrophages and CAFs: coordinating immunosuppression in gastric cancer

Dear editor, We read with great interest the article by Xue et al, titled “Disrupting of tissue-resident FOLR2+ tumor-associated macrophages-derived interleukin-10 improves anti-PD-1 efficacy in gastric cancer”[1]. The authors elegantly demonstrate that FOLR2⁺ tissue-associated macrophages (TAMs) impair CD8⁺ T-cell function by secreting PD-L1 and IL-10, thereby promoting the formation of an immunosuppressive microenvironment in gastric cancer (GC). They further show that dual blockade of PD-1 and IL-10 enhances antitumor immunity in FOLR2⁺TAMs-high tumors, providing a compelling rationale for a novel combination immunotherapy in GC. We believe that there are several aspects worth further discussion. This study complies with the TITAN Guidelines 2025 – governing declaration and use of AI[2]. While the role of TAMs is increasingly recognized, the tumor microenvironment of GC constitutes a complex ecosystem comprising various stromal cells, including cancer-associated fibroblasts (CAFs), which also significantly contribute to immunomodulation and therapeutic resistance[3,4]. To further contextualize the findings by Xue et al, we re-analyzed the public dataset GSE194261, which profiled genome-wide gene expression in CAFs or normal gastric fibroblasts derived from scirrhous or non-scirrhous GC tissues (Fig. 1). The transcripts most markedly upregulated in scirrhous GC-associated fibroblasts include WNT5A, GREM1, SPOCK1, and CCL11, which are involved in extracellular matrix organization, TGF-β signaling, and complement activation – pathways known to contribute to immune evasion and T cell dysfunction[5–7]. Gene Ontology analysis further revealed that CAFs in scirrhous GC were significantly enriched for biological processes related to metal-ion response and homeostasis, including cellular response to cadmium and copper ions. These programs are consistent with a metabolically active, stress-responsive stromal population capable of modulating redox balance and growth factor signaling in the microenvironment. These transcriptional features are consistent with previous experimental studies showing that multiple stromal components – both macrophages and fibroblasts – may converge on similar immunosuppressive mechanisms, including activation of the IL-10 and PD-L1/PD-1 axis[8,9]. Figure 1.: Differential gene expression and functional enrichment of scirrhous gastric cancer–associated fibroblasts in GSE194261. A and b, Volcano plot and heatmap of differentially expressed genes between gastric fibroblasts from cancerous tissues of scirrhous gastric cancer and fibroblasts from non-cancerous tissues of non-scirrhous gastric cancer. C and d, Gene Ontology biological process enrichment bar plot and bubble plot for upregulated genes. E and f, KEGG pathway enrichment bar plot and bubble plot for upregulated genes. These data complement the findings of Xue et al. First, combining FOLR2+ TAM features with a CAF-related gene set may improve risk stratification beyond the use of macrophage markers alone, identifying a “stroma-myeloid high” subgroup with the worst prognosis and poorest response to PD-1 monotherapy. Second, GCs enriched in CAFs often exhibit activation of TGF-β and other pathways associated with immune rejection[10], suggesting that dual PD-1/IL-10 blockade may be particularly effective when combined with CAF-modulating approaches (e.g., TGF-β signaling inhibitors). Such a strategy could suppress CAF-driven immunosuppression and unleash the antitumor potential of reactivated CD8+ T cells. In conclusion, Xue et al provide compelling evidence that targeting IL-10 in the context of FOLR2⁺ tissue-resident macrophages can enhance PD-1 blockade in GC. We believe that incorporating the fibroblast compartment into future mechanistic and clinical studies could further strengthen and extend the authors’ conclusions, ultimately supporting more precise stromal-guided immunotherapy strategies for patients with GC. Ethical approval None.