Abstract
Background: Extranodal NK/T-cell lymphoma (NKTCL) is an aggressive non-Hodgkin lymphoma subtype. Despite improved outcomes with immunotherapy, persistent resistance in a subset of patients underscores an unmet need to delineate underlying molecular drivers. Lipid metabolic reprogramming provides energy and biosynthetic precursors for tumor growth while actively remodeling the tumor microenvironment to suppress immunity. Thus, cotargeting immune checkpoints and lipid metabolism represents a promising strategy to overcome resistance; however, its mechanistic basis and clinical applicability in NKTCL remain elusive.
Methods: Using complementary multiomics profiling—single-cell RNA sequencing (scRNA-seq; n = 13 patients), bulk RNA sequencing (RNA-seq; n = 24 patients), and metabolomics(n = 22 patients)—in the prospective SPIRIT cohort (NCT04127227), we identified reprogrammed lipid metabolism as a hallmark of immunotherapy resistance in NKTCL. To mechanistically dissect this, we deployed dual parallel systems: (1) therapy-resistant models and (2) CRISPR-based gene-editing platforms across in vitro settings and humanized mouse models. Functional validation was performed by leveraging these systems with multiomics analyses, including bulk RNA-seq, scRNA-seq, RNA immunoprecipitation sequencing (RIP-seq), methylated RNA immunoprecipitation sequencing (MeRIP-seq), and metabolomics.
Results: YTHDF3 deficiency stabilizes PI3K-activating transcripts (IRS1, IGF1R, PIK3R6), driving constitutive PI3K signaling and lipid metabolic reprogramming. This hyperactivated lipid metabolism induces endoplasmic reticulum stress (ERS), promoting ERS-mediated MHC-I downregulation to confer immunotherapy resistance. Translating these findings, we designed a dual-targeting strategy synergizing PI3K inhibition (to reverse metabolic immunosuppression) and anti-PD-1 therapy (to block immune checkpoint evasion), which demonstrated efficacy in murine models. This therapeutic strategy is now being evaluated clinically in a phase Ib trial (NCT06793956).
Conclusion: Our study delineates the YTHDF3–PI3K–lipid metabolism axis as a central driver of immunotherapy resistance in NKTCL. This mechanism-guided framework, with dual targeting of PI3K and PD-1, bridges preclinical validation to clinical translation (NCT06793956).
