Cancer Research · asia
NK/T-Cell Lymphoma Has More Than One Face: Single-Cell Atlas Identifies Four Tumor Programs
Multi-omics analysis of 63 samples shows that tumors can exhibit distinct states involving immune response, proliferation, inflammation, or metabolism, potentially offering stratification clues for immunotherapy and epigenetic therapy; however, this classification has not yet been validated in prospective clinical trials.
Although NK/T-cell lymphomas share a single diagnostic label, the way these tumors function internally may differ substantially. A study published in *Nature Communications* used single-cell and multi-omics technologies to map their cellular ecosystems, identifying four recurrent tumor gene-expression programs in an effort to explain differences in disease course and treatment response among patients.
The research team analyzed 63 samples, using single-cell RNA sequencing as the foundation and integrating spatial transcriptomic, bulk tissue transcriptomic, proteomic, and metabolomic data. This approach examined not only which genes the tumor cells expressed, but also the immune environments in which they were located and the potential interactions among different cell types.
The first program, MP1, showed stronger immune-response features, but the surrounding immune cells simultaneously exhibited an exhausted state, and tumor PD-L1 expression was also higher. Based on these findings, the researchers hypothesized that this group of tumors might have a greater chance of benefiting from immune checkpoint inhibition. However, PD-L1 and immune exhaustion provide only biological clues and are not equivalent to validated predictive biomarkers of treatment efficacy.
MP2 was dominated by cell proliferation, displaying an immune-cell-poor “immune desert” accompanied by elevated HDAC2 and the proliferation marker MKI67. This finding points to epigenetic regulation as a potential therapeutic entry point, such as further testing mechanisms related to histone deacetylases, but the study itself did not demonstrate that such drugs improve patient outcomes.
The other two programs outlined distinct disease ecosystems. MP3 was dominated by myeloid immune cells, showed activation of JAK/STAT signaling, and was associated with a more aggressive clinical course. MP4, meanwhile, exhibited a distinctive amino acid metabolic profile, and tertiary lymphoid structures were also more common within these tumors. The latter are clusters of immune cells within tumors that resemble lymphoid tissue. Their presence may affect immune responses, but their therapeutic significance in NK/T-cell lymphoma remains to be clarified.
These four “metaprograms” are better viewed as a research framework for tumor states rather than clinical subtypes that can already be used to directly prescribe different drugs. The study has not demonstrated through prospective trials that selecting drugs according to these groupings can prolong survival, and some tumors may simultaneously harbor multiple programs. In addition, the version currently available is an early, unedited version. Further work is needed to establish a classification method suitable for routine specimens and to validate it in independent patient cohorts using treatment-response data.