High-plex spatial transcriptomic profiling reveals distinct immune components and the HLA class I/DNMT3A/CD8 modulatory axis in mismatch repair-deficient endometrial cancer

Abstract

Abstract Purpose Tumors bearing mismatch repair deficiency (MMRd) are characterized by a high load of neoantigens and are believed to trigger immunogenic reactions upon immune checkpoint blockade treatment such as anti-PD-1/PD-L1 therapy. However, the mechanisms are still ill-defined, as multiple cancers with MMRd exhibit variable responses to immune checkpoint inhibitors (ICIs). In endometrial cancer (EC), a distinct tumor microenvironment (TME) exists that may correspond to treatment-related efficacies. We aimed to characterize EC patients with aberrant MMR pathways to identify molecular subtypes predisposed to respond to ICI therapies. Methods We applied digital spatial profiling, a high-plex spatial transcriptomic approach covering over 1,800 genes, to obtain a highly resolved TME landscape in 45 MMRd-EC patients. We cross-validated multiple biomarkers identified using immunohistochemistry and multiplexed immunofluorescence using in-study and independent cohorts totaling 123 MMRd-EC patients and validated our findings using external TCGA data from microsatellite instability endometrial cancer (MSI-EC) patients. Results High-plex spatial profiling identified a 14-gene signature in the MMRd tumor-enriched regions stratifying tumors into “hot”, “intermediate” and “cold” groups according to their distinct immune profiles, a finding highly consistent with the corresponding CD8 + T-cell infiltration status. Our validation studies further corroborated an existing coregulatory network involving HLA class I and DNMT3A potentially bridged through dynamic crosstalk incorporating CCL5. Conclusion Our study confirmed the heterogeneous TME status within MMRd-ECs and showed that these ECs can be stratified based on potential biomarkers such as HLA class I, DNMT3A and CD8 in pathological settings for improved ICI therapeutic efficacy in this subset of patients.