Multi-modal single-cell profiling of sarcomas from archival tissue reveals mechanisms of resistance to immune checkpoint inhibitors.

e23518 Background: Single-cell RNA-seq is an enabling technology that may inform the molecular underpinnings of drug response and resistance in patient biopsies. This method is difficult to implement in the study of rare diseases such as sarcomas due to specimen requirements and technical limitations. Methods: Here, we evolved novel methods that we recently reported in melanoma ( Wang, Fan, et al., bioRxiv, 2022), which enable single-nucleus RNA, T cell receptor (snRNA/TCR)-seq, and pool-matched whole-genome sequencing (WGS) from archival, frozen sarcoma tissue. Results: This enabled profiling of 75,716 cells and 788 matched TCR clonotypes from six patients with intimal sarcoma (INS) and undifferentiated pleomorphic sarcomas (UPS), including two matched pair samples from pre/post-immune checkpoint inhibitor (ICI). Our analysis revealed substantial transcriptional cancer cell heterogeneity driven by varying copy number alterations (CNAs). In one patient with INS with a complete response to ICI followed by an isolated recurrence, we identified a rare cancer cell clone defined by CNA (confirmed with WGS) and resulting transcriptional outputs that pre-existed and emerged during resistance. Furthermore, in a patient with UPS with intrinsic resistance to ICI, we find adequate T cell clonal expansion and activation, suggesting appropriate T cell response to ICI dampened by intrinsic mechanisms of ICI resistance within the cancer cells. Non-negative matrix factorization (NMF) analysis identified cell states associated with either intrinsic or adaptive resistance to ICI that was distinct from resistance to doxorubicin. These observations are consistent with those previously reported from sequential biopsies obtained from KEYNOTE-001 in metastatic melanoma ( Wang, Fan, et al., bioRxiv, 2022), which also revealed emergence of pre-existing populations of resistant clones defined by their underlying aneuploidy patterns. Conclusions: Together, these results demonstrate feasibility of implementing single-cell genomics from archival tissue to study sarcoma and propel our understanding of drug resistance. Conceptually, this work suggests that large-scale CNAs may drive cell sub-populations associated with ICI resistance in sarcoma and in other diseases.