Biomarkers Beyond High TMB or MMR-D May Be Needed for ICIs in Sarcomas

Sarcomas with high tumor mutational burden (TMB) and mismatch repair deficiency (MMR-D) demonstrated varying degrees of responses to immune checkpoint inhibitors (ICIs), but additional biomarkers may contribute to prognosis of response, according to results from a single-institution retrospective series published in JCO Precision Oncology.¹

Among 17 patients with metastatic sarcoma meeting tumor-agnostic eligibility, objective responses occurred in 6 patients (35%), including 3 complete responses (CRs) and 3 partial responses (PRs); 2 patients had stable disease and 9 had primary progression, but responses did not scale by TMB level, and MMR-D response rates were lower than those reported across carcinomas.

“Our results confirm activity of ICI in some sarcomas and suggest that TMB and MMR-D status alone do not predict response but may operate in the context of the sarcoma subtype and other clinical or genomic features,” the investigators stated in their paper.

Background of Tumor-Agnostic ICI Treatment

Pembrolizumab (Keytruda) and dostarlimab (Jemperli) have received solid tumor-agnostic approvals for solid tumors with MMR-D status, with pembrolizumab also receiving this indication for high TMB status. However, much of the data supporting these indications come from carcinomas, with less data in sarcomas because of the rarity of these tumors that meet the FDA eligibility criteria. The likelihood of response to ICIs in patients with these biomarkers is also known to vary by disease type and within a single tumor type.

Study Design

In the single-center study, 17 patients at Dana-Farber Cancer Institute were identified by retrospective chart review among adults with metastatic sarcoma of any histologic subtype, confirmed by an expert sarcoma pathologist, with TMB of 10 mut/Mb or greater or MMR-D status by next-generation sequencing (NGS). Biomarker status was determined primarily from pretreatment tumor tissue, with 1 case assessed by liquid biopsy, using commercial or in-house next-generation sequencing panels. Response was assessed using published real-world response criteria, based mainly on CT scans performed roughly every 6 weeks, with PET-CT scans used for angiosarcoma of bone.

Among the 17 patients aged 30 to 73 at diagnosis, ICI was given as first-line therapy in 5 patients, second-line in 6 patients, and third-line or later in 6 patients. Fifteen received pembrolizumab monotherapy, 1 received eribulin plus pembrolizumab, and 1 received ipiliumab (Yervoy) plus nivolumab (Opdivo). They had a wide distribution of rare sarcomas including angiosarcoma, leiomyosarcoma, and high-grade unclassified.

MMR-D Responses Diverge From Carcinoma Benchmarks

TMB across the cohort ranged from 9 to 151 mut/Mb, and TMB among responders ranged from 15 to 75 mut/Mb, with no apparent scaling of response by TMB level; 1 patient with low TMB had stable disease and 9 with high TMB had progressive disease. MMR-D status was similarly uncoupled from response, with only 1 of 6 patients with MMR-D responding. The responding patient had Lynch syndrome and was the only patient with a germline MMR-D mutation.

This split is notable given the predicted 53% response rate to ICIs across MMR-D/microsatellite instability-high cancers, the study authors observed.² They noted that this pattern is consistent with a prior report on the sensitivity of Lynch syndrome–associated sarcomas to immunotherapy,³ and they raised the possibility that lower TMB, co-occurring genetic alterations, or insufficient neoantigen immunogenicity may explain the lack of benefit in sporadic MMR-D sarcomas specifically, although the small sample size precludes definitive conclusions.

Response Tracked With Histology and UVA Signature, Not Biomarker Status Alone

The 3 CRs occurred in MMR-proficient tumors, specifically 2 pleomorphic dermal sarcomas (PDS) and 1 cutaneous angiosarcoma, each with a TMB of 22 to 45 mut/Mb and an ultraviolet A (UVA) mutational signature. Duration of ICI treatment among responders ranged from 1.5 months, in a case with up-front progression, to 17 months, and 1 patient with PDS has remained in complete remission for more than 2 years after discontinuing pembrolizumab. The study authors noted that cutaneous angiosarcoma and PDS are already recognized as sarcoma subtypes with comparatively high ICI response rates independent of TMB or MMR-D status. They added that the combination of UVA signature and high TMB may predict response better than TMB alone, suggesting the biomarker-agnostic responses tracked more closely with histology and mutational signature than with meeting the FDA eligibility threshold itself.¹

Clinical Implications

The authors concluded that the sarcomas responding to ICI in this series generally displayed already-established, sarcoma-specific markers of sensitivity, namely histology, UVA signature, or Lynch syndrome, rather than TMB or MMR-D status on their own, which they said calls into question the standalone utility of tumor-agnostic biomarker criteria developed largely in carcinomas when applied to mesenchymal tumors. They noted that additional variables not assessed in this series, including microsatellite instability status, tertiary lymphoid structures, and cytotoxic T-cell infiltration, likely also influence ICI response in sarcomas, and they called for identification of additional or combined biomarkers to better stratify patients for ICI treatment in this setting.

REFERENCES

1. Alcindor T, Papke DJ Jr, Venkataraman V, et al. Patterns of response to immune checkpoint inhibitors in sarcomas with high tumor burden and/or mismatch repair deficiency. JCO Precis Oncol. 2026;10:e2600414. doi:10.1200/PO-26-00414

2. Le DT, Durham JN, Smith KN, et al. Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade. Science. 2017;357(6349):409-413. doi:10.1126/science.aan6733

3. Poumeaud F, Valentin T, Fares N, et al. Sarcomas developed in patients with Lynch Syndrome are enriched in pleomorphic soft-tissue sarcomas and are sensitive to immunotherapy. Eur J Cancer. 2025;216:115196. doi:10.1016/j.ejca.2024.115196