Association of progression-free survival (PFS) with microsatellite instability–high (MSI-H) or mismatch repair–deficient (dMMR) status determined by single vs dual validated companion diagnostic (CDx) assays: CheckMate-8HW post hoc analyses.

Abstract

3531 Background: First-line (1L) nivolumab (NIVO) + ipilimumab (IPI) showed PFS benefit vs chemotherapy (chemo) in patients (pts) with MSI-H or dMMR metastatic colorectal cancer (mCRC) in the phase 3 CheckMate 8HW trial. Using post hoc analyses, we examined the association of PFS with single and dual positivity in validated CDx MSI (polymerase chain reaction; PCR) or MMR (immunohistochemistry; IHC) central assays. Methods: Study details were published previously (Andre T et al. NEJM 2024). Pts with unresectable or mCRC with MSI-H or dMMR per local testing were randomized 2:2:1 to NIVO + IPI, NIVO, or chemo. Tumor tissue samples were tested centrally using validated Idylla (CDx MSI PCR; Biocartis) and Dako (CDx MMR IHC; Agilent) assays post randomization. Clinical trial assay positive (CTA+) pts were those with locally diagnosed MSI-H/dMMR mCRC randomized to NIVO + IPI or chemo. PFS and hazard ratios (HR) were evaluated in CTA+ pts; pts positive using a single central assay (CDx MSI or CDx MMR); and pts positive in both central assays (CDx dual positive). Results: Of the 303 pts randomized to 1L NIVO + IPI or chemo, 301 had locally diagnosed MSI-H/dMMR (CTA+; 2 randomized pts missed local testing and were excluded in the analysis). 263 pts had valid CDx MSI results, 284 had valid CDx MMR results; 208 were CDx dual positive (Table); result missingness was mainly due to lack of tissue samples. PFS benefit was noted with NIVO + IPI vs chemo in CTA+ pts (HR, 0.32; Table). Greater PFS benefit with NIVO + IPI vs chemo was seen in pts positive with single CDx MSI or CDx MMR assays (HR, 0.20 and 0.22, respectively), and was consistent with those from CDx dual positive pts (HR, 0.20; Table). Pts with microsatellite stable (MSS) or mismatch repair proficient (pMMR) status were low, so these data should be interpreted with caution (Table). Conclusions: In this post hoc analysis, dual positivity from CDx MSI and CDx MMR assays did not offer greater PFS benefit with 1L NIVO + IPI vs chemo over positivity from a single CDx MSI or MMR assay. Single CDx MSI or CDx MMR assays provided similar PFS benefit with 1L NIVO + IPI vs chemo in pts with MSI-H/dMMR mCRC. Clinical trial information: NCT04008030 . NIVO + IPI(n = 202) Chemo(n = 101) HR (95% CI) a n (%) Median PFS (95% CI) n (%) Median PFS (95% CI) CTA+ 200 (99) NR (34.3–NE) 101 (100) 6.2 (4.7–9.0) 0.32 (0.22–0.45) CDx b MSI+: MSI-H 147 (73) NR (38.4–NE) 71 (70) 6.2 (4.7–9.0) 0.20 (0.12–0.31) CDx b MSI−: MSS 30 (15) 1.8 (1.5–5.8) 15 (15) 7.4 (4.0–12.9) 1.51 (0.68–3.33) CDx b MMR+: dMMR 163 (81) NR (38.4–NE) 82 (81) 5.9 (4.4–7.8) 0.22 (0.14–0.34) CDx b MMR−: pMMR 27 (13) 1.8 (1.5–5.8) 12 (12) 11.5 (2.0–NE) 1.39 (0.57–3.40) CDx b : dual positive 139 (69) NR (38.4–NE) 69 (68) 6.2 (4.9–9.2) 0.20 (0.13–0.32) a Cox proportional hazards model stratified by tumor sidedness. b By central assay. NE, not estimable; NR, not reached.