The Role of MSI Testing Methodology and Its Heterogeneity in Predicting Colorectal Cancer Immunotherapy Response

Abstract

MSI is a crucial biomarker for selecting CRC patients for immunotherapy. Here, we analyze the first results from the observational prospective trial BLOOMSI (NCT06414304), which investigated the impact of MSI/dMMR testing methods and baseline tumor heterogeneity on treatment outcomes. Thirty MSI/dMMR+ CRC patients, who were candidates for immunotherapy, were enrolled. Depending on the local test used for MSI/dMMR, central PCR/IHC was performed. Baseline FFPE and liquid biopsy (LB) were analyzed with NGS. ORR (objective response rate) in the ITT population was 50% (95% CI, 31.3-68.7%). Concordance between local/central dMMR/MSI testing was 81%, and the concordance of IHC, PCR, NGS/FFPE, and NGS/LB was 68.4%. The ORR was similar for IHC+, PCR+, NGS/FFPE+, and NGS/LB+ patients (55.6%, 55.6%, 55%, and 57.9%, respectively). The ORR among patients with discordant IHC/PCR results was 0%, and the ORR among patients with NGS/LB-ORR was 25% (2/8 CR). Next, we performed quantitative MSI analysis, reflecting the clonality of MSI+ tumor cells. Multivariate analysis identified MSI clonality in FFPE (HR 0.63, 95% CI, 0.39-0.99, p = 0.0487) and LB (HR 3.05, 95% CI, 2.01-4.65, p < 0.00001) as independent predictors of progression. The ORR in patients with high clonality (≥7%, n = 4, NGS/LB) was 25%. We describe baseline methodological predictors of non-response to immunotherapy and propose a strategy for selecting potential non-responders. These findings warrant further investigation.

Keywords: MSI; colorectal cancer; dMMR; immunotherapy; liquid biopsy; next-generation sequencing.

Figure A1

Subgroup analysis of objective response rate (left) and progression-free survival (right). The gray line in the left forest plot indicates the ORR in the intention-to-treat population.

Figure A2

MSI clonality in FFPE and liquid biopsy (LB) samples based on tumor sidedness.

Figure A3

Oncoprints of somatic alterations identified following NGS analysis of the pre-treatment (A) FFPE and (B) liquid biopsy samples.

Figure 1

Study schema: (A) testing procedures; (B) treatment details; and (C) distribution of patients based on the different assays they received, showing how many patients were part of each assay group and how many received combinations of the assays. Patients with CRC were recruited based on the results of routine assessments of MSI/dMMR via gold-standard methods (IHC or PCR). Testing with an alternative gold-standard method (FFPE) and NGS (FFPE and/or LB) were performed centrally.

Figure 2

Concordance between the methods for MSI/dMMR detection used in this study.

Figure 3

Response to treatment according to the MSI/dMMR testing result.

Figure 4

Treatment efficacy of ICI according to MSI/dMMR testing results and BRAF status. Swimmer plot of the duration of treatment (in days) for each patient (n = 30). Patients marked with arrows were still under treatment as of July 2024. On the left side of the figure, information on the therapy setting, BRAF mutational status, and MSI/dMMR testing results is given for each patient.

Figure 5

Kaplan–Meier estimates of progression-free survival in patients based on (A) MSI positivity in FFPE and LB (NGS); (B) discordant MSI/dMMR results and the primary method used for MSI/dMMR testing; (C) BRAF mutations found in FFPE and/or LB; and (D) MSI clonality values in FFPE and LB.