Verification of a Cartridge-based Microsatellite Instability (MSI) Platform and Cross-platform Assessment of Cases with Rare Germline Polymorphic Markers


Session type:


Gareth Gerrard1,Diana Pelka2,Nicole Gurunlian1,Georgina Briggs1,Jack Grant1,Kevin Balibi1,Phil Bennett1
1Sarah Cannon Molecular Diagnostics,2HCA Healthcare UK



Microsatellite instability (MSI) testing is an important component of the Lynch Syndrome (LS) pre-screen for colorectal and endometrial cancers, and a pan-cancer biomarker for response to checkpoint-inhibitor immunotherapy. We previously validated the Promega MSI v1.2 system, based around the capillary electrophoretic analysis of 5 quasi-monomorphic alleles in tumour-normal pairs. However, Biocartis have recently launched the Idylla MSI CE-IVD solution, based on single-use cartridge qPCR analysis of 7 homopolymer regions in tumour-only samples. We sought to verify this platform for clinical-diagnostic use, using multi-level synthetic controls; also to assess 3 samples that had previously shown germline polymorphisms in Promega alleles.


Seven MSI DNA controls (5 levels MSI-High (MSI-H); 2 MSI-Stable (MSS)) were acquired from Horizon Discovery. These were assessed with both systems, using standard protocols and scored as MSI-H (2-or-more alleles unstable), MSI-Intermediate (1 allele), or MSS (0 alleles). For the Promega system, samples were analysed using an ABI 3500 and GeneMapper v5 software; for Biocartis, 50ng DNA was used (1 cartridge per sample), with on-platform analysis. Analysis of the lowest signal-strength control was replicated 3 times on the Biocartis, with multiple operators to assess inter-run variability. Three tumour samples with previously identified germline polymorphic alleles (ovarian, BAT-25; endometrial, NR-21; colorectal, BAT-25; all MSS) were also assessed using the Biocartis system.


Control material results were 100% concordant between the Promega and Biocartis platforms (5 MSI-H; 2 MSS). Replicated samples showed no significant differences between analytical runs (MSI-H; 5/7 markers positive; ANOVA, P=0.53). The three samples displaying germline polymorphisms were all correctly reported as MSS (0/7 markers) by the Biocartis system.


Both platforms proved to be sensitive and cost-effective; however, the Biocartis benefited from a faster and simplified workflow and apparent reduced susceptibility to targets displaying potentially interfering polymorphism at the population level, thus obviating the requirement for co-analysis of paired-normal samples.