Quantifying MLH1 promoter methylation via liquid biopsy.

Abstract

25 Background: MLH1 is one of the primary genes involved in mismatch repair (MMR). Promoter methylation of MLH1 is known to cause abnormally low expression of MLH1 resulting in deficient MMR. MLH1 promoter methylation status is informative for deciding when to screen for Lynch Syndrome as well as for characterizing the etiology of mismatch repair deficiency. While tissue-based assays for MLH1 promoter methylation are commercially available, using liquid biopsy to detect MLH1 promoter methylation is a promising non-invasive approach. Methods: The number of methylated molecules in the MLH1 promoter region was measured in 3,389 clinical samples using Northstar Response, a pan-cancer, methylation-based, liquid biopsy assay that is CLIA/CAP validated for treatment monitoring. MLH1 promoter methylation was also measured in 176 samples from 124 self-reported cancer-free subjects 50 years or older to determine the limit of blank. A threshold of 3 methylated molecules was used to determine MLH1 promoter region methylation status. To orthogonally assess concordance, MLH1 promoter methylation was compared against microsatellite instability (MSI-high) calls from Northstar Select, a CLIA/CAP validated pan-cancer comprehensive genomic profiling liquid biopsy assay, as well as tumor results if available. Results: 1,861 samples had both MLH1 promoter methylation and MSI results, with 55% of MSI-high samples (11/20) having MLH1 promoter methylation. MLH1 promoter methylation had 99.9% specificity with only one sample called as microsatellite stable but having MLH1 promoter methylation, resulting in a positive predictive value of 92% (11/12 samples). The log10 of the number of MLH1 promoter methylated molecules was highly correlated with MSI score (R = 0.93). Liquid MLH1 promoter methylation had strong concordance with MLH1 immunohistochemistry (9/9 patients) and MLH1 tissue methylation (8/8 patients). No methylated molecules were observed in any sample from cancer-free subjects. The prevalence of MLH1 promoter methylation was 1.8% (9/498) among patients with colorectal cancer in our cohort. Conclusions: MLH1 promoter methylation was quantified using the Northstar Response liquid biopsy assay, and we observed high concordance with MSI-high as measured with a separate liquid biopsy assay as well as with MLH1 immunohistochemistry and methylation via tumor testing. The prevalence of MLH1 promoter methylation in colorectal cancer is comparable to previously reported tissue and liquid measurements. Quantifying MLH1 promoter methylation via liquid biopsy over time may provide additional insight into a patient’s personalized tumor biology over the course of treatment.