High-Accuracy Determination of Microsatellite Instability Compatible with Liquid Biopsies.

BACKGROUND: Microsatellite instability (MSI) has recently emerged as a predictive pan-tumor biomarker of immunotherapy efficacy, stimulating the development of diagnostic tools compatible with large-scale screening of patients. In this context, noninvasive detection of MSI from circulating tumor DNA stands as a promising diagnostic and posttreatment monitoring tool. METHODS: We developed drop-off droplet-digital PCR (ddPCR) assays targeting BAT-26, activin A receptor type 2A (ACVR2A), and defensin beta 105A/B (DEFB105A/B) microsatellite markers. Performances of the assays were measured on reconstitution experiments of various mutant allelic fractions, on 185 tumor samples with known MSI status, and on 72 blood samples collected from 42 patients with advanced colorectal or endometrial cancers before and/or during therapy. RESULTS: The 3 ddPCR assays reached analytical sensitivity <0.1% variant allelic frequency and could reliably detect and quantify MSI in both tumor and body fluid samples. High concordance between MSI status determination by the three-marker ddPCR test and the reference pentaplex method were observed (100% for colorectal tumors and 93% for other tumor types). Moreover, the 3 assays showed correlations with r ≥ 0.99 with other circulating tumor DNA markers and their dynamic during treatment correlated well with clinical response. CONCLUSIONS: This innovative approach for MSI detection provides a noninvasive, cost-effective, and fast diagnostic tool, well suited for large-scale screening of patients that may benefit from immunotherapy agents, as well as for monitoring treatment responses.

Therapeutic potential of the human endogenous retroviral envelope protein HEMO: a pan-cancer analysis.

Human endogenous retroviruses represent approximately 8% of our genome. Most of these sequences are defective except for a few genes such as the ancestral retroviral HEMO envelope gene (Human Endogenous MER34 ORF), recently characterized by our group. In this study, we characterized transcriptional activation of HEMO in primary tumors from The Cancer Genome Atlas (TCGA) and in metastatic tumors from a Gustave Roussy cohort. Pan-cancer detection of the HEMO protein in a series of patient samples validated these results. Differential gene expression analysis in various TCGA datasets revealed a link between HEMO expression and activation of Wnt/ß-catenin signaling, in particular in endometrial cancer. Studies on cell models led us to propose that the Wnt/ß-catenin pathway could act as an upstream regulator of this retroviral endogenous sequence in tumor condition. Characterization of transcriptomic profiles of both HEMOLow and HEMOHigh tumors suggested that activation of HEMO is negatively associated with immune response signatures. Taken together, these results highlight that HEMO, as an endogenous retroviral envelope protein specifically expressed in tumors, represents a promising tumor biomarker and therapeutic target.