Post-surgery level of circulating DNA in stage III colon cancer patients: impact on the reliability of minimal residual disease detection

PurposeMinimal residual disease (MRD) assessment guided by circulating DNA (cir-nDNA) testing is strongly prognostic in operable colon cancer patients. However, further clinical research to result in fully changing the clinical management of colon cancer about the use of adjuvant therapy. Experimental DesignIn order to improve the analytical signal sensitivity of MRD detection we investigated the sources of cir-nDNA and its concentration variation in 74 stage III colon cancer patients before surgery and up to eight weeks after tumor resection. ResultsA majority of stage III colon cancer patients showed significant higher post-surgery cir-nDNA levels as compared to pre-surgery level (82.2% and 64.4% during the first month and second month period). We observed a strong association of two neutrophil extracellular traps (NETs) markers (Myeloperoxidase, MPO and neutrophil elastase, NE) during the two-months post-operative period. ConclusionsWhereas the literature previously assumed that cir-nDNA concentration significantly decreased at most one-month post-surgery in the majority of stage III colon cancer patients, our data challenge this paradigm. NETs appear to constitute a confounding factor in assessing total cir-nDNA concentration since NETs production largely varies post-surgery among patients. Thus, data suggest that it would be misleading to define an optimal post-surgery blood collection time for MRD detection. At best, we estimate that given the high level of cir-nDNA content and large inter-individual variability, the best time range for blood collection could be between the fourth and the sixth week post-surgery. Second, the value of the variant allele frequency (VAF) that is so far a criteria to select mutant cir-nDNA, should be cautiously taken into consideration. We are providing various recommendations in that regards. Translational RelevanceOur observation is of translational relevance with regard to the future practice of cancer medicine, specifically minimal residual disease (MRD) assessment guided by cir-nDNA testing, which shows considerable promise in that domain. Given high inter-individual variation, as well as the high levels of cir-nDNA concentration arising from neutrophil extracellular traps (NETs) formation, we conclude that no optimal post-surgery blood collection time for MRD detection can be determined for all patients, and the use of the mutation allele frequency (MAF) should be limited. Our data suggest that best practice would be to accurately test for cir-nDNA levels, to monitor for NETs-derived inflammation, and to use the absolute mutant DNA quantification instead of MAF in the post-operative period.