Genomic landscape and survival analysis of ctDNA “neo-RAS wild-type” patients with originally RAS mutant metastatic colorectal cancer

Background: The term “neo-RAS wild-type” refers to the switch to RAS wild-type disease in plasma circulating tumor DNA (ctDNA) from originally RAS mutant colorectal cancers. Consistently, the hypothesis to re-determine RAS mutational status in ctDNA at disease progression in RAS mutant mCRC opened to a new perspective for clinically-based selection of patients to be treated with EGFR inhibitors. Currently, the genomic landscape of “neo-RAS wild-type” is unknown. This is a prospective study aimed to investigate clinical and genomic features associated with RAS mutation clearance in a large cohort of RAS mutant mCRC patients who converted to RAS wild- type in liquid biopsy at failure of first-line treatments. Secondary aim was to investigate the long term prognostic significance of “true neo-RAS wild- type”. Patients and methods: 70 patients with stage IV RAS mutant colorectal cancer were prospectively enrolled. Plasma samples were collected at progression from first-line treatment. RAS/BRAF mutations in plasma were assessed by RT-PCR. In RAS/BRAF wild-type samples, ctDNA was used to generate libraries using a 17 genes panel whose alteration has clinical relevance. To investigate the prognostic significance of RAS mutation clearance, test curves for PFS and OS were represented by Kaplan-Meier estimator plot and Log-rank test. Results: The most commonly detected actionable mutations in “neo-RAS wild-type” were: PIK3CA (35.7%); RET (11.9%); IDH1 (9.5%); KIT (7%); EGFR (7%); MET (4.7%); ERBB2 (4.7%); FGFR3 (4.7%). Both OS and post-progression survival were longer in patients with “neo-RAS wild-type” compared to those who remained RAS mutant (p<0.001 for both). Conclusions: De-novo-targetable mutations occured in a large percentage of “neo-RAS wild-type”, being PIK3CA the most commonly detected. RAS mutation clearance in ctDNA is associated with long- term improvement of overall survival.