P-171 Prognostic value of conversion from RAS-mutated to RAS wild-type during treatment of metastatic colorectal cancer using liquid biopsies – real-world data of two Portuguese institutions

Abstract

Determination of RAS and BRAF mutational status is standard-of-care in the management of metastatic colorectal cancer (mCRC). Liquid biopsy (LB) is a powerful technology to evaluate tumor genetic heterogeneity in a noninvasive manner. Because RAS-mutations (mut) play a major role in resistance to anti-EGFR-therapy in mCRC, serial monitoring of RAS-mut with LB may be useful to guide treatment. A retrospective observational multicenter study was conducted in patients (pts) with mCRC between January/2018-December/2021. RAS and BRAF mutations were examined in the tissue biopsy at the time of metastatic diagnosis (TMD), and by LB throughout treatment (according to physician’s choice of timing), using the Idylla™ test (Biocartis). Our aim was to evaluate the prognostic impact of loss of RAS-mut status (conversion to RAS wild-type (wt) in LB) during treatment of mCRC, and to identify pts and tumor characteristics that correlate with RAS-conversion. A total of 58 LB were done in mCRC patients. Forty-one pts initially with RAS-mut tumors were studied, median (MED) age was 65 years (41-86), 26 (63%) were men. The majority had stage IV disease at diagnosis (N=25, 61%), MED 1 site of metastases (1-4), mostly hepatic (N=26, 63%) and pulmonary (N=13, 32%). Primary tumor of right-colon occurred in 18 pts (44%). Forty-pts (97,5%) underwent surgery of primary tumor, 13 (32%) complemented with metastasectomy. Curative-intent strategies were performed in 30 pts (73,2%), all receiving perioperative systemic therapy. LB was performed after a MED of 3 lines (0-7) of systemic treatment and all pts were previously treated with anti-VEGF antibodies (Abs). Conversion to RAS-wt occurred in 13 pts (32%); 8 of them (61,5%) received further treatment with anti-EGFR-Abs and 3 obtained a response (2 ongoing after 2 and 41 months). None had BRAF mutations in LB. MED follow-up after LB is 12 months, 1-year-OS is 58% (50-66) and MED OS is 15 months (11-19). A non-statistically significant numerical difference in MED OS was found between RAS-wt conversion and persistent RAS-mut pts (23 (4-40) vs 11 (5-17) months; log rank 3,8; p=0,05). Multivariate Cox regression analysis identified RAS-wt conversion as an independent prognostic factor for OS (HR 0,27; CI95 0,09-0,80; p=0,02), as well as presence of peritoneal metastases (HR 2,7; CI95 1,0-7,0; p=0,04). Moreover, more than one site of metastases at TMD was associated with RAS-conversion. This study confirms RAS-wt conversion as an independent prognostic factor for OS, with a potencial 12 months improvement in OS and a 73% relative-risk reduction of death in heavily pretreated pts with RAS-wt conversion. Although it was a retrospective study, with a small sample, these results are promising, with a rate of RAS conversion to wt superior to published data (32% versus 20% ). Several clinicopathological factors were correlated with RAS-wt conversion, which may help select pts most likely to benefit from LB during treatment. Doing LB in pts with Ras-mut tumors should be considered, preferentially early in the disease-course, since loss of mutation represents an opportunity for treatment with anti-EGFR-therapy otherwise unlikely.