Liquid biopsies have shown that, in RAS mutant colorectal cancer, the conversion to RAS wild-type * status during the course of the disease is a frequent event, supporting the concept that the evolutionary landscape of colorectal cancer can lead to an unexpected negative selection of RAS mutant clones. The aim of the present study was to clarify whether the negative selection of RAS mutation in plasma might be drug-dependent. For this purpose, we used liquid biopsy to compare the rate of conversion from RAS mutant to RAS wild-type * in two groups of originally RAS mutant mCRC patients: the first treated with chemotherapy alone, while the second was treated with chemotherapy combined with bevacizumab. Serial liquid biopsies were performed at 3 months (T1), 6 months (T2), 9 months (T3), and 12 months (T4) after starting first line treatments. We found that the only independent variable significantly associated to RAS status conversion was the use of bevacizumab. RAS conversion was not found associated to tumor burden reduction, although bevacizumab-treated patients who converted to RAS wild-type * had a significantly longer PFS compared to patients who remained RAS mutant. The appearance of a "RAS wild-type * window", mainly in bevacizumab-treated patients, might present them as candidates for second line treatment with anti-EGFR, which was otherwise precluded.
Keywords: RAS conversion; bevacizumab; colorectal cancer; liquid biopsy.