Detecting the disappearance of RAS-mutant clones in the plasma of patients with RAS-mutant mCRC to select patient candidates for anti-EGFR treatment and to monitor resistance to treatment

Abstract

Introduction: The pattern of clonal evolution of RAS mutations in cancer is highly heterogeneous, with some tumors retaining and some others losing RAS mutations as the disease progresses. This mixed pattern of positive and negative selection of RAS mutations observed at relapse appears to be related to the selective pressure of treatments. To date, while the mutational landscape of RAS wild-type metastatic colorectal cancer (mCRC) characterized by the appearance of RAS mutation at disease progression has been widely investigated through liquid biopsy, that of mutant RAS cancers losing RAS mutations at relapse has rarely been exploited. Given that metastatic sites are often difficult to re-biopsy in clinical practice, genomic profiling from circulating tumor DNA offers potential advantages when compared with tumor tissue-based sequencing, being easily and repeatedly collected on almost any patient. To date, mCRC harboring RAS mutations has been judged as ineligible for anti-EGFR targeted therapies and are doomed to rely on the inhibition of angiogenesis as the only actionable molecular target. Recommendations for medical oncologists take no notice of the molecular background of a priori resistant, RAS-mutant mCRC and ignore potential strategies to inhibit the addition of cancer clones to the EGFR pathway. We aimed to explore relapse-associated mutated key genes in mCRC characterized by negative selection of RAS mutations in plasma at the failure of anti-angiogenesis based regimens and thus treated with EGFR inhibitors. Although preliminary, this study might identify additional oncogenic driver mutations with the potential to be exploited for establishing future therapeutic strategies. Methods: We prospectively enrolled 13 RAS-mutant mCRC patients at the time of progression of disease from bevacizumab containing treatments (any line of therapy). The Idylla™ system was used to select patient candidates for anti-EGFR treatments, on the basis of the disappearance of RAS-mutant clones in plasma. With the same method, we re-evaluated RAS genes mutational status at the progression of disease after anti-EGFR treatment. Results: RAS genes were found to be wild-type in 9/13 (69%) patients. Two of these patients died before the beginning of treatment. Seven patients who converted to RAS wild type status in plasma at the time of progressive disease received EGFR inhibitors, achieving clinical benefit. Plasma DNA samples were available from 4/7 patients for Idylla™ analysis at both baseline (before starting anti-EGFR treatment) and at progression of disease. At failure of EGFR inhibitors, RAS genes wild type status was maintained in 2 out of 4 patients, a KRAS G12C mutation was maintained in one patient and a new RAS mutation (NRAS A146 T) was detected in the last one. Conclusion: We propose a hypothetical algorithm which accounted for the transient disappearance of RAS-mutant clones over time, which might extend the continuum of care of mutant RAS colorectal cancer patients through the delivery of a further line of therapy. We further investigated relapse-associated mutated key genes in mCRC treated with EGFR inhibitors. The persistent absence of RAS mutations at relapse might lead to the identification of additional oncogenic driver mutations as potential targets for future therapeutic strategies.