Clinical validation of prospective liquid biopsy monitoring in patients with wild-type RAS metastatic colorectal cancer treated with FOLFIRI-cetuximab

Abstract

Cancer genomics and translational medicine rely on the molecular profiling of patient's tumor obtained during surgery or biopsy. Alternatively, blood is a less invasive source of tumor DNA shed, amongst other ways, as cell-free DNA (cfDNA). Highly-sensitive assays capable to detect cancer genetic events from patient's blood plasma became popularly known as liquid biopsy (LqB). Importantly, retrospective studies including small number of selected patients with metastatic colorectal cancer (mCRC) patients treated with anti-EGFR therapy have shown LqB capable to detect the acquired clonal mutations in RAS genes leading to therapy resistance. However, the usefulness of LqB in the real-life clinical monitoring of these patients still lack additional validation on controlled studies. In this context, we designed a prospective LqB clinical trial to monitor newly diagnosed KRAS wild-type (wt) mCRC patients who received a standard FOLFIRI-cetuximab regimen. We used BEAMing technique for evaluate cfDNA mutations in KRAS, NRAS, BRAF, and PIK3CA in twenty-five patients during a 2-y period. A total of 2,178 cfDNA mutation analyses were performed and we observed that: a) continued wt circulating status was correlated with a prolonged response; b) smoldering increases in mutant cfDNA were correlated with acquired resistance; while c) mutation upsurge/explosion anticipated a remarkable clinical deterioration. The current study provides evidences, obtained for the first time in an unbiased and prospective manner, that reinforces the utility of LqB for monitoring mCRC patients.

Keywords: anti-EGFR; cetuximab; cfDNA; colorectal cancer; liquid biopsy.

Figure 1. Flowchart of patient disposition

A total of twenty-five patients were included in the study, and twenty-three patients were evaluable. Response to FOLFIRI-cetuximab and clinical outcome are shown to each patient, as well as tumor and cfDNA mutation status. Patient 8 had a driver mutation in an unknown resistance-associated gene and later had a KRAS cfDNA mutation explosion, therefore was classified accordingly.

Figure 2

cfDNA mutation analyses of the twenty-five mCRC patients included in the study is shown divided in prior (upper figure) and after FOLFIRI-cetuximab treatment (below figure). Patients are separated accordingly to somatic mutation status (wt tumors shown in left panels, and BRAF/PIK3CA-mutated tumors in right panels) prior treatment. Stablished cfDNA mutation cutoff is 0,02%. Black dots are cfDNA mutation levels < 0,02%, blue dots correspond to prior treatment tumor mutations observed in the cfDNA at > 0,02% levels. Red dots are cfDNA mutation levels > 0,02% newly acquired during the FOLFIRI-cetuximab therapy.

Figure 3. Results of cfDNA from the twenty-three evaluable mCRC patients of the study are shown

Stablished cfDNA mutation cutoff is 0,02%. Black dots are cfDNA mutation levels < 0,02%, blue dots correspond to prior to treatment tumor mutations observed in the cfDNA at > 0,02% levels. Red dots are cfDNA mutation levels > 0,02% newly acquired during the FOLFIRI-cetuximab therapy. Patients 9, 12, 17 and 18 carried somatic BRAF mutation prior treatment.

Figure 4. Treatment response and cfDNA status of four patients that acquired KRAS/PIK3CA during FOLFIRI-cetuximab treatment mutations are shown

Patients 2 presented gradual increase of cfDNA PIK3CA mutation that anteceded disease relapse. Patient 3 had an intermediated increase in cfDNA KRAS mutation coinciding with tumor grow. Patients 7 and 8 suffered very rapid cfDNA mutation increase (mutation explosion) followed by fast metastasis spread and clinical deterioration that culminated in death. Clinical and genetic detailed information of each patient are described in the results section..

Figure 5. Kaplan-Meir survival curve of patients separated by somatic and cfDNA mutation status

All six patients with BRAF somatic mutation or cfDNA mutation explosions progressed fast and died. None of the seventeen patients with continued wt cfDNA status (N = 15) or with intermediate/gradual increase in cfDNA mutations (N = 2) died during the study.