Phase II study of single-agent cetuximab in KRAS g13d mutant metastatic colorectal cancer

RAS mutant metastatic colorectal cancer (mCRC) patients are excluded from treatment with anti-epidermal growth factor receptors. Nevertheless, preclinical experiences and retrospective data from large phase III trials suggested that patients carrying KRAS G13D mutation might derive benefit from cetuximab in first and advanced lines of treatment [1–3]. In particular, cetuximab has anti-proliferative activity in KRAS G13D mutant cell lines and in wild-type ones. Consistently, the post hoc analysis of a pooled dataset of 579 chemo-refractory patients treated with cetuximab between 2001 and 2008 showed that KRAS G13D mutant patients (n = 32) achieved longer progression free survival (PFS) [4.0 versus 1.9 months, hazard ratio (HR) = 0.51, P = 0.004] and overall survival (OS) (7.6 versus 5.7 months, HR = 0.50, P = 0.005) when compared with patients bearing other KRAS mutations. A significant interaction between KRAS mutational status (KRAS G13D versus other KRAS mutations) and OS benefit from cetuximab was also detected (P for interaction = 0.003). Moving from such strong rationale, we conducted this hypothesisconfirmatory phase II single-arm trial to provide a prospective proof of the clinical benefit of cetuximab in KRAS G13D mutant mCRC patients and to definitively ascertain if such drug might deserve further consideration in this subgroup of patients. The primary end point was 4 months-progression-free rate (PFR), defined as the proportion of patients alive and free from disease progression 4 months after treatment start. Responses were determined according to RECIST 1.1. On the basis of a systematic review of literature reporting a 4-month PFR of 10% in patients with KRAS mutations other than G13D, we assumed that cetuximab would have been a promising option if 50% of patients with G13D KRAS mutation would have been free from progression 4 months after treatment start. According to the phase II Fleming single-stage design, setting α and β errors as 0.05 and 0.10, respectively, 12 patients were needed. Cetuximab would have been considered worth of further investigation if at least four patients had been alive and progression free at 4 months. Patients carrying G13D KRAS mutation, progressed after treatment with fluoropyrimidine, oxaliplatin, irinotecan and bevacizumab or with no other valid therapeutic options, with histologically confirmed adenocarcinoma, measurable metastatic disease and life expectancy >3 months, were considered eligible and were prospectively enrolled to receive cetuximab 500 mg/m2 biweekly. Twelve consecutive patients were enrolled from January 2011 to April 2013. The median number of administered cycles was 5 (range 2–12). The primary objective of the trial was not met: three patients (25%) were progression-free at 4 months. None of 12 patients achieved response and three patients (25%) had disease stabilisation, for an overall disease control rate (DCR) of 25%. Notably, DCR at 6 months was 0%. Median PFS and OS were 1.9 (95% CI 1.7–3.8) and 7.2 months (95% CI 5.7–9.7), respectively. No unexpected toxicities were observed (grade 3 or 4 skin rash 17%). To our knowledge, this is the first prospective trial addressing the issue of the efficacy of cetuximab in patients affected by KRAS G13D-mutated mCRC. Among 12 treated patients, no responses were observed; thus, cetuximab does not provide a clinically relevant benefit to these patients. We should acknowledge that the phase II ICE CREAM trial addressing the same question is currently ongoing (https://www.anzctr.org.au/Trial/ Registration/TrialReview.aspx?id=362339). The results will help to draw a definitive conclusion. However, such findings underline that data obtained from subgroup and retrospective analyses are only hypothesis-generating and always need to be interpreted with caution and prospectively validated.