Abstract 446P
Background
HER2 positivity is found in 3-5% of mCRC. The phase II HERACLES-A trial showed that dual anti-HER2 therapy with T + L has 30% response rate (RR) in HER2+ RAS wt mCRC after failure of standard care (Sartore-Bianchi 2016). These data led to inclusion of T+ L among recommended treatments by NCCN and other guidelines, but still lack confirmation by large trials comparing treatments. The latter, especially for this uncommon subset of mCRC, would require efforts unbearable in the setting of independent clinical research. We designed this study to confirm HERACLES-A data of efficacy through a Bayesian approach allowing to monitor longitudinally efficacy of T + L in the practice setting.
Methods
We adopted a Bayesian design for an observational cohort study in order to report the RR of T + L in HER2+ mCRC by updating the prior probability of response observed in the HERACLES-A trial with the likelihood of the RR in all consecutive patients with the same characteristics treated at Niguarda Cancer Center after the HERACLES-A closure. We simultaneously monitored efficacy and toxicity using the Bayesian optimal phase II (BOP2) design (Zhou, Lee, e Yuan 2017), with futility boundaries for efficacy and safety and a total sample size of 40 patients. We planned 3 interim analyses at 10, 20 and 30 patients evaluable for RR. Type I = II error was ∼10%, calculated after 10000 simulations under H0 and H1 scenario.
Results
From May 2019 to Jan 2021, we collected data of HER2+ mCRC patients treated with T + L according to HERACLES-A inclusion criteria (Sartore-Bianchi et al. 2016). Patients were followed for RR and adverse events (AE). On May 1st 2021, at the first interim analysis, 2/10 evaluable patients had PR according to RECIST criteria, with an updated likelihood of response of 26.2% and a 95% credible interval of 13.7-39.2%. No AE G3 drug-related were reported.
Conclusions
At the first planned interim analysis, treatment with T + L for HER2+ mCRC patients was confirmed to be safe and effective. A Bayesian approach, allowing to monitor results accounting for previously available data, can support the process of approval by regulatory authorities for treatments targeted to uncommon subsets such as HER2+ mCRC.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Fondazione Oncologia Niguarda.
Disclosure
A. Amatu: Financial Interests, Personal, Advisory Board: Roche; Bayer; Financial Interests, Personal, Invited Speaker: CheckmAb. S. Siena: Financial Interests, Personal, Advisory Board: Daiichi Sankyo; Amgen; Bayer; Bristol Myers Squibb; CheckmAb; Merck; Roche-Genentech; Seattle Genetics. A. Sartore-Bianchi: Financial Interests, Personal, Advisory Board: Amgen; Bayer; MSD; Sanofi; Servier. All other authors have declared no conflicts of interest.