Efficiency of biomarker screening for enriched metastatic colorectal cancer trials: The ATTACC program experience.

Abstract

450 Background: Incorporation of multiple enrichment biomarkers into prospective clinical trials for metastatic colorectal cancer (mCRC) has been proposed or initiated by NSABP, EORTC, NCI Colon Task Force, and UK MRC. The feasibility of large scale screening efforts in mCRC has not been previously assessed. Methods: Patients (pts) with 5-FU refractory mCRC at MD Anderson Cancer Center were offered screening in the Assessment of Targeted Therapies Against Colorectal Cancer (ATTACC) program to identify eligibility for companion clinical trials with a therapy targeted to an aberration detected in the patient. Results: Between 8/2010 and 8/2013, 400 heavily pretreated mCRC pts were enrolled onto ATTACC for screening and allocation onto 10 separate Phase I or II companion trials based on banked FFPE tumor testing by IHC, gene sequencing, and CpG island methylation phenotype (CIMP) assay. Outside tissue was obtained with a median of 6 calendar days (8% had insufficient tumor). IHC, sequencing, and CIMP results required a median of 11, 12, and 20 days, respectively. In this population, 53% tumors had KRAS mutations, with BRAF, NRAS, and PI3KCA mutations, and complete PTEN loss present in 8.1%, 8.6%, 21%, and 12%, respectively. CIMP-high status (≥2 markers) was noted in 38% of pts. Given molecular results, trial-specific eligibility, and intermittent study openings, fewer than 25% of screened pts were eligible for enrollment onto a companion protocol. Of these eligible patients, approximately half (13%) enrolled onto a companion trial. Within the entire cohort, 19% enrolled in unenriched trials, and 51% received standard-of-care therapies. ATTACC screening, with an expanding biomarker panel, remains ongoing at a rate of 25-30 patients per month. Conclusions: Universal screening programs such as ATTACC represent an optimal strategy for enrolling patients in multiple biomarker enrichment studies. Even in the setting of multiple companion treatment studies, motivated patients/physicians, and a dedicated screening infrastructure, a majority of patients did not ultimately participate on a companion trial, emphasizing the lack of drugs targeting a number of common aberrations.