448PD - Genomic profiling of three pathways through molecular profiling-based assignment of cancer therapy (NCI- MPACT)

Emerging clinical data show prediction of response to therapies targeting specific genetic aberrations have unexpectedly variable outcomes. This multicenter, double-blind, randomized trial opened in 2013 to compare response rates (RR) between 2 groups of patients (pts) identified to have an actionable mutation of interest (aMOI) in one of 3 genetic pathways (DNA repair, PI3K, or RAS/RAF/MEK): group A) Pts treated with agent(s) targeting one selected pathway (experimental arm-A) and B) Pts treated with agent(s) not targeting that pathway (control arm-B). Based on the data available at the time, the aMOI selection criteria encompassed alterations throughout the entire selected pathway instead of specific genetic changes. Primary objective is to compare the RR (CR and PR) and 4 months PFS between treatments arms A and B. A CLIA-certified genetic analysis of a fresh tumor biopsy was performed at entry. Pts with an aMOI were randomized 2:1 to arm A vs. B. Study drugs were: 1) DNA repair-a) veliparib & temozolomide (VT); b) AZD1775 & carboplatin (AC); pts with p53 mutations were preferentially selected for AC; 2) PI3K- everolimus (E); 3) RAS- trametinib(T). At disease progression, Arm B pts could cross over to their target arm (A). 193 pts underwent biopsies; >90% of samples completed DNA sequencing. 96 pts (50%) had an aMOI and were randomized to treatment. Cohort VT had insufficient accrual on the experimental arm to be evaluable. AC, E and T cohorts were closed due to futility. Enrollment rate after treatment assignment was 77% for Arm A and 53%, for arm B. Attrition analysis between arms A and B are ongoing. The increasing availability of genetic sequencing and bias toward expected benefit of highly specific treatment agents may account for the large number pt withdrawal from Arm A. This imbalance made comparison of arms A and B uninterpretable. The trial has been amended to employ a non-randomized design to complete the assessment of VT’s activity. Analysis of the aMOIs are ongoing to develop a more stringent selection criteria for future precision medicine trials. NCT01827384. Christina Rosenberger, PhD. NCI. NCI. S. Kummar: Advisory / Consultancy: Corvus Pharmaceuticals; Advisory / Consultancy: MedTree; Advisory / Consultancy: Nodus Therapeutics; Advisory / Consultancy: Genentech; Advisory / Consultancy: ShangPharma Innovation; Advisory / Consultancy: Seattle Genetics; Travel / Accommodation / Expenses: Bayer; Shareholder / Stockholder / Stock options: Dh