Successful implementation of a novel trial model: The Signature program.

Abstract

106 Background: Here we describe 8 ongoing single agent clinical protocols under Novartis’ “Signature” program involving buparlisib (BKM120, PI3Ki), dovitinib (TKI258, multikinase inhibitor), binimetinib (MEK162, MEKi), encorafenib (LGX818, RAFi), ribociclib (LEE011, CDK4/6i), BGJ398 (FGFRi), ceritinib (LDK378, ALKi) and sonidegib (LDE225; SMOi). These are tissue-agnostic, genetic alteration-specific (mutation, amplification, translocation, etc.) protocols using patients (pts) identified via standard-of-care profiling. This brings the ‘Protocol to the Patient’ for pts with actionable genetic alterations and who would like access to drugs targeting those alterations. Methods: Pts with advanced solid and hematologic cancers and no standard therapeutic options are eligible. Pts are preidentified with a local test performed in a CLIA laboratory for an actionable genetic alteration. Indications where existing data showed no benefit, or key studies were planned, were excluded from accrual. The primary objective is to assess clinical benefit (SD or better for ≥ 16 weeks) for each compound. A novel adaptive statistical design is used to cluster pts of like indications into cohorts for independent analysis for futility (minimum 10 pts) or efficacy (minimum 15 pts). Results: Between March 2013 – January 2015, 16 academic and 151 unique community/network sites have dosed 368 pts; buparlisib (142), dovitinib (73), binimetinib (90), encorafenib (9), ribociclib (30), BGJ398 (12), ceritinib (3) and sonidegib (9) with completed cohorts for buparlisib (CRC, ovarian, sarcomas, HNSCC, cervix), dovitinib (CRC, GIST), and binimetinib (lung). The average startup timeline was 5.2 weeks.The most frequent genetic alterations were RAS mutation (68%), PIK3CA mutation (55%), and PTEN loss (41%). Preliminary activity was observed in various tumors; buparlisib (vaginal, HNSCC), dovitinib (ovarian), and binimetinib (AML, ovarian, thyroid). Conclusions: This program allows rapid enrollment of molecularly profiled pts with genetic alterations linked to cognate targeted agents. Early signs of clinical activity suggest potential for detection of new indications using a pt-sparing design that could lead to subsequent confirmatory trials.