Development of a second-generation TRAIL agonist and predictive biomarker profile for colorectal cancer

In colorectal cancer (CRC), RAS mutated cancers comprise 30-50%, for which there are limited treatment options. Tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) has been found to selectively induce cell death in 50% of the CRC cell lines tested in vitro and this appears to be independent of RAS mutation status. Minimal clinical success was observed for first generation TRAIL receptor agonists primarily due to poor agonist activity and poor bioavailability. In addition, a biomarker to identify responsive patients was lacking. Here we present the therapeutic development of MM-201, a next generation TRAIL receptor agonist, and its companion biomarker strategy. MM-201 is composed of an IgG1 Fc fused to a single chain TRAIL trimer (Fc-scTRAIL) engineered for improved stability and activity. Using TRAIL surface display on yeast, we identified three mutations (R130G/N228S/I247V) from a random mutagenesis library that lead to enhanced surface expression and DR5 binding. Incorporation of these three mutations into each TRAIL protomer of MM-201 resulted in an improvement of greater than 10°C in TM and an extended serum half-life of 33 hours in mice compared to wild-type Fc-scTRAIL and comparable activity in a panel of colorectal cancer (CRC) cell lines. To uncover predictive biomarkers for MM-201, we used continuous logic gate analysis on a publicly available dataset of cell survival for 27 CRC cell lines treated with cross-linked TRAIL cytokine. Using this dataset and gene expression data from Cancer Cell Line Encyclopedia, we explored combinations of 33 differentially expressed genes in the apoptosis pathway and identified a two gene biomarker signature that positively correlated with TRAIL response in CRC cell lines. The combined signature of high DR4 and low cIAP-1 was identified with a correlation coefficient of 0.99 with MM-201 response across 15 CRC cell lines both in vitro and in vivo. Subsequently we evaluated MM-201 and our biomarker approach in vivo, in a group of five CRC patient-derived xenograft (PDX) models. Four of the five models were sensitive to MM-201, with an average tumor growth inhibition (TGI) of 91%. The relative expression of DR4 and cIAP-1 in each of the PDX models was measured by RT-qPCR and the biomarker scores were observed to be significantly higher for all responding models. Given the responses observed in preclinical models, we used TCGA data to assess biomarker prevalence in patient tumor tissues. This analysis predicted that 59% of samples would have a biomarker score greater than 0.5, which correlates to a 75% reduction in cell viability in vitro. In contrast, only 12% of normal tissues had a score in this range. Our biomarker signature identifies a potentially responsive patient population and when combined with our next generation TRAIL agonist, MM-201, has the potential for development in CRC, especially for RAS mutated cancers. Citation Format: Sara Ghassemifar, Yasmin Hashambhoy-Ramsay, Haluk Yuzugullu, Tamara Utermark, Violette Paragas, Tim Maiwald, Lia Luus, Maja Razlog, Hannah Hudson, Diane Chai, Birgit Schoeberl, C. Patrick Reynolds, Peter Sorger, Andrew J. Sawyer, Daryl C. Drummond, Eric Tam. Development of a second-generation TRAIL agonist and predictive biomarker profile for colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 700.