Abstract 4367: Discovery and evaluation of pharmacodynamic and predictive biomarkers for anti-RSPO3, a treatment that reduces tumor growth and cancer stem cell frequency in patient derived xenograft tumor models

Abstract

Abstract R-spondins (RSPOs) are a novel family of secreted ligands that bind to LGR receptors and are important in human tumor development. We screened our tumor bank of close to 100 patient-derived xenografts (PDXs) and found that a subset of breast, lung, ovarian, and pancreatic cancer models highly express RSPO3, which activates the Wnt/β-catenin signaling pathway by blocking internalization and degradation of Frizzled receptors. We generated an anti-RSPO3 antibody that significantly reduces cancer stem cell frequency, as shown by an in vivo serial transplantation experiment performed in a lung PDX model. Briefly, fifty tumor cells from four different treatment groups (control antibody, anti-RSPO3, Paclitaxel, or the combination of anti-RSPO3 and Paclitaxel) were injected into a new cohort of mice. After 59 days, all mice in the control group developed palpable tumors while the mice implanted with cells harvested from anti-RSPO3 alone (p = 0.0009, vs. Ctr) or in combination with Paclitaxel (p = 0.0007, vs. Ctr) groups developed fewer and smaller tumors. The strongest effect in decreasing tumorigenicity was observed in the combination group (p = 0.0046 vs. Paclitaxel). We also tested anti-RSPO3 treatment in a range of RSPO3 expressing PDX cancer models. Our data show that compared to standard of care the combination of our antibody with Paclitaxel is more effective in promoting tumor growth inhibition (TGI) in both ovarian (93%<TGI<98%) and lung (78%<TGI<93%) PDX models (Paclitaxel: 48%<TGI<69% and 60%<TGI<68% respectively). In addition, anti-RSPO3 treatment is effective in delaying tumor growth both as a single agent (47%<TGI<76%) and in combination with Irinotecan (66%<TGI<100%) when compared to the Irinotecan alone group (44%<TGI<74%) in colorectal cancer models with and without PTPRK-RSPO3 genomic fusion. Using Gene Set Enrichment Analysis (GSEA) we also identified pharmacodynamic biomarkers in xenograft tumors and surrogate tissues. GSEA shows down-regulation of stem cell, oncogenic, and Wnt pathway signatures. Examples of individual genes identified are: AXIN2, LGR5 and ZNRF3 from tumors, LEF1, FZD3 and SOX4 from hair follicles and SOX4, UBR5 and MYST3 from the blood. Similarly, several protein tumor markers (including CA19-9, CEA and IGFBP2), were down-regulated in plasma samples. These RSPO3-related genes and proteins from hair and blood could be used as biomarkers to monitor target engagement in the clinic. Collectively the efficacy observed in our PDX models suggests that RSPO3 expression levels correlates with sensitivity to anti-RSPO3 and could be used as a predictive biomarker. We are currently developing assays to be used in the clinic to measure RSPO3 levels, and we intend to initiate a phase 1a trial for anti-RSPO3 in 2015. Citation Format: Fiore Cattaruzza, Pete Yeung, Wan-Ching Yen, Alayne Brunner, Min Wang, YuWang Liu, Marcus Fischer, Gilbert O'Young, Cecile Chartier, Austin Gurney, Tim Hoey, John Lewicki, Ann M. Kapoun. Discovery and evaluation of pharmacodynamic and predictive biomarkers for anti-RSPO3, a treatment that reduces tumor growth and cancer stem cell frequency in patient derived xenograft tumor models. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4367. doi:10.1158/1538-7445.AM2015-4367