Cabozantinib (cabo), an inhibitor of tyrosine kinases including MET, VEGFR2, RET, KIT, and AXL, is undergoing clinical investigation in patients with castration-resistant prostate cancer (CRPC). A phase II randomized discontinuation trial of cabo showed clinical activity in CRPC patients. However, not all patients responded to cabo, and most of the responding patients eventually experienced disease progression. The objective of our study was to evaluate the efficacy of cabo in a preclinical setting using xenograft models of CRPC with differential phenotypes to obtain insight into which tumor types might be sensitive to cabo, and to investigate mechanisms of action of this agent.

We used LuCaP 35CR, a PTEN- ERG+ Rb+ CRPC patient-derived xenograft (PDX) that exhibits amplification of the androgen receptor (AR); LuCaP 96CR, a PTEN+ ERG- Rb+ CRPC PDX with high levels of intra-tumoral androgens; LuCaP 86.2, a PTEN- ERG+ Rb- CRPC PDX that expresses the AR v5-7 variant and is not responsive to endocrine therapy; and LuCaP 93, a PTEN- ERG- Rb- neuroendocrine CRPC PDX. Expression of cabo targets in these models was determined by qPCR. Animals were treated with 30 mg/kg cabo p.o. for six weeks (5d on/2d off), and tumor volume (TV), serum PSA and body weights were monitored. IHC and expression arrays were used to analyze effects of cabo. Cabo was effective in halting tumor growth in all four PDXs. Significant inhibition of TV was evident after one week of cabo treatment compared to control groups. PSA values were also significantly lower after one week of the treatment in all three adenocarcinomas. Furthermore, cabo treatment resulted in significant survival benefits in LuCaP 35CR, LuCaP 96CR and LuCaP 93 without significant decreases in body weight. IHC analysis showed decreased microvessel density in all cabo-treated groups vs control groups, suggesting effects on tumor environment. qPCR analysis showed that all models expressed some cabo targets though at different levels, but we did not observe any associations between magnitude of cabo effects and levels of the its targets in the tumors; inhibition of LuCaP 93, which expresses the highest levels of cabo targets, was similar to that of LuCaP 96CR, which expresses low levels of cabo targets. Our analyses of expression profiles of tumors following cabo treatment showed very little overlap between the models. However, E2F1 signaling, was inhibited by cabo in Rb- PDXs LuCaP 93 and LuCaP 86.2, but not in the Rb+ PDXs LuCaP 35CR and LuCaP 96CR. Also, expression of genes regulated by androgen was increased in the Rb- tumors and decreased (LuCaP 35CR) or was not altered (LuCaP 96CR) in the Rb+ tumors, and the levels of nuclear AR showed a similar pattern. Further investigations are ongoing to obtain more detailed insight into mechanisms of cabo effects.

Citation Format: Lisha G. Brown, Holly M. Nguyen, Ilsa M. Coleman, Peter S. Nelson, Robert L. Vessella, Dana T. Aftab, Eva Corey. Assessment of cabozantinib activity in diverse prostate cancer xenograft models. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 737. doi:10.1158/1538-7445.AM2014-737