Abstract 1894: Antitumor effects of everolimus, sorafenib, and axitinib in human renal cells carcinoma xenografts resistant to first line treatment by sunitinib

Abstract

Abstract Background. Acquired resistance to sunitinib stands as a major clinical challenge in renal cell carcinoma. This study aimed to evaluate the antitumor effects of everolimus, sorafenib, and axitinib as second line treatment in renal cancer xenografts acquiring resistance to sunitinib. Materials and Methods: CAKI1 renal carcinoma cells (VHL+) were engrafted in nude mice and treated continuously with tumor-static doses of oral sunitinib (60 mg/kg) until progression (define as 100% increase tumor volume on 3 subsequent measurements). Mice that progressed were randomized to receive second line treatment with oral everolimus (2.5, 5, 10 and 20 mg/kg), sorafenib (60mg/kg), or axitinib (60mg/kg) for 3 weeks. Tumor responses and progression-free survival were calculated using Kaplan-Meier analysis. Tumors were analyzed by immunohistochemistry (IHC), Western blot, and q-RT-PCR. Results. First line sunitinib induced significant tumor growth delay compared with placebo (medians PFS: 50 and 17 days, respectively, p<0.0001) without toxicity. Tumors that progressed under sunitinib treatment showed higher mRNA levels of HIF-1α, VEGFA, CXCR4, and CA-IX as compared to tumors displaying response to sunitinib. An increase in CXCR4, vimentin, and HER3 protein expression was associated with sunitinib-resistance. No change in mouse CD31 mRNA was observed contrasting with an increased human CD31 mRNA expression in tumors developing resistance to sunitinib compared to sensitive tumors, suggesting that human cancer stem cells may dedifferentiate to gain angiogenic capacities at progression. Second line treatment with everolimus, sorafenib, and axitinib were well tolerated as defined by mouse weight, mortality, and behavior. Median PFS for 20 mg/kg everolimus was significantly higher than that of control group (13 and 5 days, respectively, p<0.001). Growth inhibition using 5 mg/kg/day everolimus was similar to that of axitinib. Doses >10mg/kg everolimus display higher effects on tumor growth than sorafenib and axitinib. Phospho-S6 expression was reduced in tumors treated with everolimus and axitinib but not with sorafenib, indicating that everolimus and axitinib inactivate the PI3k/AKT/mTOR signaling. IHC analysis of everolimus- and sorafenib-treated tumors showed no apoptosis, large hypoxic (by CA-IX staining) and necrotic areas associated, and a decreased number of tumor vessels (CD31 staining) compared to controls. Conversely, axitinib neither induce hypoxia nor necrosis in this model. Conclusions. This study shows that, using an appropriate tumor progression methodology, human tumor models can be used to evaluate first and second line treatment with targeted drugs. This setting also shows that xenograft models may allow testing molecular markers of response and resistance prior to clinical trials. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 1894. doi:1538-7445.AM2012-1894