Single nucleotide polymorphism associations with response and toxic effects in patients with advanced renal-cell carcinoma treated with first-line sunitinib: a multicentre, observational, prospective study

Background Sunitinib is a tyrosine kinase inhibitor with proven efficacy in renal-cell carcinoma, but some patients do not respond or need dose reductions due to toxicity. Because there are no validated molecular predictors of response or toxicity to sunitinib, we aimed to identify genetic markers predictive of outcome and toxic effects. Methods In our observational, prospective study we enrolled previously untreated adults (>= 18 years) with clear-cell renal-cell carcinoma at 15 institutions in the Spanish Oncology Genitourinary Group in Spain. Patients received sunitinib according to local practice guidelines. We assessed RECIST response, progression-free survival (PFS), overall survival, and toxicity of sunitinib with 16 key polymorphisms in nine genes: VEGFR2 (rs2305948 and rs1870377), VEGFR3 (rs307826, rs448012, and rs307821), PDGFR-a (rs35597368), VEGF-A (rs2010963, rs699947, and rs1570360), IL8 (rs1126647), CYP3A4 (rs2740574), CYP3A5 (rs776746), ABCB1 (rs1045642, rs1128503, and rs2032582), and ABCB2 (rs2231142). We assessed associations with efficacy and toxicity by use of univariable and multivariable analyses (with clinical factors associated with outcomes as covariates). We adjusted for multiplicity using the Bonferroni method; p values of less than 0.0031 before adjustment were deemed to still be significant after adjustment. Findings We enrolled 101 patients between Oct 10, 2007, and Dec 13, 2010. 95 of these patients were included in toxicity analyses and 89 in the efficacy analyses. Two VEGFR3 missense polymorphisms were associated with reduced PFS with sunitinib on multivariable analysis: rs307826 (hazard ratio [HR] per allele 3.57, 1.75-7.30; p(unadjusted) = 0.00049, p(adjusted) = 0.0079) and rs307821 (3.31, 1.64-6.68; p(unadjusted) = 0.00085, p(adjusted) = 0.014). The CYP3A5*1 (rs776746) high metabolising allele was associated in a multivariable analysis with an increased risk of dose reductions due to toxicity (HR per allele 3.75, 1.67-8.41; p(unadjusted) = 0.0014, p(adjusted) = 0.022). No other SNPs were associated with sunitinib response or toxicity. Interpretation Polymorphisms in VEGFR3 and CYP3A5*1 might be able to define a subset of patients with renal-cell carcinoma with decreased sunitinib response and tolerability. If confirmed, these results should promote interventional studies testing alternative therapeutic approaches for patients with such variants.