Use of STAT3 polymorphisms to predict overall survival in patients treated with interferon-alpha for metastatic renal cell carcinoma.

e15048 Background: We previously reported that the single nucleotide polymorphisms (SNPs) in signal transducer and activator 3 (STAT3) gene were most significantly associated with better response to interferon (IFN)-α in patients with metastatic renal cell carcinoma (mRCC) in our retrospective analysis (JCO 25:2785, 2007). We conducted this trial to prospectively confirm the results, and reported the significant association between the SNPs of STAT3-2 and clinical benefits (CR, PR, and SD more than 24 weeks) of IFN-α (p = 0.039) (ASCO 2011, abstract No. 4590). In this study, we have further analyzed the correlation between overall survival (OS), progression-free survival (PFS) and the 11 SNPs reported previously. Methods: In this multicenter, prospective study, patients with histologically confirmed RCC that was metastatic, measurable disease, age > 20 years, ECOG PS 0-1 and adequate organ function received 3 dosages of 5 million U per week of IFN-α treatment. We evaluated the correlation between OS, PFS and SNP allele frequencies of STAT3-2, STAT3-0, SOCS3-1, IL4R-34, PTGS1-3, PTGS1-4, PTGS1-5, PTGS2-12, IRF2-67, ICSBP-38, and TAP2-5 (JCO 25:2785, 2007) in mRCC patients treated with IFN-α. Association between OS, PFS and genetic polymorphisms was analyzed using Logrank test for trend. Results: Two hundreds three eligible patients were enrolled between December 2006 and October 2009. All patients had prior nephrectomy, and 88.7% had ECOG PS 0. Ninety-four percent of patients had clear cell RCC, and 5% had papillary RCC. At the time of this analysis the central review assessed response rate was 13.8% (28/203) (9 CR, 19 PR). CR rate of 4.4% (9/203) was more than we expected. PFS was not associated with any of the 11 SNPs examined. However, 2 SNPs of STAT3 gene, STAT3-2 and STAT3-0, were significantly associated with OS, and the correlation of STAT3-2 (p = 0.0347) was stronger than that of STAT3-0. Namely, C/C genotype of STAT3-2 was significantly associated with prolonged OS compared with the other genotypes (C/T and T/T). Conclusions: This is the first prospective study demonstrating that STAT3 polymorphisms can predict OS in mRCC patients treated with IFN-α.