Q-TWiST analysis of tivozanib (T) versus sorafenib (S) in patients with advanced renal cell carcinoma (RCC) in the TIVO-3 study.

Abstract

298 Background: In the randomized phase III study TIVO-3, the VEGFR-TKI tivozanib (TIVO) increased progression-free survival with better tolerability but no difference in overall survival (OS) relative to sorafenib (SORA) as third- or fourth-line therapy in patients with metastatic RCC. These results provide motivation to apply quality-adjusted time without symptoms of disease and toxicity (Q-TWiST) methods to quantify the net health benefits of TIVO, in the presence of similar survival, when compared to SORA. Methods: In application of Q-TWiST, patient-level OS was subdivided into three mutually exclusive states: time with toxicity (TOX), time without symptoms and toxicity (TWiST), and time after progression/relapse (REL). Mean Q-TWiST was calculated by applying utility coefficients of 0.5, 1.0, and 0.5 to the restricted mean (max 36 months follow-up) health states of TOX, TWiST, and REL, respectively; 95% CIs for the means and mean differences were estimated by bootstrap distributions. Relative Q-TWiST gain was defined as the mean absolute Q-TWiST difference divided by the SORA mean OS. Results: Mean TWiST was significantly longer for TIVO than for SORA (10.30 months v.5.35 months; Table). Mean REL time was significantly shorter for TIVO, with no difference in mean TOX time. Mean Q-TWiST was 15.04 and 12.78 months for TIVO and SORA, respectively, a statistically significant difference (p=0.0493). The relative gain for TIVO was 11.2%. Clinical trial information: NCT02627963 . Values in table are mean (95% CI) in months or p-value for difference in treatment group means. Conclusions: The difference in Q-TWiST in TIVO-3 was primarily driven by benefits of TIVO in TWiST, partially offset by superiority of SORA in REL time. As a third- or fourth-line treatment for RCC, TIVO significantly increased Q-TWiST relative to SORA, primarily through an increase in TWiST, which is generally considered to be the state with highest utility to patients. Consequently, Q-TWiST may be considered an alternative patient-centered measure of benefit of TIVO in these settings. [Table: see text]