Novel statistical analysis of long-term survival to account for crossover in a phase III trial of sunitinib (SU) vs. placebo (PL) in advanced GIST after imatinib (IM) failure
2008; Lippincott Williams & Wilkins; Volume: 26; Issue: 15_suppl Linguagem: Inglês
10.1200/jco.2008.26.15_suppl.10524
ISSN1527-7755
AutoresG. D. Demetri, Xiaoyan Huang, Christopher R. Garrett, Patrick Schöffski, Martin E. Blackstein, Manisha H. Shah, Jaco J. Verweij, Vanessa Tassell, Charles M. Baum, Paolo G. Casali,
Tópico(s)Lymphoma Diagnosis and Treatment
Resumo10524 Background: SU is an oral multitargeted tyrosine kinase inhibitor of KIT, PDGFRs, VEGFRs, FLT3, CSF-1R and RET, approved multinationally for treatment of advanced GIST after IM failure. Interim analysis of a double-blind, PL-controlled phase III trial in this pt population showed a significant difference in OS between pts randomized to SU (50 mg/d administered in 6-wk cycles of 4 wk on treatment, 2 wk off; N=207) or PL (N=105) favoring SU (HR 0.49; P=0.007), without medians having been reached. The study design allowed PL pts to cross over to SU at PD or unblinding of the trial. While efficacy and safety/tolerability continue to be assessed, conventional statistical methods yield biased treatment-effect estimates for mature OS data because of the crossover. Methods: In addition to the log-rank test, Cox model and Kaplan-Meier method, OS was analyzed using the rank-preserving structural failure time (RPSFT) method to account for crossover. Results: Across the entire study (double-blind + open-label phases), 243 pts were randomized to receive SU and 118 pts to receive PL, 104 of whom crossed over to SU. As of Nov. 2007, conventional analysis showed that OS converged in the two treatment groups (SU median 74.7 wk, 95% CI 61.4–85.7; PL 64.9 wk, 45.7–98.4; HR 0.82, P=0.128) as expected, given the crossover design. However, RPSFT analysis yielded an estimated median OS for PL of 36.0 wk (95% CI 25.9–51.0), revealing a significant SU treatment effect (HR 0.46, P<0.0001) comparable to that of the blinded phase. As in the blinded phase, the most common treatment-related AEs throughout the entire study were fatigue, diarrhea, nausea and skin discoloration, mainly grade 1/2; incidences increased slightly with extended SU treatment. The frequencies of hematologic laboratory abnormalities were similar for shorter-term and extended SU treatment. The incidence of cardiac AEs was low with SU across the study (4.3%). Conclusions: Use of RPSFT analysis to account for crossover confirmed the long-term OS benefit provided by SU vs. PL to pts with IM-resistant/intolerant GIST in this phase III study, with implications for other crossover trials. SU demonstrated acceptable, predictable safety with long-term treatment. Author Disclosure Employment or Leadership Consultant or Advisory Role Stock Ownership Honoraria Research Expert Testimony Other Remuneration Pfizer Inc. GlaxoSmithKline, Infinity, MedImmune, Novartis, Pfizer Inc., sanofi-aventis Pfizer Inc. Novartis, Pfizer Inc., PharmaMar, SigmaTau Infinity, Jansen Cilag, Novartis, Pfizer Inc., SigmaTau Infinity, Novartis, Pfizer Inc. Novartis, PharmaMar
Referência(s)