PCN9 VALIDITY OF PROPORTIONAL HAZARDS (PH) WEIBULL MODELS FOR ANALYZING PROGRESSION FREE SURVIVAL (PFS) AND OVERALL SURVIVAL (OS) IN PATIENTS WITH TRASTUZUMAB (TZ)-REFRACTORY ERBB2+ (HER2+) METASTATIC BREAST CANCER (MBC) RECEIVING LAPATINIB PLUS CAPECITABINE (L+C) VERSUS CAPECITABINE ONLY (C-ONLY)
| Autor: | M Walker, Paul Tappenden, Thomas E. Delea, Jonathan Karnon, Oleg Sofrygin, Mayur M. Amonkar |
|---|---|
| Rok vydání: | 2008 |
| Předmět: |
Oncology
medicine.medical_specialty business.industry Health Policy Hazard ratio Public Health Environmental and Occupational Health Accelerated failure time model Lapatinib medicine.disease Metastatic breast cancer Capecitabine Trastuzumab Internal medicine medicine Progression-free survival skin and connective tissue diseases business neoplasms Survival analysis medicine.drug |
| Zdroj: | Value in Health. 11(3) |
| ISSN: | 1098-3015 |
| DOI: | 10.1016/s1098-3015(10)70189-3 |
| Popis: | VALIDITY OF PROPORTIONAL HAZARDS (PH)WEIBULL MODELS FOR ANALYZING PROGRESSION FREE SURVIVAL (PFS) AND OVERALL SURVIVAL (OS) IN PATIENTSWITH TRASTUZUMAB (TZ)-REFRACTORY ERBB2+ (HER2+) METASTATIC BREAST CANCER (MBC) RECEIVING LAPATINIB PLUS CAPECITABINE (L+C)VERSUS CAPECITABINE ONLY (C-ONLY) Sofrygin O, Delea T,Tappenden P, Karnon J,Amonkar MM, Walker M Policy Analysis Inc. (PAI), Brookline, MA, USA, University of Sheffield, Sheffield, UK, GlaxoSmithKline, Collegeville, PA, USA, GlaxoSmithKline, London, UK OBJECTIVES: Lapatinib is an oral small molecule dual targeted therapy that binds intracellularly to the ATP binding site of the EGFR and ErbB2 (HER2) receptors. In the EGF100151 trial, L+C improved time to progression (TTP) and progression free survival (PFS) vs C-only in women with ErbB2+ MBC who had received prior therapy including TZ. Following achievement of the primary endpoint, enrollment was halted, preventing demonstration of a significant difference in OS. METHODS: To inform ongoing analyses of the cost-effectiveness of L+C vs. C-only, Weibull survival functions for PFS and OS were fitted to observed failure time data from EGF100151 using Accelerated Failure Time (AFT) regression. Survival function parameters were estimated using a single regression equation for each outcome with treatment groups entered as an independent variable. Hazard Ratios (HRs) for progression and death with L+C were assumed to be proportionate to HRs for C-only. Expected PFS, OS, and post-progression survival (PPS) were calculated for each group. The validity of the Weibull model and PH assumption were assessed using graphical and analytical methods. RESULTS: Expected PFS, PPS and OS for L+C were 36.89, 43.78, and 80.67 weeks, respectively. Corresponding values for C-only were 22.49, 45.03, 67.47 weeks, respectively. Graphical tests of transformed survival functions supported use of Weibull distribution. Correlation tests between the ranked failure times and Schoenfeld residuals, supremum test for proportional hazards assumption, and comparisons of HRs for L+C vs. C-only by quarter postrandomization, provided no strong evidence of nonproportionality. CONCLUSION: Proportional hazards Weibull survival models are valid for modeling survival time data in patients with trastuzumab-refractory ErbB2-overexpressing MBC receiving L+C versus C-only, and suggest that lapatinib provides substantial benefit in terms of PFS and OS in patients with ErbB2+ MBC. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect