Application of Weibull Accelerated Failure Time Model on the Disease-free Survival

Background and Purpose: The goal of this study is application of the proportional hazards model (PH) and accelerated failure time model (AFT), with consideration Weibull distribution, to determine the level of effectiveness of the factors affecting on the level of disease-free survival (DFS) of the patients with breast cancer. Materials and Methods: Based on the retrospective descriptive studies, 377 female patients with primary breast cancer that had been treated at oncology section (Omid Hospital) in Mashhad, Iran, were considered for participation in the study. Results: The median was 2 years and 10 months (mean 2 years and 4 months, interquartile range 3 years and 7 months, range 1-2920 days or 8 years). The average age of patients was 49.2 ± 11.5 (range 25-80 years). The rate of DFS determined with 52.5%. Base on the result of PHs Weibull model, the risk of the metastasis after surgery in the patients with invasive tumors to the skin and chest and positive lymph node was more than 2 times [hazard ratio (HR) = 2.53; confidence interval (CI): 1.26, 5.06] and (HR = 2.37 CI: 1.42, 3.98), respectively. Hormone therapy after surgery decreases the risk of the metastasis to 0.63% (HR = 0.63; CI: 0.41, 0.95). In the AFT model, characteristic tumor decelerated (? < 1) the DFS time by a factor of 0.30 and 0.33, respectively. Furthermore, the hormonal therapy accelerated (? > 1) the DFS time by factor of 1.8. Conclusion: In this study, the PH and AFT assumption have been satisfied by data and characteristics tumor and the hormonal therapy decrease and increases time until recurrence, respectively.