Background and Objective: Acute lymphoblastic leukemia (ALL) is the most common malignancy in children, accounting for about a quarter of all malignancies. This study aimed to identify the prognosis of patients with relapsed acute lymphoblastic leukemia in Dr. Sheikh Hospital in Mashhad from 2006 to 2015 using competing risk models.

Materials and Methods: Patients with ALL from March 2006 to February 2015 were reviewed retrospectively, yielding 424 children with acute lymphoblastic leukemia. We considered the recurrence of cancer as a main event and death without recurrence as a competing risk. We employed Cox models and sub-distributions by univariate and multivariate analysis. Data analysis was performed using Stata and R softwares.

Results: The event free 5-year survival and the overall 5-year survival rates of all 424 patients were 70.2% and 81.9% respectively. The 5-year cumulative incident probability of recurrence and death prior to recurrence were 12 and 17 percent respectively. Based on the subdistribution models, white blood cells and blood platelets have been known to affect the recurrence of cancer.

Conclusion: Our results showed that Cox and competing risk models have different results to determine the prognosis factors affecting leukemia patients’ survival despite the competing risks. Also subdistribution models reduce the duration of response to therapy and thus reduce the rate of mortality by providing significant health policies.