Steroid response as prognostic factor and its correlation with molecular assessment of childhood acute lymphoblastic leukemia

Background: Survival rate of children with acute lymphoblastic leukemia (ALL) in Indonesia remains low. Risk stratification accuracy is important to improve survival. In developed countries, risk stratification is determined based on gene fusion that is known related to steroid resistency. Steroid response at day-8 correlates with prognosis. The assessment can be applied in centers that cannot perform molecular assessment. This study aims to evaluate whether steroid response correlated to molecular assessment. Methods: A cross-sectional study was performed at Child Health Department, Cipto Mangunkusumo Hospital (January 2013-March 2014), a total of 73 patients were enrolled. Steroid was given for 7 days. Peripheral blast count at day 8 was evaluated, good response if blast count < 1000 /μL and poor if ≥1000 /μL. Fusion gene detection was also performed. The data was analysed using Statistical Package for Social Sciences (SPSS) version 20.0. Results: Fusion gene was detected in 45 patients. In 1-10 years age group, 26/32 (81%) subjects had good response, while 75% in < 1 year age group and 7/9 (78%) in =10 years age group had poor response. 5/7 (71%) subjetcs had leukocyte count > 100,000 /μL and 7/8 (88%) with T-cell showed poor response. Age, leukocyte count, and T-cell were statistically correlated with steroid response (p < 0.05). E2A-PBX1 fusion gene was the most common 19/45 (42%), followed by TEL-AML1 17/45 (38%), BCR-ABL 5/45 (17%), and MLL-AF4 1/45 (3%). Four of five subjects (80%) with BCR-ABL and one subject with MLL-AF4 had poor steroid response. On the other hand, 12/19 (63%) with E2A-PBX1 and 13/17 (77%) with TEL-AML1 had good response. There was no correlation between steroid response and molecular assessment. Conclusion: Steroid response correlates with age, leukocyte count, and T-cell but not with molecular assessment.