[Evaluation of early response to treatment and its prognostic value in childhood acute lymphoblastic leukemia]

This study was purposed to investigate the prognostic value of early response to treatment in childhood acute lymphoblastic leukemia (ALL). Four indexes were used to assess early response to treatment including response to prednisone on day 8 (D8-PR), percentage of lymphoblast in bone marrow on day 22 (D22-BM) and day 33 (D33-BM), the level of minimal residual disease (MRD) on day 33 (D33-MRD) by morphological and molecular biological method in 426 children with ALL. Prognostic impact of early response to treatment was analyzed, and multivariate analysis of the predictive value was performed by Cox-regression analysis. All patients were followed up until October 31, 2013, with a median follow-up time of 80 months (0.5 to 106 months). The results showed that there were significant differences between event free survivals (EFS) of the sub-groups divided according to the four indexes. The 8 years-EFS in patients with prednisone good response (PGR) was significantly higher than that in patients with prednisone poor response (PPR);patients with M1 in bone marrow on day 22 or day 33 had the better outcomes than that of patients with M2/M3;patients with high level of MRD ( ≥ 10(-4)) had the worse outcomes as compared with patients with low level of MRD (<10(-4)) (P < 0.001). Cox proportional hazard model analysis showed that BCR/ABL fusion gene positive, D8-PR, D33-BM and D33-MRD were the independent prognostic factors for childhood ALL, and the hazard ratio of D33-MRD ≥ 10(-2) was highest (HR:11.886, P < 0.001). It is concluded that early response to treatment is an independent prognostic factor with important prognostic values, and it has important clinical guiding instructive significance for risk stratification in the treatment of children ALL.