OBJECTIVETo ovaluate the prognostic value of prednisone response in treatment regimes of children with acute lymphoblastic leukemia.

METHODSA total of 598 newly diagnosed ALL patients were enrolled and received prednisone pre-treatment. Based on the peripheral lymphoblast count on day 8, these patients were divided into 2 groups: prednisone good response (PGR) and prednisone poor response (PPR). PPR patients were classified into high risk group immediately and then received intensed chemotherapy. The all enrolled patients were followed up and the clinical features and treatment outcomes of the two groups were analyzed.

RESULTSCompared with PGR group, PPR group had different characteristics. They were older in age and had higher initial white blood cell count (P<0.05). T-cell ALL (T-ALL) and Philadelphia chromosome positive ALL (Ph+ ALL) were frequent in PPR group (P<0.05). Event-free survival (EFS) rate of PPR group was significantly lower than that of PGR group (P<0.05). 2 year event-free survival(EFS) rate of PGR group was (88.3±1.5)%, while the 2-year EFS rate of PPR group was (58.4±5.3)%. 5 year EFS rates of PGR and PPR were (80.8±2.1)% and (53.4±6.0)%, respectively. The EFS rate of PPR group was falling rapidly within 2 years. PPR group had higher relapse rate, and most relapses occurred within 18 months (P<0.05). PPR group had more high incidence of minimal residual disease (MRD) both on day 33 and on week 12 (P<0.05). No significant difference of EFS and relapse time was found between PPR and high risk PGR patients (P>0.05). In multi-variate regression analysis, the PPR, the presence of BCR-ABL1 and MLL were significantly unfavorable factors (P<0.05).

CONCLUSIONPrednisone response has been confirmed to be still great prognostic value and PPR children patients have poor outcomes generally. It is likely that the response to prednisone does not make much sense to high risk ALL patients.