Objective: To evaluate the prognostic factors in predicting relapse of childhood acute lymphoblastic leukemia(ALL) treated with CCLG-2008 protocol in order to optimize the chemotherapy for ALL to improve long-time survival.Methods: Three hundred fifty eight patients diagnosed with ALL and treated with the CCLG- ALL 2008 protocol between December 1st 2008 and December 31 st 2012 were enrolled in this study. All patients were followed up until September 1st, 2015. Prognostic impact of clinical features, response to treatment, biologic features were analyzed and multivariate analysis of predicted value was performed by Cox- regression analysis.Results: 358 patients were included in the study. Among them, 219 patients were boys and 139 patients were girls. The ratio of boy to girl was 1.57:1.00. The median age was 51 months(5 to 199 months). The median follow-up tine was 52 months(4 to 82 months). After treatment of CCLG-ALL2008 protocol, 79 patients suffered from relapse. Among 279 patients who didn’t suffer from relapse, 4 patients died because of treatment-related deaths, 5 patients were lost to follow-up. The 9 patients would not admitted into the univariate analysis. The relapse rate was 22.1%(79/358), the over-all survival rate was 86.3%(309/358).There were 48 boys and 31 girls who suffered from relapse.The ratio of boy to girl of relapse was 1.55: 1.00. 66 patients were younger than 10-year old and 13 were older or 10-year old. The negative control was the 270 patient who didn’t undergo relapse. Among them, 166 patients were boys and 104 patients were girls. The ratio of boy to girl was 1.60:1.00.250 patients were younger than 10-year old and 20 patients were older or 10-year old. No significant statistical differences were found in sex between the relapse and non-relapse group. The relapse rate of patients who were older than 10-year old were significant higher than the other, χ~2=5.8,P =0.015.64 patients suffered from extramedullary relapse, among them 6 patients have testis relapse; 2 patients have CNS relapse; 3 patients have both bone marrow and testis relapse, 4 patients have both bone marrow and CNS relapse.The relapse rate in the standard-risk, intermediate-risk and the high-risk groups were 13.3%, 17.6%, and 41.3%, respectively(P < 0.05). The relapse patients at very early stage, early stage, and late stage were 25 cases, 29 cases, and 25 cases, respectively, accounting for 31.6 %, 36.7 %, and 31.6 %.The median time of relapse time(the period from relapse to diagnose) of standard-risk, intermediate-risk and the high-risk groups were 31.2, 20.5, 22.7 months respectively. In the standard-risk group, 4 patients recurred in the very early stage, 5 patients recurred in the early stage, 9 patients recurred in the late stage. In the intermediate-risk group, 9 patients recurred in the very early stage, 8 patients recurred in the early stage, 6 patients recurred in the late stage. In the high-risk group, 12 patients recurred in the very early stage, 16 patients recurred in the early stage, 10 patients recurred in the late stage. The Fisher’s exact test showed no significant statistic differences and P =0.45. But by analysing the chart of frequency distribution of the relapse time of B-ALL patients, we found that the peak of relapse of the standard-risk patients was in the late stage, the intermediate-risk patients was in the very early and early stage, and the high-risk patients have two peaks,one in the early stage, the other was in the late stage.The relapse rates in patients of B-ALL with initial leukocyte counts >100×109/L, non-remission in 15 th day of induction(M3), and the level of minimal residual disease(MRD) on 12w(12w- MRD) > 10-4 were significant higher, the hazard ratio was 3.17(1.58~6.36), 1.87(1.07~3.30), 1.90(1.12~3.20), respectively(P <0.05).Conclusions: Patients with high-risk had the highest relapse rate treated by the CCLG-ALL2008 protocol. High initial leukocyte counts, minimal residual disease on D15 12 w were the independent prognostic factors for relapse of childhood B-ALL. |