Relapse
The CIR in pediatric patients was significantly lower than that in adult patients (21.8%±5.4% vs. 34.1%±5.5%, P=0.037, figure 1A1). The difference in CIR was also seen at 2 year interim (30.8%±6.4% in children vs. 43.1%±6.1% in adults, P=0.046, supplemental figure 1A). Consistent with previous report [13], MDAC-based consolidation greatly lowered disease relapse compared to SDAC-based regimen (P<0.001, supplemental figure 2). To investigate the impact of age on CIR, we analyzed the CIR in patients receiving MDAC- vs. SDAC-based regimens respectively. When consolidation regimens were taken into consideration, there was no statistical difference in CIR between the two age cohorts (MDAC-based regimen: children 11.9%±5.0% vs. adults 16.3%±6.7%, P=0.467; SDAC-based regimen: children 45.8%±12.0%, P=0.517 vs. adults 47.5%±7.6%, figure 1A2 and A3).
Multivariate analysis indicated that both EML (with vs. without, HR 2.20, 95% CI 1.11-4.35, P=0.024) and consolidation regimens (MDAC-based vs. SDAC-based, HR 0.20, 95% CI 0.10-0.39, P<0.001) were independent risk factors for disease relapse. Patients harboring c-KIT mutations (mutant vs. wild type [WT], HR 1.80, 95% CI 0.92-3.50, P=0.088) also showed a trend toward higher risk of recurrence, although this was not statistically significant. Other variables, including gender (male vs. female), age (≥14 vs. <14 years), WBC count at the time of diagnosis (≥20×109/L vs. <20×109/L), immunphenotype including CD34, CD33, CD13, CD117, HLA-DR, CD19 and CD56 (positive vs. negative), additional chromosome abnormalities (with vs. without) and FLT3-ITD status (mutant vs. WT) did not affect disease recurrence. Detail was shown in table 3.
To examine whether EML and c-KIT mutations behaved similarly on relapse in terms of patients’ age, we performed subgroup analysis which confirmed EML as the risk factor for relapse only in adult patients (with 54.6%±10.8% vs. without 26.0%±6.1%, P=0.002, figure 2A2), but not in pediatric patients (with 38.3%±18.0% vs. without 19.3%±5.5%, P=0.082, figure 2A3). Similarly, c-KIT mutations were associated with higher risk of relapse only in adults (adults, mutant 52.4%±10.2% vs. WT 22.8%±7.3%, P=0.005, figure 3A2; children, mutant 20.0%±17.9% vs. WT 23.6%±6.3%, P=0.372, figure 3A3). The adverse effects of EML and c-KIT mutations on disease recurrence in adults were not improved by choice of consolidation regimens (supplemental figure 3 and 4).