Oncogenetic mutations combined with MRD improve outcome prediction in pediatric T-cell acute lymphoblastic leukemia

Abstract

Risk stratification in childhood T-cell acute lymphoblastic leukemia (T-ALL) is mainly based on minimal residual disease (MRD) quantification. Whether oncogenetic mutation profiles can improve the discrimination of MRD-defined risk categories was unknown. Two hundred and twenty FRALLE2000T-treated patients were tested retrospectively for NOTCH1/FBXW7/RAS and PTEN alterations. Patients with NOTCH1/FBXW7 (N/F) mutations and RAS/PTEN (R/P) germ line (GL) were classified as oncogenetic low risk (gLoR; n = 111), whereas those with N/F GL and R/P GL mutations or N/F and R/P mutations were classified as high risk (gHiR; n = 109). Day 35 MRD status was available for 191 patients. Five-year cumulative incidence of relapse (CIR) and disease-free survival were 36% and 60% for gHiR patients and 11% and 89% for gLoR patients, respectively. Importantly, among the 60% of patients with MRD <10^-4, 5-year CIR was 29% for gHiR patients and 4% for gLoR patients. Based on multivariable Cox models and stepwise selection, the 3 most discriminating variables were the oncogenetic classifier, MRD, and white blood cell (WBC) count. Patients harboring a WBC count ≥200 × 10⁹/L, gHiR classifier, and MRD ≥10^-4 demonstrated a 5-year CIR of 46%, whereas the 58 patients (30%) with a WBC count <200 × 10⁹/L, gLoR classifier, and MRD <10^-4 had a very low risk of relapse, with a 5-year CIR of only 2%. In childhood T-ALL, the N/F/R/P mutation profile is an independent predictor of relapse. When combined with MRD and a WBC count ≥200 × 10⁹/L, it identifies a significant subgroup of patients with a low risk of relapse.