Clinical Impact of Additional Cytogenetic Aberrations, cKIT- and RAS Mutations and Other Factors in Pediatric t(8;21)-AML

Introduction

Whilst pediatric t(8;21)-positive AML has an overall survival rate of up to 80-90%, the relapse rate at 30-40% remains unacceptably high. Factors that reliably discriminate between patients with t(8;21) who do or do not relapse have been poorly described. The prognostic impact of additional cytogenetic aberrations and of cKIT or RAS mutations in this subgroup remains unclear. In this large retrospective cohort study, supported by the International-BFM study group, we aimed firstly to determine outcome in relation to these aberrations and, secondly to identify any treatments associated with outcome.

Methods

Data on diagnostics, treatment, response, relapse or other events and follow-up were collected from 19 childhood cancer study groups, who provided data from 916 pediatric AML-patients positive for t(8;21) and/or AML1-ETO fusion gene. Karyotypes were reviewed and classified (CH&BB). Aberrations occurring in at least 10 patients were considered for analyses and involved loss of sex chromosome, deletions of chromosome 9q (del9(q)), abnormalities of chromosome 7q, gain of chromosome 4 (+4) and gain of chromosome 8. A total of 228 samples could be evaluated for cKIT and/or RAS mutations. Significant characteristics in explorative analysis were used in 2 multivariate Cox proportional-hazard regression models for assessing effects of induction and overall treatment elements. Induction was defined as first two courses of chemotherapy. Endpoints were probabilities of achieving complete remission (pCR), of event free survival (pEFS), and of overall survival (pOS) and cumulative incidence of relapse (CIR). Only results of multivariate analyses are shown.

Results

In final analyses 838 patients with a full karyotype were included. CR was achieved in 92% of the patients. Five-year pOS was 74% (SE 1.5%), five-year pEFS 58% (SE 1.7%) and five-year CIR 26% (SE 1.5%). WBC count >20 x10⁹/L was associated with poorer outcome for all endpoints (p<0.01). CNS involvement at presentation led to higher CIR (p<0.01). Twenty five percent of the patients had cKIT mutation(s). No differences in clinical or leukemic characteristics between mutated and wild-type patients were identified. Seven out of 10 patients with +4 had cKIT mutation(s) (p<0.01). Patients with RAS mutations (13%) had higher WBC counts (p=0.03). No clinical differences between N-RAS (n=17) and K-RAS (n=5) mutated patients were found. No associations between other additional aberrations and cKIT- or RAS mutations were identified. Patients with del(9)(q) (n=104) had lower pCR (p=0.01), without influencing pEFS or pOS. Additional +4 (n=21) was strongly associated with higher CIR (p<0.01) and poorer survival rates (5ypEFS 18% versus 59%, p_logrank<0.01 and 5ypOS 33% versus 75%, p_logrank<0.01). Presence of other cytogenetic aberrations including complex karyotype or cKIT or RAS mutations did not affect outcome.

Anthracyclines >150mg/m² used in first induction course was associated with higher pCR (p=0.04) and higher pOS (p=0.03). Etoposide >500mg/m² used in first course was associated with higher pEFS (p=0.01) and pOS (p=0.02). The use of HD-AraC 3g/m² during induction was associated with higher pCR (p=0.02). When including consolidation therapy, cumulative dosages of cytarabine >30g/m² and etoposide >1500mg/m² were associated with lower CIR (p=0.02 and p<0.01 respectively).

Conclusion

This study identifies novel negative prognostic factors in pediatric t(8;21)-AML patients - del(9q) associated with an inferior pCR, and +4 significantly associated with an increased relapse rate and inferior survival. This challenges the favorable classification of patients with t(8;21) and +4. The presence of del(9q) may indicate the need for intensified induction therapy in view of the lower pCR. Although cKIT or RAS mutations did not influence outcome in conventionally treated t(8;21)-patients in this study, mutated patients may benefit further from treatment with tyrosine kinase inhibitors.

This study suggests benefit from protocols containing high-dose anthracyclines, etoposide and HD-AraC 3g/m² during induction, and the use of cumulative dosages of cytarabine >30g/m² and of etoposide >1500mg/m² in pediatric t(8;21)-positive AML. Optimal dosages and combinations of these agents should be established in prospective randomized trials.