Clinical Impact of Additional Cytogenetic Aberrations and Complex Karyotype In Pediatric 11q23/MLL-Rearranged AML: Results from an International Retrospective Study

Abstract 762

Pediatric AML is a heterogeneous disease. We showed that differential outcomes of 11q23/MLL-rearranged AML depend on the fusion partner involved in the translocation in a large international retrospective study from 11 collaborative groups comprising 756 children (Balgobind et al, Blood 2009). In the current analysis we focused on the clinical characteristics and prognostic impact of additional cytogenetic aberrations (ACA). All patients with complete karyotypes (n=733) were centrally reviewed and classified for type and number of aberrations. Cases with 2 or more aberrations (including 11q23/MLL-rearrangement), were included in the ACA group (n=344/733, 47%). Numerical and structural ACA were divided into gains and losses of full or partial chromosomes, and balanced translocations were defined by the breakpoints. A complex karyotype was defined as 3 or more aberrations. ACA occurring in at least 10 patients were considered for statistical analysis, including differences in presenting characteristics (chi-square or Fisher's exact tests) and survival estimates (Kaplan-Meier method and Cox proportional hazards model for Event free survival (EFS), 5 year estimates and 95% confidence intervals (CI) are given). Multivariate analysis included the fusion partners with independent prognostic significance, as identified by Balgobind et al. Patients with ACA were found to be older (median age, 3.0 y vs 1.8 y, p=.001) and had a higher frequency of FAB M7 phenotype (4.5 % vs 1.1 %, p=.007) than those without ACA. For the complete cohort (n=733) estimates of EFS and overall survival (OS) were .44 and .56, and cumulative incidence of relapse (CIR) was .35 (standard errors .03). Patients with ACA showed significantly poorer clinical outcome than cases without ACA (EFS .38 vs .48, p=.002; OS .47 vs .62, p<.001; CIR .52 vs .38, p<.001). The most frequent specific ACA that was identified was trisomy 8 (130/344 cases, 38%). Patients with trisomy 8 were older (median age, 4.4 y vs 1.9 y, p<.001), presented with lower WBC counts (median 2.3 vs 27.1*10⁹/l, p<.001), were found more frequently in association with t(9;11)(p22;q23) (58% vs 42%, p=.01) and had a higher frequency of FAB M5 (78% vs 61%, p=.01) compared to patients without trisomy 8 cases. Trisomy 8 appeared to independently predict better clinical outcome within the ACA group (EFS .53 vs .29, p<.001; OS .61 vs .39, p=.003; CIR .35 vs .62, p<.001. and Hazard Ratio (HR) .57, CI .36-.92, p=.02). Trisomy 19 was found in 37/344 cases (11%). Trisomy 19 independently predicted poor clinical outcome in ACA cases. This effect appears to be mainly determined by refractory disease rather than relapse (probability of complete remission 62% vs 87%, EFS .17 vs .40, p=.003; OS .24 vs .50, p<.001; CIR .54 vs .51, p=.88; HR 1.77, CI 1.13–2.78, p=.01). Structural ACA were identified in 204 cases. Analyses in which all structural ACA were grouped together showed that they are of independent prognostic significance compared to only numerical aberrations (EFS .32 vs .47, p=.02; OS .42 vs .56, p=.10; CIR .59 vs .41, p=.003; HR 1.36, CI 1.07–1.80, p=.01). However, no specific structural ACA was present in more than 10 patients, precluding a more detailed analysis. Complex karyotype was found in 192/733 (26%) patients. Outcome was worse in these patients than those without complex karyotype (EFS .37 vs .46, p=.02; OS .45 vs .59, p=.003; and CIR .53 vs .42, p<.001), however, no independent prognostic significance was shown. In the context of the newly identified prognostic factors, the translocation partners 10p12, 6q27, 1q21 and 10p11.2 still showed independent prognostic value (HR 1.36, 2.29, .12 and 2.12 respectively). In conclusion, here we show that in addition to fusion partners, ACA have independent prognostic significance in pediatric 11q23/MLL-rearranged AML. Trisomy 8 was identified as an independent indicator of better prognosis in pediatric 11q23/MLL-rearranged AML. The presence of any ACA other than trisomy 8 predicts for worse clinical outcome. More specifically, trisomy 19 and structural aberrations were independent negative prognostic factors. Complex karyotype was a frequent finding and was a negative prognostic factor in univariate analysis only. This study shows that MLL-translocation partner and additional cytogenetic aberrations are important for patient stratification. Future studies should aim to understand the biology underlying these prognostic differences.