Somedeb Ball, MD from the H. Lee Moffitt Cancer Center & Research Institute speaks about Molecular annotation of extramedullary acute myeloid leukemia to identify prevalence of targetable mutations.
Link to Abstract:
https://meetinglibrary.asco.org/record/195837/abstract
Background:
In individuals with acute myeloid leukemia, extramedullary (EM) involvement, such as myeloid sarcoma (MS) and leukemia cutis (LC), is uncommon (AML). The mutational landscape of EM-AML is unknown, as is the concordance of sequencing data from EM vs. non-EM sites (blood or bone marrow) and the possibility of customized targeted treatment in this patient population.
Methods:
Clinical and genomic data on EM-AML patients treated at Moffitt Cancer Center, Memorial Healthcare System, and University of Miami, as well as sequenced cases at a central laboratory, were collected in a multicenter retrospective analysis. Data from next-generation sequencing (NGS) panels examined 24 406 genes, with 15 genes covered by all panels, including IDH1, IDH2, KIT, KRAS, NPM1, NRAS, and TP53, among others. SPSS was used to calculate survival estimates using Kaplan-Meier statistics and multivariate analysis using Cox-regression (v.26).
Results:
There were 58 individuals with EM-AML in our research. The median age at diagnosis was 62 years, and males made up 55 percent of the patients. Our group included 34 (59%) MS patients and 19 (33%) LC patients. 60 percent of evaluable patients (n=45) had EM-AML during relapse, and 31 percent had isolated EM illness. In a multivariate analysis that included disease setting (new diagnosis vs. relapse) and ELN risk category, patients with LC had a significantly worse median overall survival (OS) than those with MS (5.7 months vs. 21.9 months, p= 0.008). The pattern of EM involvement (MS vs. LC) remained an independent prognostic factor for OS (p= 0.04). In 48 individuals, the results of NGS done at EM presentation were accessible, with 19 of them having NGS data from the EM location. On EM site NGS, the most often mutated genes were NRAS (37 percent) and NPM1 (non-EM site NGS) (28 percent ). According to EM NGS, 52 percent of patients exhibited a targetable genomic change, with IDH mutations accounting for 37 percent, NPM1 for 21%, FLT3 for 5%, and MLL-PTD for 11%. At EM-AML, there was considerable discordance in targetable mutations between EM and non-EM NGS in five out of nine evaluable patients (two with concurrent M+EM illness). Three of the four patients who were given IDH1/2 inhibitors based on EM NGS had a full response.
Conclusions:
EM-AML has a different molecular architecture, and its prognosis is worse in LC patients than in MS patients. We conclude that EM site NGS is crucial in EM-AML patients since 52 percent have potentially targetable mutations and might benefit from particular targeted medicines.