NUP98::NSD1 fusion is associated with poor outcomes in Acute Myeloid Leukemia regardless of co-occurring mutations and treatment regimen

Abstract

Abstract Introduction: Fusions of the NUP98 gene are a recurring feature in acute myeloid leukemia (AML) and occur in approximately 7% of children and young adults. NUP98::NSD1 (NUP/NSD) is the most common NUP98 fusion and associated with co-occurring FLT3/ITD (ITD+) mutations. In a large cohort of children and young adults with AML, we evaluated outcomes for NUP/NSD AML and whether changes in therapy or co-occurring mutations affected survival. Methods: A total of 2231 diagnostic specimens from patients treated on 4 consecutive CCG/COG AML trials were screened, of which 111(5%) were positive for the NUP/NSD fusion; CCG2961 (n=15), AAML03P1 (n=3), AAML0531 (n=35), and AAML1031 (n=58). Outcomes were known for 107 (96%) patients. All ITD+ were included regardless of allelic ratio. Co-occurring genomic features were collated and outcome data reported. The Kaplan-Meier method was used to determine 5-year EFS and OS from study entry, RR from CR (morphology <5% blasts) and DFS and OS from HCT. Results: In the 111 NUP/NSD patients, 84 (76%) had a co-occurring ITD, and 57 (51%) had WT1 mutations. There was no difference in prevalence of WT1 mutations in those with and without ITD (53% v. 48%, p=0.824). Compared to non-NUP/NSD patients, copy number alterations in chromosomes 5, 8 and 9 were more prevalent in NUP/NSD patients with del5q, tri8 and del9q seen in 5.2%, 18.6% and 14.4% of patients (p=0.002, p=0.0003 and p=0.04, respectively). Patients with NUP/NSD had a CR rate of 40% and MRD+ rate of 74% compared to that of 79% and 27% for those without NSD1 (p<0.001 for both). The EFS and OS for patients with NUP/NSD were 17% and 39% compared to 48% and 64% for those without NUP/NSD fusion (p<0.001 for both). Relapse risk was significantly higher in the NUP/NSD cohort (64% vs. 44%, p=0.002). There were 75 (68%) NUP/NSD patients with available end of induction MRD and outcome data (MRD positive, n=55; MRD negative, n=20). Those with and without MRD had a DFS of 15% and 30% (p=0.019) and OS of 36% and 51% (p=0.16). NUP/NSD patients with and without ITD had a nearly identical EFS and OS of 17% v. 19% and 40% v. 38%, respectively (p=NS for both). The presence or absence of WT1 mutations did not impact outcomes. Three distinct interventions (GO exposure, HCT in first CR (CR1) and sorafenib treatment for ITD+ were evaluated for potential efficacy in this high-risk cohort. GO-exposure: 25 NUP/NSD patients received GO on AAML03P1 or AAML0531. Those with and without GO exposure had a near identical CR, MRD, EFS, OS and RR (all p=NS), highlighting lack of efficacy of GO in this cohort. Analysis of GO exposure in NUP/NSD/ITD+ patients demonstrated no benefit. Similar comparison in NUP/NSD/ITD neg patients was limited due to the small cohort size (n=8). HCT in 1st CR: 22 patients received HCT in CR1 vs. 11continued on consolidation chemotherapy. Those who received HCT had near identical EFS and OS from time of transplant as chemotherapy recipients (p=NS for both). Analysis in the NUP/NSD/ITD+ patients also found that HCT in CR1 did not modify outcome. Sorafenib therapy: AAML1031 study demonstrated efficacy of sorafenib in ITD+ patients. We evaluated the outcome of 19 NUP/NSD/ITD+ patients who received sorafenib on AAML1031 compared to patients treated on a similar backbone without sorafenib (n=61). EFS from study entry and RR from CR for those treated with and without sorafenib was nearly identical at 16% and 17%, and 75% and 70% (p=NS for both) highlighting that sorafenib also lacked efficacy in this cohort. Conclusion: Patients with NUP/NSD AML have dismal outcomes with no improvement in survival with interventions that have been shown to be effective in other patient subsets. Even those who achieve an MRD negative CR had poor outcomes, suggesting that initial response is not predictive of eventual outcome. Lack of benefit with HCT in CR1 is particularly concerning. Historically, co-occurring ITD has been considered determinant of outcome. Our analysis demonstrates that ITD does not modify the outcome in NUP/NSD patients, and we detected no therapeutic benefit of sorafenib in NUP/NSD/ITD+ patients. Thus, consideration of alternative targeted therapies (e.g. menin inhibitors, venetoclax, gilteritinib, palbociclib) and novel pre and post HCT approaches should be studied in this very high-risk group in future cooperative trials to optimize outcomes. Studies to identify therapeutic targets in NUP/NSD AML are ongoing.