Impact of Non-ABL1 Mutations on Outcomes in Patients with Chronic Myeloid Leukemia

Introduction:

Small retrospective studies have shown that myeloid mutations like ASXL1 and RUNX1 can impact progression, rates of major molecular remission (MMR), success of treatment free remission (TFR) and, more recently, rates of cardiovascular complications. Our aim was to study baseline disease and mutation characteristics of patients with CML who had DNA next generation sequencing (NGS) and to study the impact of non-ABL1 mutations on overall survival (OS), event-free survival (EFS) and failure free survival (FFS).

Methods:

Patients (pts) diagnosed with CML (with translocation 9;22 or the Philadelphia chromosome) who had a NGS panel done during their course of disease were selected for this study and data was collected on patient-, disease- and treatment characteristics as well as outcomes. EFS was calculated from the date of diagnosis to the date of loss of MMR, failure to achieve major cytogenetic remission by 12 months, failure to achieve bcr-abl <1% on international scale by 12 months, progression to accelerated or blast phase (by the modified MD Anderson criteria), or death, whichever came first. FFS was calculated as duration from initiation of first tyrosine kinase inhibitor (TKI) to second TKI either for intolerance or progression or date of failure of TFR. OS was calculated from the date of treatment initiation with first-line TKI to date of death. Time-to-event analyses were assessed using the Kaplan-Meier method. The log-rank test was used to compare OS, EFS and FFS.

Results:

In this single center retrospective study, a total of 129 pts with CML between 2016 – 2024, were included in this analysis: 115 pts were in chronic phase at diagnosis (CML-CP), 7 were in accelerated phase (CML-AP) and 7 were diagnosed with de novo blast phase CML (CML-BP). CML-AP and CML-BP were grouped as advanced phase CML.

From the 115 pts who were initially diagnosed in chronic phase, 67 (58.3%) were male and 48 (41.7%) were female. The median age at diagnosis was 54 years (range 16-92 years). Frontline treatment was imatinib in 51 (44.3%), dasatinib in 37 (32.2%), nilotinib in 16 (13.9%), bosutinib in 9 (7.8%) and asciminib in 1 (0.8%) pt. Allogeneic stem cell transplantation was performed in 8 pts, 6 of whom had progressed to advanced CML. 51 (44.3%) pts had at least one mutation on NGS testing. The most common mutations seen in CP-CML pts were ASXL1 (n=13), DNMT3A (n=7), RUNX1 (n=3), KMT2A (n=3), BCOR (n=3), NF1 (n=2), CSF3R (n=2), GATA2 (n=2) and CUX1 (n=2). Other pts had one mutation each in SUZ12, SH2B3, KDM6A, EZH2, IDH2, ETV6, U2AF1, GNAS, NRAS, NOTCH1, PHF6, WT1, ZRSR2 and TP53.

Of the 115 pts with CML-CP, 16 (13.9%) progressed to BP-CML and 4 (3.5%) progressed to AP-CML. A significantly greater number of pts with CP-CML harboring an ASXL1 mutation (4/13; 30.8%) or other non-ABL1 mutations (12/38; 31.6%) progressed to advanced phase CML compared to pts with no mutations (4/64; 6.25%) (p=0.0019). At a median follow up of 57 months (range 1-227 months), the OS of pts who did not have any mutations trended towards being longer compared to pts with ASXL1 mutation and other mutations (not reached vs 69 and 111 months; p=0.0662). The EFS of pts without any mutations was significantly longer compared to pts with ASXL1 or other mutations (221 vs 26 and 33 months; p=0.0004). In addition, the FFS of pts with ASXL1 and other mutations was significantly shorter when compared to pts with no mutations on (19 and 19 vs 66 months; p=0.0209).

From the 14 pts who were diagnosed with de novo advanced phase CML, 12 (85.7%) had at least one non-ABL1 mutation on NGS. The most common mutations seen in this subgroup were ASXL1 (n=5), RUNX1 (n=1), TP53 (n=1), BCORL1 (n=1), TET2 (n=1) and CUX1 (n=1).

Conclusion:

Confirming the findings of other small retrospective analyses, ASXL1 is the most common mutation seen in pts with CP-CML. Pts with ASXL1 or other non-ABL1 mutations were more likely to progress to advanced phase CML. The OS of pts with ASXL1 mutation trended towards being shorter and the EFS and FFS of pts with ASXL1 and other non-ABL1 mutations was significantly shorter than pts with CP-CML with no mutations. Our data corroborates the work of others in identifying a potential role for NGS analysis in the prognostication and risk-based management of pts with CML. Larger multi-institutional and prospective studies are needed to better determine the significance of these mutations in relationship to ELN response milestones in pts with CP-CML.