Paper Details

ABSTRACT

Background: As outcomes of frontline regimens for acute lymphoblastic leukemia (ALL) improve, myeloid neoplasms (MNs) arising after ALL therapy are increasingly recognized. These MNs are hypothesized to result from various mechanisms; some may represent true therapy-related MNs (t-MNs) due to cytotoxic chemotherapy, while others may be due to lineage switching. However, their risk factors and underlying clonal architecture and trajectories remain incompletely defined.

Methods: We retrospectively analyzed adults with newly diagnosed ALL who achieved remission after receiving hyper-CVAD- or mini-hyper-CVD-based therapy. For pts with available samples, bulk next-generation sequencing (NGS) was performed at one or more of the following time points: ALL diagnosis, remission, and MN diagnosis. In selected pts, targeted single-cell DNA and antibody sequencing (DAb-seq) was performed at ALL and/or MN diagnosis to determine the cellular distribution of shared mutations.

Results: From August 1992 to December 2024, 868 pts were evaluable. 744 pts (86%) had B-ALL and 124 (14%) had T-ALL. High-risk cytogenetic abnormalities were present in 110 pts (13%). Ph-like ALL occurred in 57/244 evaluable pts (23%), and TP53 mutations (TP53mut) in 58/337 (17%). Overall, 773 pts (90%) received hyper-CVAD, and 95 (10%) received mini-hyper-CVD-based therapy.

After a median follow-up of 135 months (95% CI, 127–149), 39 pts developed MNs; 4 pts who developed MNs within 1 year of ALL diagnosis were excluded because mixed-lineage disease at initial diagnosis could not be ruled out. Overall, 35 cases of post-ALL MN were observed, for an incidence of 4.0%. The median interval from ALL diagnosis to MN diagnosis was 39 months (range, 12–179). In univariable analysis, age ≥60 years (sHR, 2.6; 95% CI, 1.3–4.9; p=0.005) and a TP53mut at ALL diagnosis (sHR, 5.0; 95% CI, 1.5–16.9; p=0.008) were associated with a higher incidence of MNs.

When pts were stratified by age and TP53 status, the 5-year cumulative incidence (CuI) of MN was 16.4% (CI 95%, 4.8–34.1) among pts aged ≥60 years with TP53mut ALL (n=27), 5.3% (CI 95%, 1.4–13.5) among those aged ≥60 years with TP53-wild-type (WT) ALL (n=61), 0% (CI 95%, NA–NA) among those aged <60 years with TP53mut ALL (n=31), and 0.5% (CI 95%, 0.1–2.5) among those aged <60 years with TP53-WT ALL (n=218).

There were 3 cases of acquired KMT2Ar at MN diagnosis, suggesting a true chemotherapy-induced t-MN after Hyper-CVAD. 9 pts underwent paired bulk NGS at both ALL and MN diagnosis; 7 (78%) had at least one shared mutation; 5 pts (56%) shared at least one identical TP53mut, 1 (11%) shared DNMT3A and IDH2 mutations, and 1 (11%) shared a DNMT3A mutation. 2 of 4 pts who were TP53 WT at ALL diagnosis acquired TP53mut at MN diagnosis. Of 3 evaluable pts with TP53mut ALL who developed MNs, 2 had persistent TP53mut during ALL remission, suggesting the TP53mut was present in the non-lymphoid compartment.

Targeted DAb-seq was performed on 10 samples from 6 pts (ALL diagnosis, n=6; MN diagnosis, n=4). Among 4 pts with paired samples, DAb-seq identified 2 distinct patterns of ALL-to-AML transformation. In two pts, the AML arose by polyclonal expansion, in which a myeloid clone (genetically distinct from the ALL clones) was selected and expanded. In one pt, prior therapy selected a PPM1D-mutant clone within the myelomonocytic compartment, showing a classic t-AML. In the other pt, an IDH1-mutant clone within the myeloid compartment was selected and expanded at AML, showing an example of AML leukemogenesis driven by a driver mutation. In the remaining two pts, the AML arose by monoclonal expansion, in which the same clone persisted from ALL to AML while switching the immunophenotype, retaining DNMT3A and biallelic TP53 mutations in one pt and a FLT3-ITD in the other. Two single-timepoint B-ALL samples similarly harbored myeloid lesions, such as NRAS or DNMT3A mutations. Additional ALL and remission samples are undergoing sequencing. DAb-seq of TP53mut ALL is also ongoing, and updated analyses will be presented at the meeting.

Conclusion: Although MNs were rare among adults who achieved remission after frontline chemotherapy for ALL, older age and TP53mut at ALL diagnosis were associated with an increased incidence of MNs. Among pts with paired sequencing, most shared at least one mutation at both time points, suggesting a common clonal origin. DAb-seq suggested that multiple mechanisms may underlie the development of MNs after ALL therapy.