Paper Details

ABSTRACT

BCR::ABL1 tyrosine kinase inhibitors (TKIs) have markedly improved outcomes in Philadelphia chromosome-positive acute lymphoblastic leukemia (Ph+ ALL). Common practice is to continue TKI therapy indefinitely in patients (pts) who do not undergo allogeneic stem cell transplantation, exposing them to cumulative toxicity, financial burden, and reduced quality of life. Unlike chronic myeloid leukemia, where treatment-free remission (TFR) is an established outcome with validated eligibility criteria, evidence supporting TFR in Ph+ ALL is limited.

We conducted a multicenter retrospective study of non-transplanted adults with Ph+ ALL who achieved molecular remission and subsequently discontinued TKI therapy to evaluate treatment discontinuation outcomes. Pts aged ≥18 years diagnosed between 2005 and 2024 at five U.S. centers were included. Complete molecular remission (CMR) was defined as undetectable BCR::ABL1 transcript by reverse transcription quantitative polymerase chain reaction (RT-qPCR) with assay sensitivity of at least MR4 (10?4). Loss of CMR was defined as re-emergence of BCR::ABL1 transcript on two consecutive RT-qPCR assessments using the same sensitivity.

Seventeen pts were identified; 2 were lost to follow-up shortly after TKI discontinuation (TKI-D) and excluded. Among the remaining 15 pts, TKI-D occurred due to treatment-related adverse events (n=10) or patient preference (n=5). Median age at TKI-D was 67 years (range, 32–91); 10 pts (67%) were female. BCR::ABL1 transcript type was p190 in 12 pts (79%), p210 in 1 (7%), both in 1 (7%), and unknown in 1 (7%). Cytogenetic abnormalities included BCR::ABL1 rearrangements with otherwise normal karyotypes in 7 pts (46%), complex karyotypes in 2 (14%), hyperdiploidy in 1 (7%), and were unknown in 5 (34%). Initial therapy consisted of chemotherapy + TKI in 12 pts (CALGB 10701 n=4; CALGB 10403 n=1; BFM-inspired n=1; Hyper-CVAD n=5; vincristine + dexamethasone n=1), inotuzumab + TKI in 1 patient, ruxolitinib and dexamethasone + TKI in 1 patient, and blinatumomab + TKI in 1 patient. All pts achieved CMR before TKI-D. Next-generation sequencing (NGS)-based measurable residual disease (MRD) assessment (sensitivity 10?⁶) was available in 3 pts; all were MRD negative at TKI-D. At TKI-D, 7 pts were receiving dasatinib, 3 ponatinib, 3 nilotinib, and 2 imatinib. Median TKI therapy duration before TKI-D was 57 months (range, 5–154).

After a median follow-up of 21 months (range, 4–111) from TKI-D, 6 pts (40%) lost CMR, including 1 morphologic relapse, and 14 (93%) were still alive. Median time to loss of CMR was 4 months (range, 1–8). Pts who lost CMR had a numerically shorter median duration of sustained CMR before TKI-D than those who maintained TFR (30 months [range, 3–55] vs 46 months [range, 26–129]), although this difference was not statistically significant (p=0.145). Notably, 4 of 6 pts (67%) who lost CMR had <48 months of sustained CMR before TKI-D.

All pts who lost CMR after TKI-D restarted TKI therapy. Two (33%) regained CMR within 9 months, including one patient with an IZKF1 deletion who regained CMR after dasatinib reinitiation but relapsed years later with acquisition of an ABL1 V299L kinase domain mutation and was successfully salvaged with ponatinib, inotuzumab, and dexamethasone. Three (50%) continue to have persistent quantifiable BCR::ABL1 transcripts without morphologic relapse. NGS MRD data after TKI-D were available for 2 of these 3 pts; both remained NGS MRD negative despite loss of CMR by BCR::ABL1 transcript monitoring. The only morphologic relapse after TKI-D occurred in a patient with an IZKF1 deletion who had the shortest pre-TKI-D CMR duration (3 months) and died 4 months after relapse.

TKI-D carried a substantial risk of loss of CMR (40%), whereas morphologic relapse was infrequent. All loss of CMR events occurred within the first year after TKI-D, most in pts with <48 months of sustained CMR before TKI-D, consistent with prior reports. Molecular findings in our study, including IZKF1 alterations and persistent NGS MRD negativity despite loss of CMR by BCR::ABL1 transcript monitoring, support integrating genomic risk stratification and NGS MRD to refine candidate selection for TKI-D. These results suggest that TKI-D may be feasible in carefully selected non-transplanted Ph+ ALL pts, particularly after >48 months of sustained CMR, although prospective validation incorporating molecular risk stratification is needed.