Mutations in IDH1 and IDH2 portend a poor prognosis in adult patients with acute lymphoblastic leukemia Free

Background While prior studies suggest that IDH1 and IDH2 mutations confer a poor prognosis in acute lymphoblastic leukemia (ALL), data in adult patients (pts) remain limited. We evaluated the impact of these mutations in a cohort of adults with B- or T-cell ALL and compared their outcomes with IDH^wtpts.

Methods A retrospective analysis of pts with newly diagnosed B- or T-cell ALL treated with Hyper-CVAD- or BFM-based regimens from 8/2010-6/2025 was conducted to determine the incidence and clinical impact of IDH1 or IDH2 mutations compared to a cohort of IDH^wt pts.

Results Out of 449 pts with untreated T- or non Ph+ B-ALL, 21 (5%) had an IDH1 (n=7, 32%) or IDH2 (n=14, 67%) mutation; 11/385 tested (3%) among B-ALL (n=5 IDH1, n=6 IDH2) and 10/86 tested (12%) among T-ALL (n=2 IDH1, n=8 IDH2) (p=0.0007). Of the 7 IDH1 mutations, 6 (86%) were R132C and 1 (14%) was R132G; of the 14 IDH2 mutations, 12 (86%) were R140Q, and 2 (14%) were R140L.

The median age was 66 years (range, 28-83) in IDH^mut pts vs. 42 (18-87) in IDH^wt pts (p<0.0001). There was no significant difference in baseline white blood cell count in IDH^mutvs. IDH^wt (p=0.28). The majority of IDH^mutpts (n=9, 42%) had diploid karyotype; only 1 pt (5%) with B-ALL had a complex karyotype vs. the majority of IDH^wtpts having diploid/other karyotype (p=0.34). Out of IDH^mut pts with either a 28- or 81- next generation sequencing panel, the most common co-mutations were: 7/20 (35%) DNMT3A, 6/20 (30%) NOTCH1, 5/20 (25%) NRAS, 5/20 (20%) KRAS, 3/15 (20%) IKZF1, 3/20 (15%) TET2, 2/15 (13%) BCOR, 2/15 (13%) JAK1, 2/15 (13%) SRSF2, 1/20 (5%) TP53. Of the 7 pts with DNMT3A^mut, 1 (14%) was B-ALL and 6 (86%) were T-ALL (p=0.06). Of the 6 pts with NOTCH1^mut, all were T-ALL (p=0.03). Of the 10 pts with N/KRAS^mut, 8 (80%) were T-ALL (p=0.1).

The complete remission (CR) rate was 90% (19/21) overall: 10/11 (91%) in B-ALL and 9/10 (90%) in T-ALL, with a CR rate of 71% for IDH1^mutand 100% for IDH2^mut(p=0.09). Measurable residual disease (MRD)-negativity by multiparameter flow cytometry (FCM) after 1 cycle was achieved in 13/19 pts (68%) of pts overall: 70% in B-ALL and 67% in T-ALL. The FCM MRD-negativity rate at any time was 78% for both B-ALL and T-ALL. The rate of NGS-MRD negativity was 75% (3 out of 4 pts) overall, with 25% (1 out of 4 pts) achieving NGS-MRD negativity after cycle 1. Two out of 2 pts (100%) with B-ALL achieved NGS-MRD negativity overall, with 1 pt (50%) achieving after cycle 1. One out of 2 pts (50%) with T-ALL achieved NGS-MRD negativity overall, with no pts achieving NGS-MRD negativity after cycle 1.

After a median follow-up of 62 months (27-77), of the 10 responding B-ALL pts, 4 (40%) relapsed; all (100%) died. Two pts (20%) received allogeneic stem cell transplant (ASCT) consolidation (1 relapsed, received salvage treatment and is still alive in remission, and 1 died in CR) and 1 pt (9%) received CAR T, however relapsed, was salvaged and is alive in remission. Two pts (18%) are alive in CR without further consolidation, 2 pts (18%) died in CR, and 1 pt (8%) who attained a PR died.

After a median follow-up of 19 months (11-31) of the 9 responding T-ALL pts, 5 (56%) relapsed; all died. Three pts (33%) proceeded to ASCT (1 died in CR and 2 are alive in CR). One pt (11%) experienced bi-phenotypic switch and died. The 1 pt (10%) who achieved a PR died. One pt with IDH1^mutT-ALL and 2 pts IDH2^mutT-ALL received ivosidenib and enasidenib, respectively, in salvage: the IDH1^mutpt had no response and the 2 IDH2^mutpts achieved CR with a median duration of response of 17 months (3-31) before relapsing.

The median event-free survival (EFS) for IDH^mutpts was 16 months, with a 3-year EFS rate of 28%, vs. a median EFS of 69 months and a 3-year EFS rate of 60% in IDH^wtpts (p=0.0005). The median overall survival (OS) for IDH^mut pts was 33 months, with a 3-year OS rate of 43%, vs. a median OS of 78 months and a 3-year OS rate of 68% in IDH^wt pts (p=0.01). For IDH1 vs. IDH2, median OS was 33 months vs. 26 months (p=0.99) and median EFS was 11 months vs. 15 months (p=0.99).

ConclusionAdults with newly diagnosed B- or T-cell ALL harboring IDH1 or IDH2 mutations have poor outcomes, with 3-year OS of 40%. Co-occurring mutations include DNMT3A, NOTCH1, and NRAS, which were enriched in pts with T-ALL. Novel therapeutic strategies, including venetoclax and IDH inhibitors, are warranted in this high-risk subgroup.