Real World Use of IDH2- Targeted Inhibitors in a Single Academic Medical Center Experience Since Enasidenib FDA-Approval

Background

Enasidenib, a novel inhibitor of isocitrate dehydrogenase 2 (IDH2), offers targeted therapy to the ~12% of acute myeloid leukemia (AML) patients in which a recurrent IDH2 mutation exists. While currently only FDA approved for use in relapsed/refractory disease, emerging data supports use among the treatment naïve. Adoption of and clinical experience with the drug outside of the clinical trial setting remains to be characterized. We reviewed the clinical histories of patients with IDH2 hotspot mutations identified by next generation sequencing (NGS) panel following FDA approval of and commercial availability of enasidenib.

Methods

We included all AML patients for whom an NGS panel was performed for the detection of somatic mutations in targeted regions of 42 genes between August 2017 and March 2019. Clinical and demographic data were obtained from the clinical records and abstracted for analysis. Statistical analyses were performed using the SAS/STAT (version 15.1) software.

Results

Twenty-eight individuals with IDH2 mutations were identified from 146 patients with available genomic data (14% of patients had IDH2 hotspot mutations). The cohort was predominantly male (n=18,64%) and white (n=24,86%). The median age was 69 years [range, 36-89 years], while median ECOG was 1 [range 0-3]. Six cases represented AML secondary to prior myelodysplastic syndrome or myeloproliferative neoplasm while the remainder represented primary disease. The most commonly identified disease subtypes were those with minimal differentiation (n=8,29%), myelomonocytic subtype (n=4,17%), and monocytic subtype (n=3,13%). Thirteen demonstrated adverse cytogenic risk while intermediate risk and favorable risk represented 11 and 4 cases respectively.

NGS panel results were available to providers at a median of 14 days [range 7-22] after bone marrow biopsy. The median number of coexistent mutations identified was 3 [range 0-11]. The most common coexistent mutations included NPM1 (n=7,25%), RUNX1 (n=6,21%), and FLT3 (n=5,18%). The median blast percentage on concomitant bone marrow biopsy was 65% [range, 3-80%]. Twenty-one percent of mutations (n=6) occured at residue R172 while 79% (n=22) altered R140.

In 9 cases an IDH2 mutation was detected at initial disease diagnosis. Only one of which was initially treated with enasidenib. The remaining newly diagnosed patients received 7 +3 induction (n=4, 44%), clinical trial protocol (n=3, 33%), or azacitadine monotherapy (n=1, 11%) respectively. Five of these treatment naïve patients (56%) achieved an initial clinical remission (i.e. CR or CRh or CRi). The remaining 19 (68%) patients had IDH2 mutations detected upon disease progression or relapse. These individuals had previously been treated with a median of one prior line of therapy [range,1-3].

Among the 19 with relapsed/refractory disease enasidenib was discussed with 14 (74%), all of whom would go on to receive the drug. Enasidenib was used as monotherapy in 9 (64%). Among the 5 patients with whom enasidenib was not discussed as potential therapy two enrolled in a clinical trial, two received a hypomethylating agent, and one received targeted therapy for another actionable mutation (e.g. midostaurin). When used as combination therapy, edasidenib was most often combined with hypomethylating agents (n=6) or venetoclax (n=3). Among those treated with enasidenib for relapse/refractory disease a complete remission was achieved in 5 (36%) compared to those treated with another agent (n=3,60%). Those treated with enasidenib spent a median 6.2 months [range 1.7-11.6] on next therapy while those treated with other agents spent a median 6.9 months [range 2.9-13.5]. Only two cases of differentiation syndrome were identified in this cohort.

Discussion

In a real-world cohort of AML with IDH2 mutant patients treated at a single center we identified a significant proportion of patients receiving enasidenib at some point in time. When considering enasidenib therapy, there appeared to be no barrier to treatment initiation. Those not receiving enasidenib were likely to enroll in a clinical trial, receive a hypomethylating agent, or another targeted therapy. IDH2 mutation rate, response rate, and toxicities in our cohort were comparable to published reports from clinical trials of enasidenib in AML. Further study is needed to directly compare enasidenib to other agents as salvage therapy for relapsed/refractory AML.