PF261 A RANDOMIZED, OPEN‐LABEL, PHASE II STUDY OF SELINEXOR VERSUS PHYSICIAN'S CHOICE (PC) IN OLDER PATIENTS WITH RELAPSED OR REFRACTORY ACUTE MYELOID LEUKEMIA (AML)

Background: Despite advances in AML, options for relapsed disease are limited. Inhibition of nuclear export protein 1 (XPO1), which transports tumor suppressor proteins out of the nucleus, represents a novel therapeutic target. Selinexor has demonstrated activity in AML cell lines, animal models, and patient samples. A phase I dose‐escalation study of selinexor in patients with advanced hematologic malignancies showed an CR/CRi of 9.7% in 41 patients. Aims: Selinexor was compared with treatment of PC in patients aged ≥60 years with relapsed/refractory AML who were ineligible for intensive chemotherapy and/or transplantation. Methods: Patients were randomized in a 2:1 ratio to receive either selinexor or PC treatment (Best supportive care [BSC] and hydroxyurea, or BSC plus low‐dose cytosine arabinoside, or BSC plus DNA methyltransferase (DNMTi: azacitidine or decitabine). Patients had to receive at least 2 cycles of DNMTi to be eligible but were not necessarily refractory to DNMTi. Stratification factors included the duration of their first complete remission (CR) on prior therapy and age (<70 years). Patients were eligible for study inclusion following one prior AML treatment including treatment with an HMA. Primary objective was OS. Results: 175 patients were in ITT population, 118 received treatment with selinexor 60 mg twice weekly and 57 patients were enrolled in PC. Withdrawal prior to randomization occurred in 2% on selinexor and 21% on PC. Baseline characteristics between the arms included lower baseline ANC (<0.5 x 10 9 /L, 42.4% vs. 21.1%); greater TP53 abnormalities (11.9% vs 5.3%); and increased proportion of prior MDS (11.0% vs 5.3%) in the selinexor group. The median OS for selinexor versus PC was 94 days versus 170 days, (HR = 1.18 [95% CI 0.79–1.75]; P  = 0.4221). CR/CRi was observed in 12% of patients on selinexor, and 4% of the PC‐treated patients. The most common TEAEs in the selinexor arm included nausea (59.1%), decreased appetite (55.7%), fatigue (46.1%), diarrhea (40.0%), thrombocytopenia (33.9%), vomiting and pyrexia (27.8%). In PC‐treated patients, the most common TEAEs included febrile neutropenia (35.6%), constipation (33.3%), fatigue (28.9%), pyrexia (28.9%), anemia (26.7%), thrombocytopenia (24.4%), and dyspnea (22.2%). The most common SAEs in the selinexor and PC groups were febrile neutropenia and pneumonia. Febrile neutropenia occurred at a lower rate in the selinexor group (17.4%) than in the PC group (35.6%). The incidence of TEAE leading to death was identical in each arm (20%). Summary/Conclusion: The primary endpoint was not met and selinexor treatment did not show a significant difference in median OS compared with treatment of PC in this patient population. There was no increased TEAE of death between the 2 arms. There were several factors that may have led to this result. First, the trial was designed to allow DNMTI‐experienced patients, but not necessarily DNMTi‐refractory patients, so the survival estimates in the PC were subsequently influenced by responses to SOC DNMTi therapy in patients who were not refractory to DNMTi therapy. Second, the imbalances in the presence ofTP53, depth of neutropenia at enrollment, and transformed MDS indicate a higher risk population in the selinexor arm. Therefore, the results of this study support further investigation of selinexor in carefully selected populations of AML patients.