A phase 1 clinical trial of single-agent selinexor in acute myeloid leukemia

Abstract

Selinexor is a novel, first-in-class, selective inhibitor of nuclear export compound, which blocks exportin 1 (XPO1) function, leads to nuclear accumulation of tumor suppressor proteins, and induces cancer cell death. A phase 1 dose-escalation study was initiated to examine the safety and efficacy of selinexor in patients with advanced hematological malignancies. Ninety-five patients with relapsed or refractory acute myeloid leukemia (AML) were enrolled between January 2013 and June 2014 to receive 4, 8, or 10 doses of selinexor in a 21- or 28-day cycle. The most frequently reported adverse events (AEs) in patients with AML were grade 1 or 2 constitutional and gastrointestinal toxicities, which were generally manageable with supportive care. The only nonhematological grade 3/4 AE, occurring in >5% of the patient population, was fatigue (14%). There were no reported dose-limiting toxicities or evidence of cumulative toxicity. The recommended phase 2 dose was established at 60 mg (∼35 mg/m²) given twice weekly in a 4-week cycle based on the totality of safety and efficacy data. Overall, 14% of the 81 evaluable patients achieved an objective response (OR) and 31% percent showed ≥50% decrease in bone marrow blasts from baseline. Patients achieving an OR had a significant improvement in median progression-free survival (PFS) (5.1 vs 1.3 months; P = .008; hazard ratio [HR], 3.1) and overall survival (9.7 vs 2.7 months; P = .01; HR, 3.1) compared with nonresponders. These findings suggest that selinexor is safe as a monotherapy in patients with relapsed or refractory AML and have informed subsequent phase 2 clinical development. This trial was registered at www.clinicaltrials.gov as #NCT01607892.

Figure 1.

Selinexor PK. (A) Selinexor plasma concentration in patients with AML as a function of time over 48 hours for doses as listed administered on day 1. Points are means and standard error for the number of patients at each dose listed in parentheses on the graph. Median in vitro IC₅₀ is 86.4 ng/mL (or 195 nM). (B) C_max, AUC_{0-8 h}, and AUC_{0-48 h} were different across doses with high statistical significance (P < .0001 by ANOVA).

Figure 2.

PDn. (A) XPO1 mRNA levels in PBMCs isolated from patients before and after the first dose of selinexor. Each point represents data from patients across all doses normalized to the predose level, with the number of patients per time point indicated on the graph. The Student t test was used to determine statistical significance (*P < .05; ***P < .0001 vs predose). (B-D) Assessment of protein and mRNA levels in patient-derived BM blasts. XPO1, p53, FLT3, and Kit protein levels were analyzed by western blot at baseline and near the end of cycle 1 (day 24). GAPDH was used as a loading control. (E) FLT3 and Kit mRNA levels from corresponding BM blast samples assayed in panel D.

Figure 3.

Efficacy. (A) CR, CRi, MLFS, PR, SD, PD, ORR = (CR+CRi+MLFS+PR)/number of evaluable patients; disease control rate (DCR) = (CR+CRi+MLFS+PR+SD)/number of evaluable patients. (B) The best response based upon change in BM blast percentage is depicted for the 65 evaluable patients with available blast data postselinexor administration. Best response category is color-coded as indicated in the legend. The remaining 30 patients lacking post–drug blast data were assessed as SD (2), PD (14), or were NE (14).

Figure 4.

PFS and OS. Kaplan-Meier plots of PFS and OS. (A) For all patients, median PFS was 1.7 mo and median OS was 2.7 mo. (B) PFS for OR patients (5.1 mo) differed significantly from nonresponders (PD/SD) (1.3 mo; P = .008; HR, 3.1). (C) OS for OR patients (9.7 mo) differed significantly from nonresponders (PD/SD) (2.7 mo; P = .01; HR, 3.1). Statistical significance was determined using the Mantel-Byar method. Black marks indicate censored data at the last date a patient was known to be alive.