Higher-risk myelodysplastic syndrome (MDS) carries a significant risk of progression to acute myeloid leukemia (AML) and decreased survival [1]. Hypomethylating agents (HMAs) are first-line therapies for this population, demonstrating tolerability and improved overall survival (OS) [2,3,4,5]. However, HMAs are not curative for MDS, and thus, treatment of refractory or relapsed higher-risk MDS patients after HMA therapy represents a profound unmet medical need, with no approved therapeutics and median OS (mOS) of 4–5.6 months [6, 7].

Selective inhibitor of nuclear export (SINE) compounds are a class of novel and specific antagonists of the karyopherin exportin 1 protein (XPO1). Eltanexor is a reversible, oral SINE compound, which in non-clinical AML mouse studies showed robust single-agent activity, with improved tolerability and overall survival [8, 9]. We therefore analyzed eltanexor treatment in a phase 1 cohort of patients with higher-risk, HMA-refractory MDS.

Twenty patients with MDS were treated with single-agent eltanexor at a starting dose of 20 mg (n = 15) or 10 mg (n = 5) on days 1–5 of each week of each 28-day cycle (Additional file 1: Figure S1). The majority of patients were aged ≥ 75 years with de novo MDS, and all patients had MDS primary refractory to HMAs (Additional file 1: Table S1).

Of the 20 patients, seven (35%) had marrow complete remission (mCR), including two patients that also achieved hematologic improvement (HI). One (5%) additional patient had HI and stable disease (SD), and four patients (20%) had SD. Overall, the disease control rate (DCR) was 60% (95% CI 36.1–80.9), and the overall response rate (ORR) was 40% (95% CI 19.1–64.0). Of the 20 patients enrolled, 15 patients were evaluable for efficacy, as five 20 mg patients discontinued the study and were thus not evaluable for disease response; ORR for the efficacy-evaluable patients was 53.3% (95% CI 26.6–78.7; Table 1).

Three patients (20%) had HI and became transfusion independent ≥ 8 weeks. Among patients on eltanexor at the 20 mg dose, four patients (40%) reached mCR, including one patient who reached mCR with HI, one patient (10%) had HI and SD, and two (20%) had SD. Of patients on the 10 mg dose, three patients (60%) reached mCR, including one patient who had mCR with HI, and two patients had SD (40%).

After 3–4 cycles of treatment, levels of normal neutrophils and platelets improved, while blasts decreased, and patients had a median reduction in bone marrow blasts from baseline of -78.2% (range −86% to 33%; Additional file 1: Figure S2). mOS for the entire efficacy-evaluable cohort was 9.86 months. OS for patients who reached mCR was longer than for those who did not reach mCR, with a median of 11.86 vs 8.67 months, respectively (HR = 0.28, p = 0.11; Fig. 1). Additionally, mOS for patients with mCR was longer than mOS for patients with progressive disease (PD; 3.15 months [HR = 0.23, p = 0.08]), and patients with SD had longer OS than patients with PD (9.86 vs 3.15 months, HR = 0.38, p = 0.09).

Overall Survival. The median OS for all patients was 9.86 months. OS was higher in mCR patients (n = 7) than patients who did not reach mCR (n = 8): median 11.86 vs 8.67 months (HR = 0.28, p = 0.11), and longer than OS for patients with PD (n = 3, mOS = 3.15 months, HR = 0.23, p = 0.08). mCR = marrow complete remission; OS = overall survival; HR = hazard ratio; PD = progressive disease

None of the patients experienced dose-limiting toxicities (DLTs). The most frequently occurring adverse events (AEs) of any grade were non-hematologic and occurred more often in the 20 mg treatment group than the 10 mg group (Additional file 1: Table S2). Serious AEs (grade 3/4) were infrequent, and there were no grade 5 events. Of the 20 patients initially treated with eltanexor, seven patients (35.0%) required a dose delay, and eight patients (40.0%) required a dose reduction to 10 mg due to adverse events.

Novel agents to treat relapsed/refractory MDS patients are currently in development, though recent phase 3 trials to develop novel treatments for higher-risk MDS patients were unsuccessful, failing to meet the primary endpoints of improving OS [10, 11]. All patients in this study had MDS primary refractory to HMAs, 19/20 were considered higher risk by IPSS scoring, and 25% of patients had poor risk cytogenetics. Despite these factors, the ORR was still 53.3% in efficacy-evaluable patients, demonstrating an encouraging initial response to eltanexor monotherapy. Consistent with the selectivity of eltanexor for leukemic cells, responses included both a substantial reduction in blasts and commensurate increases in normal neutrophils and platelets, even for patients who achieved SD rather than a response [12]. A limitation of this report is the small patient population.

Overall, eltanexor represents a mechanistically novel, oral agent with good tolerability and robust single-agent activity conferring improved survival relative to historical controls, supporting the continued development of this novel agent.