Olutasidenib alone or combined with azacitidine in patients with mutant IDH1 myelodysplastic syndrome

Key Points

• •Olutasidenib alone or with azacitidine elicited high response rates and durable remissions in intermediate- to very high-risk mIDH1 MDS.
• •The safety profile of olutasidenib + azacitidine was comparable to that of either monotherapy; no new safety signal reported in combination.

Visual Abstract

Abstract

Olutasidenib, a potent, selective, oral small-molecule inhibitor of mutant isocitrate dehydrogenase 1 (mIDH1), is US Food and Drug Administration approved for mIDH1 relapsed/refractory (R/R) acute myeloid leukemia based on results from the pivotal cohort of a multiarm phase 1/2 trial that also enrolled patients with myelodysplastic syndrome (MDS). We report pooled data evaluating olutasidenib as monotherapy or combined with azacitidine in R/R and treatment-naïve (TN) higher-risk mIDH1 MDS. Twenty-two patients (median age, 74 years; 59% male) with intermediate- to very high-risk MDS (monotherapy, n = 6 [4 R/R and 2 TN]; combination, n = 16 [11 R/R and 5 TN]) were analyzed. The most frequent adverse events were fatigue and cytopenias. Differentiation syndrome occurred in 3 patients (14%), including 1 (5%) with grade 3 severity. QT prolongation occurred in 1 patient receiving combination therapy. ORR was 59% (complete remission [CR], 6/22 [27%]; marrow CR, 7/22 [32%]) in intent-to-treat (ITT; n = 22) and 68% (CR, 6/19 [32%]; marrow CR, 7/19 [37%]) in response-evaluable patients (n = 19). ORR (ITT population) was 33% (2/6) for monotherapy (3/6 patients received half the recommended dose or less) and 69% (11/16) for combination therapy. Median time to response was 2 months (range, 1-13), median duration of response was 14.6 months (95% confidence interval [CI], 5.8-32.8), and median overall survival was 27.2 months (95% CI, 6.9-37). Sixty-two percent and 67% of patients who were transfusion dependent at baseline achieved 56-day red blood cell and platelet transfusion independence, respectively. Olutasidenib with or without azacitidine demonstrated encouraging clinical activity and tolerability in patients with higher-risk mIDH1 MDS. This trial was registered at www.ClinicalTrials.gov as #NCT02719574.

Introduction

Myelodysplastic syndrome (MDS) is a heterogeneous group of hematologic malignancies associated with high morbidity and mortality, with potential for transformation into acute myeloid leukemia (AML). The incidence of MDS is 4.5 per 100 000 people per year overall and increases with advanced age (26.9 per 100 000 for ages 70-79 years).¹ Survival of untreated patients with MDS ranges from <6 months to several years.²

MDSs are classified into risk categories based on the severity of cytopenias, percentage of bone marrow blasts, and cytogenetic abnormalities.³ Patients with lower-risk MDS are managed with therapies aimed at improving cytopenias, whereas those with high- or very high-risk MDS are treated with hypomethylating agents (HMAs) such as azacitidine (AZA) or decitabine.⁴ Although many patients initially respond to HMAs, most do not achieve complete remissions (CRs), and the disease typically progresses.5, 6, 7, 8 Allogeneic hematopoietic stem cell transplantation (HSCT) remains the only putatively curative option for MDS, but it is not an option for many patients.^9,10 Given the limited treatment options for higher-risk MDS, there is a need for novel therapies, especially in the relapsed/refractory (R/R) setting.6, 7, 8

Targeted agents against an identified, recurrent genetic alteration have demonstrated clinical benefit in patients with AML and now increasingly in MDS.11, 12, 13, 14, 15 Mutations in the gene encoding isocitrate dehydrogenase 1 (mIDH1) enzymes occur in ∼3% to 4% of patients with MDS^16,17; this is less frequent than their occurrence in AML, which is estimated at 7% to 14%.18, 19, 20, 21, 22, 23, 24 Studies have shown that patients with mIDH1 MDS have a higher risk of leukemic transformation and shorter event-free survival (EFS) and overall survival (OS).¹⁷ Wild-type IDH1 (wtIDH1) has critical functions in cellular metabolism and homeostasis by converting isocitrate to α-ketoglutarate. Mutant IDH1-R132 has gain-of-function enzymatic activity that converts α-ketoglutarate to the oncometabolite, 2-hydroxyglutarate.²⁰ Abnormal production of 2-hydroxyglutarate disrupts epigenetic regulation and impairs hematopoietic differentiation, which contributes to oncogenesis.^20,25,26

Olutasidenib is a small-molecule inhibitor of mutant IDH1 protein that is approved by the US Food and Drug Administration for the treatment of mIDH1 R/R AML. Olutasidenib is highly selective, targeting only mIDH1 and leaving intact wtIDH1 function (50% inhibitory concentration to wtIDH1, 22 400 nM).²⁷ Inhibition of wtIDH1 can induce oxidative stress, disrupt homeostatic function, and decrease cell viability, particularly affecting hematopoietic stem cells under stress conditions.^27,28 Olutasidenib is characterized by a low molecular weight (formula weight 355) and small structure, which allows it to occupy less space in the binding pocket of mIDH1-R132 dimers, making it resistant to displacement by some second-site mutations. A phase 1/2 study was conducted to evaluate olutasidenib alone or in combination with AZA in patients with AML or MDS harboring mIDH1-R132 (ClinicalTrials.gov identifier, NCT02719574). The pivotal cohort of olutasidenib monotherapy in patients with R/R AML has been previously reported, with 32% of patients achieving CR, 35% achieving CR + CR with partial hematologic recovery, and a median duration of CR + CR with partial hematologic recovery of 25.9 months.²⁹ Herein, we present the final results of the phase 1/2 trial of patients with R/R or treatment-naïve (TN) MDS harboring mIDH1.

Methods

Study design

This subset analysis was conducted on patients with MDS pooled from multiple cohorts of a completed phase 1/2 open-label, nonrandomized, multicohort, multicenter study (ClinicalTrials.gov identifier, NCT02719574) that enrolled patients aged ≥18 years with pathologically confirmed AML or MDS harboring mIDH1-R132. The phase 1 dose-escalation stage tested single-agent olutasidenib starting at doses of 150 mg once daily, with a parallel escalation arm initiated later for olutasidenib in combination with AZA (additional details see supplemental Appendix). Olutasidenib administered orally at 150 mg twice daily in continuous 28-day cycles was selected as the recommended phase 2 dose whether given alone or in combination with AZA (AZA given IV or subcutaneously at 75 mg/m² daily for 7 days).

This study was conducted following the Declaration of Helsinki and good clinical practice guidelines. The protocol was approved by institutional review boards or local ethics committees at participating study sites. All patients provided written informed consent before screening and enrollment.

Patients

Inclusion criteria were adults (≥18 years) with intermediate- to very high-risk MDS, as defined by a revised International Prognostic Scoring System (IPSS-R) score >3, Eastern Cooperative Oncology Group performance status 0 to 2, acceptable liver and renal function, and baseline corrected QT interval by Fridericia of ≤450 milliseconds. Key exclusion criteria included any prior treatment with an IDH1 inhibitor; systemic anticancer therapy or an investigational agent within 21 days before the first dose of study drug; radiation therapy or immunotherapy within 1 month before the first dose; the presence of symptomatic central nervous system involvement requiring urgent intervention; uncontrolled infections or metabolic disorders; and clinically significant heart disease or other serious nonmalignant disease that could compromise study objectives.

The patients described were enrolled at 13 sites in the United States, Europe, United Kingdom, and Australia.

Safety end points

Safety was a primary end point for both phase 1 and phase 2. Patients who received at least 1 dose of olutasidenib (safety/full analysis set [FAS]/intent-to-treat [ITT] population) were monitored for incidence and severity of adverse events (AEs), clinical laboratory abnormalities, and changes in electrocardiogram parameters. Additional details on safety end points are provided in the supplemental Appendix.

Efficacy end points

The phase 1 portion of the study monitored response to therapy (CR, marrow CR, partial remission [PR], and stable disease [SD]) as a secondary end point. The phase 2 portion specified a primary efficacy end point based on CR rate and a secondary efficacy end point of best overall response rate (ORR; CR + marrow CR + PR). Response and hematologic improvement (HI) definitions were derived from the modified International Working Group criteria for MDS (2006),³⁰ provided in supplemental Table 1, and were assessed by investigators. Additional secondary end points in the phase 2 study were time to response, duration of response (DOR), EFS, OS, and 56-day transfusion independence (TI). TI was defined as 56 consecutive days after baseline with no transfusion (red blood cells [RBCs] or platelets) in patients who were transfusion dependent at baseline (ie, required a transfusion during the 8 weeks before receiving study drug). Post hoc analyses included response in patients meeting the MDS diagnosis criteria by the World Health Organization (WHO) 2022 definition and a composite response rate combining patients with CR + marrow CR + HI in SD responders.

Statistical analysis

Descriptive statistics were summarized for all analyses unless otherwise indicated. Binomial end points were presented by number and percentage of patients with Clopper-Pearson 95% confidence intervals (CIs). Median time-to-event values (DOR, OS, EFS, time to response) were calculated using the Kaplan-Meier method in the FAS population with 95% Brookmeyer-Crowley CIs. Response rates were determined in the FAS/ITT and response-evaluable populations, which excluded patients who did not receive a postbaseline response assessment. DOR was calculated only for patients who achieved a response (PR or better), from the time of first response until relapse or death. Patients without an event were censored at their last response assessment. Patients who discontinued study drug to undergo HSCT were followed until relapse or censored at their last response assessment. For patients who had no event before the end-of-study cutoff, EFS was censored on the date of last visit, and OS was censored at the last known date the patient was alive.

Results

Patients

A total of 22 patients with mIDH1 MDS enrolled in the phase 1/2 study, including 6 patients who received monotherapy (R/R, n = 4; TN, n = 2) and 16 patients who received combination therapy (R/R, n = 11; TN, n = 5; Figure 1). The median age was 74 years (range, 59-87), and 59% were male. The median time from diagnosis was 15.3 months (range, 1-77) for the monotherapy group, 14.3 months (0-191) for the combination therapy group, and 15.2 months (range, 0-191) for the pooled population. Demographics and baseline characteristics are summarized in Table 1. Patients were R/R to a median of 1 prior regimen (range, 1-4). Notably, all but 1 patient with R/R MDS (n = 15) had received prior HMA. The patient without prior HMA received prior cytarabine and idarubicin (7+3) induction therapy. Three patients (14%) received 2 prior HMA regimens. No patient had prior IDH1 inhibitor therapy. Per eligibility criteria, all patients had an IPSS-R score >3, with 86% of patients in the “high” (IPSS-R, >4.5-6) or “very high” (IPSS-R, >6) risk categories at baseline.

Figure 1.

Study design and patient flow. The study pooled 13 patients from phase 1 and 9 patients from phase 2 who entered the study with MDS. Three patients from the phase 1 study received a suboptimal, lower than recommended dose of olutasidenib (OLU) monotherapy (100-150 mg daily). The remaining 3 monotherapy patients received the full dose of OLU at 150 mg, twice daily. Seven patients from the phase 1 study received combination therapy of OLU + AZA (OLU + AZA). When combined with 9 patients from the phase 2 study, 16 total patients with MDS received OLU + AZA. The safety population or FAS comprised 22 patients, of whom 6 received some form of OLU monotherapy and 16 received OLU + AZA combination therapy. The response-evaluable population excluded 1 patient from the monotherapy group and 2 patients from the combination therapy group who did not have any response assessments during the study. The response-evaluable population consisted of 2 TN patients who received OLU monotherapy and 5 TN patients who received OLU + AZA for a total of 7 TN patients evaluated for response; and 3 R/R patients who received OLU alone and 9 R/R patients who received OLU + AZA for a total of 14 R/R patients evaluated for response. BID, twice daily; QD, daily.

Table 1.

Baseline demographics and disease characteristics (full analysis population)

Parameter	Monotherapy OLU∗ (n = 6)	Combination OLU + AZA (n = 16)	Pooled (N = 22)
Age, median (range), y	77 (66-87)	72 (59-82)	74 (59-87)
<65 y, n (%)	0	3 (19)	3 (14)
65 to <75 y, n (%)	2 (33)	7 (44)	9 (41)
≥75 y, n (%)	4 (67)	6 (38)	10 (45)
Sex, n (%)
Male	4 (67)	9 (56)	13 (59)
Female	2 (33)	7 (44)	9 (41)
Race, n (%)
White	6 (100)	13 (81)	19 (86)
Not specified/other	0	3 (19)	3 (14)
Region
North America	6 (100)	11 (69)	17 (77)
EU	0	2 (13)	2 (9)
Asia Pacific	0	3 (19)	3 (14)
MDS disease, n (%)
R/R	4 (67)	11 (69)	15 (68)
TN	2 (33)	5 (31)	7 (32)
MDS risk category per IPSS-R, n (%)
Intermediate, >3 to 4.5	0	3 (19)	3 (14)
High, >4.5 to 6	5 (83)	10 (63)	15 (68)
Very high, >6	1 (17)	3 (19)	4 (18)
ECOG performance status, n (%)
0	0	4 (25)	4 (18)
1	4 (67)	11 (69)	15 (68)
2	2 (33)	1 (6)	3 (14)
Bone marrow blast percentage, median (range)	6.5 (0-16)	7.5 (0-16)	7.5 (0-16)
Renal function, n (%)
Normal, ≥90 mL/min	2 (33)	3 (19)	5 (23)
Mildly impaired, 60-89 mL/min	2 (33)	9 (56)	11 (50)
Moderately impaired, 30-59 mL/min	2 (33)	4 (25)	6 (27)
Prior number of regimens, n (%)
0	2 (33)	5 (31)	7 (32)
1	3 (50)	5 (31)	8 (36)
2	0	5 (31)	5 (23)
≥3	1 (17)	1 (6)	2 (9)
Prior HMA, n (%)	4 (67)	10 (63)	14 (64)
Prior HSCT, n (%)	0	1 (6)	1 (5)
IDH1 mutation type
R132C	3 (50)	4 (25)	7 (32)
R132H	2 (33)	10 (63)	12 (55)
R132G	1 (17)	1 (6)	2 (9)
Unknown	0	1 (6)	1 (5)
Cytogenetic risk classification, n (%)
Good	1 (17)	9 (56)	10 (45)
Intermediate	1 (17)	3 (19)	4 (18)
Poor	0	2 (13)	2 (9)
Very poor	2 (33)	0	2 (9)
Unknown	2 (33)	2 (13)	4 (18)
Cytogenetic result, n (%)
Normal	1 (17)	8 (50)	9 (41)
Abnormal	3 (50)	6 (38)	9 (41)
Complex (≥3 abnormalities)	2 (33)	1 (6)	3 (14)
Unknown	2 (33)	2 (13)	4 (18)
No. of comutations, n (%)
None	1 (17)	0	1 (5)
1-3	4 (67)	10 (63)	14 (64)
4-7	1 (17)	3 (19)	4 (18)
Not done	0	3 (19)	3 (14)

ECOG, Eastern Cooperative Oncology Group; OLU, olutasidenib.

^∗During phase 1 dose escalation with OLU monotherapy, 3 patients received a suboptimal dose that was ≤50% of the approved dose of OLU. One patient received 100 mg daily, and 2 patients received 150 mg daily.

The study is complete (data cutoff, 15 June 2023). The median duration of follow-up was 53.8 months (95% CI, 12.3 to not reached [NR]) for the overall study. In the monotherapy and combination groups, the median follow-up was 60.3 months (95% CI, 11.8 to NR) and 53.8 months (95% CI, 12.3 to NR), respectively. At study end, 3 patients remained on treatment and transitioned to posttrial access or commercial supply, whereas 19 patients discontinued treatment. Primary reasons for discontinuation were progressive disease (PD; 6/22 [27%]), death (3/22 [14%]; 1 disease progression, 1 pulmonary fungal infection, and 1 pneumonia; all deemed not related to treatment), AEs (2/22 [9%]), physician’s decision (2/22 [9%]), patient withdrawal (2/22 [9%]; 1 of whom chose a different treatment option but continued with study follow-up), and transition to palliative care (1/22 [5%]). Additionally, 3 patients (14%) discontinued treatment to proceed to postremission HSCT transplant.

Safety

In the pooled safety analysis set (n = 22), comprising patients from both monotherapy and combination therapy cohorts, all patients (100%) experienced ≥1 treatment-emergent AE (TEAE), and 21 (95%) experienced a grade ≥3 TEAE (Table 2). The most frequently reported TEAEs of all grades in the monotherapy and combination cohorts included fatigue (67% and 63%), nausea (67% and 56%), and arthralgia (67% and 31%, respectively); additionally, constipation was common (56%) in the combination group. The most frequent grade ≥3 TEAEs with olutasidenib monotherapy were thrombocytopenia and neutropenia, each occurring in 2 of 6 patients (33%). In the combination therapy cohort, the most common grade ≥3 TEAEs were cytopenias: neutropenia in 6 of 16 (38%); febrile neutropenia in 5 of 16 (31%); and thrombocytopenia in 4 of 16 (25%). Overall, 16 of 22 patients (73%) experienced ≥1 treatment-related adverse events (TRAEs), with 15 (68%) experiencing a grade ≥3 TRAE (supplemental Table 2).

Table 2.

TEAEs of all grades (in 25% patients or more overall) and grade 3 or higher severity (in 10% patients or more overall)

TEAE	All grades, monotherapy OLU (n = 6), n (%)	All grades, combination OLU + AZA (n = 16), n (%)	Grade ≥3, monotherapy OLU (n = 6), n (%)	Grade ≥3, combination OLU + AZA (n = 16), n (%)
Patients with any TEAE	6 (100)	16 (100)	6 (100)	15 (94)
Fatigue	4 (67)	10 (63)	1 (17)	2 (13)
Nausea	4 (67)	9 (56)	0	1 (6)
Constipation	1 (17)	9 (56)	0	0
Arthralgia	4 (67)	5 (31)	1 (17)	0
Platelet count decreased	2 (33)	7 (44)	2 (33)	4 (25)
Vomiting	2 (33)	7 (44)	0	0
Neutrophil count decreased	2 (33)	6 (38)	2 (33)	6 (38)
Cough	1 (17)	6 (38)	0	0
Diarrhea	1 (17)	6 (38)	0	1 (6)
Dizziness	1 (17)	6 (38)	0	0
Dyspnea	2 (33)	5 (31)	0	1 (6)
Febrile neutropenia	1 (17)	5 (31)	1 (17)	5 (31)
Headache	2 (33)	4 (25)	0	0

OLU, olutasidenib.

Two patients (9%) had AEs as the primary reason for discontinuation of the study drug. One patient in the olutasidenib monotherapy group experienced a grade 3 alanine aminotransferase increase that resolved and was deemed unrelated to the study drug, and the other patient, in the combination therapy group, experienced elevated alanine aminotransferase (resolved), elevated bilirubin (resolved), and elevated gamma-glutamyl transferase (unresolved), all of which were grade 2 and deemed related to the study drug.

Differentiation syndrome occurred in 3 patients (14%) overall. One grade 1 event on day 34 in a patient from the combination cohort resolved after dose reduction and concomitant oral dexamethasone (10 mg daily); 1 grade 2 event on day 25 in a patient receiving monotherapy resolved after drug was held (drug resumed after patient recovered) and concomitant IV dexamethasone (10 mg every 12 hours [q12h]) was administered; and 1 grade 3 event on day 26 in a patient receiving combination therapy did not resolve due to patient death. This grade 3 differentiation syndrome event occurred in a patient who had previously discontinued treatment on day 23 due to co-occurring pneumonia, leading to respiratory failure and death (unrelated to treatment). The patient also received IV dexamethasone for concomitant leukocytosis (10 mg q12h) and IV methylprednisolone for pneumonia prophylaxis (60 mg once). Grade 3 QT prolongation occurred in 1 patient (5%) receiving combination therapy. No patient receiving monotherapy had QT prolongation. There were no treatment-related deaths on study.

Response

In the pooled full analysis population, the ORR was 59% (13/22), with a 27% CR (6/22) and 32% marrow CR rates (7/22; Table 3). The composite response rate (CR + marrow CR + HI in SD patients) was 68%. Among the 7 patients who achieved marrow CR, 4 (57%) achieved HI in ≥1 lineage (erythrocyte [HI-E], platelet [HI-P], and/or neutrophil [HI-N]) according to modified International Working Group 2006 criteria; 3 of 7 evaluable patients (43%) had HI-E, 1 of 4 (25%) had HI-P, and 3 of 6 (50%) had HI-N. Most patients achieved HI in 2 lineages, with 1 patient having HI-E + HI-P and 2 patients having HI-E + HI-N. An additional 2 patients whose best response was SD achieved HI (40%), with 1 HI-E + HI-N and 1 HI-N alone. Individual patient data on type and duration of treatment and response are depicted in Figure 2A. Three patients with a short duration on trial (<1-2.5 months) and no response assessment were excluded from the response-evaluable set (supplemental Table 3). In the pooled MDS response-evaluable population (n = 19; supplemental Table 3; Figure 2A), the ORR was 68% (13/19), with all responders experiencing either CR (6/19 [32%]) or marrow CR (7/19 [37%]), and the composite response rate (CR + marrow CR + HI in SD responders) was 79% (15/19).

Table 3.

Response rates and duration

FAS	N = 22		N = 22		Pooled (N = 22)
	Monotherapy OLU (n = 6)∗	Combination OLU+AZA (n = 16)	TN (n = 7)∗	R/R (n = 15)∗
Best overall response, n (%)
ORR (CR + marrow CR + PR)	2 (33)	11 (69)	6 (86)	7 (47)	13 (59)
CR	1 (17)	5 (31)	5 (71)	1 (7)	6 (27)
Marrow CR	1 (17)	6 (38)	1 (14)	6 (40)	7 (32)
Marrow CR with HI	1 (17)	3 (19)	1 (14)	3 (20)	4 (18)
PR	0	0	0	0	0
SD	2 (33)	3 (19)	1 (14)	4 (27)	5 (23)
SD with HI	1 (17)	1 (6)	1 (14)	1 (7)	2 (9)
PD	1 (17)	0	0	1 (7)	1 (5)
Not done	1 (17)	2 (13)	0	3 (20)	3 (14)
Composite response rate (CR + marrow CR + SD HI)	3 (50)	12 (75)	7 (100)	8 (53)	15 (68)
DOR, median (95% CI), mo
Duration of CR	NR (NR to NR)†	14.2 (4.9 to NR)	14.5 (4.9 to NR)	20.5‡	20.5 (4.9 to NR)
Duration of CR + marrow CR	NR (6.7 to NR)	14.6 (2.8 to NR)	23 (5.8 to NR)	14.6 (2.8 to NR)	14.6 (5.8-32.8)
Time to response, median (range), mo
Time to CR	8.3 (8.3-8.3)	5.1 (2.5-14.3)	5.1 (2.5-8.3)	14.3‡	5.7 (2.5-14.3)
Time to CR + marrow CR	4.7 (1-8.3)	2 (1-13)	2.2 (1-8.3)	2 (1-13)	2 (1-13)
OS, median (95% CI), mo
OS, total	14 (4.5 to NR)	27.5 (5-36.6)	NR (7.3 to NR)	16.3 (3.1-36.6)	27.2 (6.9-37)
CR responders	NR (NR to NR)†	26.9 (7.3 to NR)	26.9 (7.3 to NR)	37‡	37 (7.3 to NR)
CR + marrow CR	NR (14 to NR)	36.6 (7.3 to NR)	NR (7.3 to NR)	36.6 (5 to NR)	36.6 (14 to NR)
Nonresponders	NR (4.5 to NR)	3.1 (1 to NR)	NR (NR to NR)§	5.7 (1 to NR)	6.9 (1-30.8)
EFS, median (95% CI), mo
EFS, total	4.9 (1.9 to NR)	10.2 (3.7-25.8)	25.8 (7.3 to NR)	20.3 (3.7 to NR)	7.6 (3.7-25.8)
CR responders	NR (NR to NR)†	19.9 (7.3 to NR)	19.9 (7.3 to NR)	34.8‡	25.8 (7.3 to NR)
CR + marrow CR	NR (7.6 to NR)	20.3 (7.3 to NR)	25.8 (7.3 to NR)	20.3 (3.7 to NR)	20.3 (7.3-34.8)
Nonresponders	4.3 (1.9 to NR)	1.9 (1 to NR)	NR (NR to NR)§	2.8 (1 to 4.9)	3.7 (1-4.9)

AZA added to the regimen and achieved a CR.

^∗Three patients from the phase 1 dose-escalation portion of the study (all received monotherapy OLU) received a suboptimal dose that was ≤50% of the approved dose of OLU. One patient (R/R) received 100 mg daily, and 2 patients (1 TN and 1 R/R) received 150 mg daily.

^†There were no patients with an event, and 1 patient was censored at study end but continued to receive study drug/transitioned to commercial use.

^‡Range and 95% CI not applicable (n = 1)

^§The patient was categorized as SD while receiving suboptimal OLU monotherapy (150 mg, daily) but at investigator’s decision had AZA added to the regimen and achieved a CR.

Figure 2.

Response duration and transfusion requirements. (A) Per patient response, survival events, and duration are depicted in a swimmer plot. The patients are grouped by disease state and treatment. Reasons for no response assessment are as follows: 1 discontinued drug on day 32 due to physician’s decision to initiate a prohibited medication; 1 developed pneumonia (unrelated to study drug) on day 22, leading to respiratory failure and death; and 1 had an AE on day 45 and discontinued drug and never resumed or underwent any assessments before dropping from the study at 10 months. (B) RBC transfusion requirements at baseline (left) and after baseline (right). Eight of 13 patients (62%) requiring RBC transfusions at study entry achieved 56-day RBC TI; 3 of 13 (23%) remained RBC transfusion dependent; and 2 of 13 (15%) were not evaluated. Maintenance of RBC TI from baseline to postbaseline occurred in 8 of 9 patients (89%). (C) Platelet transfusion requirements at baseline (left) and postbaseline (right). Six of 9 patients (67%) requiring platelet transfusions at baseline achieved 56-day platelet TI; 1 of 9 (11%) remained platelet transfusion dependent; and 2 of 9 (22%) were not evaluated. Maintenance of platelet TI from baseline to postbaseline occurred in 11 of 13 patients (85%). Asterisk (∗) indicates received lower than optimal dose (≤50%) of OLU; ◊ indicates patient was started on suboptimal dose of monotherapy, then added AZA to regimen; HSCT, proceeded to HSCT after remission.

The monotherapy group (FAS/ITT, n = 6; TN, n = 2; R/R, n = 4) included 3 patients enrolled in the dose-escalation portion who received ≤50% of the approved dose of olutasidenib (150 mg per twice daily). By ITT, the ORR in the monotherapy group was 33% (2/6), and the composite response rate was 50% (3/6). Among the 3 patients who received monotherapy at the recommended olutasidenib dose, responses included CR (TN), marrow CR with HI (R/R), and no response assessment (R/R). The 3 patients who received a suboptimal olutasidenib dose had SD with HI, SD, and PD (see supplemental Table 3).

In the combination group (FAS), the ORR was higher in TN MDS (5/5 [100%]) than R/R MDS (6/11 [54.6%]). Of the 5 CR responses with combination therapy, 4 were in TN MDS, and 1 was in R/R MDS; of the 6 marrow CR responses with combination therapy, 3 patients had HI (TN MDS, 1/1; R/R MDS, 2/5; Table 3; Figure 2A). The combination therapy group included 3 patients who achieved remission (2 CR and 1 marrow CR) and subsequently proceeded to HSCT.

Kaplan-Meier analysis of the FAS population demonstrated a CR + marrow CR duration of 14.6 months (95% CI, 5.8-32.8), which was equivalent to the duration of overall response due to the absence of patients with PR (Table 3). The median duration of CR was 20.5 months (95% CI, 4.9 to NR). In the primary analysis, DOR was not censored for transplant; however, 2 of 3 patients who underwent postremission HSCT were censored at the time of study treatment discontinuation before transplant, because they had no further response assessment recorded. In a sensitivity analysis censoring for HSCT, the median DOR (ORR/CR + marrow CR) was 23 months (95% CI, 6.7-32.8). The median time to CR + marrow CR and to CR alone was 2 months (range, 1-13) and 5.7 months (range, 2.5-14.3), respectively.

TI

Of 13 patients dependent on RBC transfusions at baseline, 8 (62%) achieved 56-day TI, and 3 (23%) remained dependent on RBC transfusion (Figure 2B). Of 9 patients dependent on platelet transfusions at baseline, 6 (67%) achieved 56-day TI (Figure 2C). Two patients who were transfusion dependent for both RBCs and platelets at baseline were not evaluable due to discontinuing treatment before a postbaseline evaluation could occur.

Of the patients who were RBC TI at baseline, 8 of 9 (89%) maintained 56-day TI during the study (Figure 2B). Of the patients who were platelet TI at baseline, 11 of 13 (85%) maintained 56-day TI during the study (Figure 2C).

All responders (CR + marrow CR) except 1 patient (12/13 [92%]) were RBC TI after baseline (RBC dependent, n = 4; RBC independent at baseline, n = 9). The sole responder who did not achieve RBC TI after baseline had best response of marrow CR with HI-E + HI-N. All responders in the study (13/13 [100%]) achieved or maintained 56-day platelet TI after baseline (platelet dependent, n = 1; platelet independent at baseline, n = 12).

Survival

The median OS in the total population (n = 22) by Kaplan-Meier analysis was 27.2 months (95% CI, 6.0-37), with a 1-year survival probability of 68% (95% CI, 45-83; Table 3). The median OS was 27.5 months (95% CI, 5-36.6) for patients receiving combination therapy and 14 months (95% CI, 4.5 to NR) for monotherapy (Table 3). As expected, the median OS of responders (CR + marrow CR) was longer than nonresponders (36.6 months [95% CI, 14 to NR] vs 6.9 months [95% CI, 1-30.8]). The median OS in TN patients (n = 7) was NR (95% CI, 7.3 to NR; range 7.3-67.2), with 12- and 24-month survival probabilities both at 86%. R/R patients with MDS (n = 15) had a median OS of 16.3 months (95% CI, 3.1-36.6), with 12- and 24-month survival probabilities of 60% and 43%, respectively (Figure 3A; supplemental Table 4). In patients treated with the combination therapy, the median OS was 26.9 months (n = 5; 95% CI, 7.3 to NR) in TN MDS and 27.5 months (n = 11; 95% CI, 1.5-36.6) in R/R MDS (Figure 3B). In patients treated with monotherapy, the median OS was NR (n = 2; minimum, 60.3 months; maximum, 67.2 months) in TN MDS and 10.45 months (n = 4; 95% CI, 4.5 to NR) in R/R patients. The 2-year OS probability was 100% (95% CI, 100-100) in TN patients with MDS receiving monotherapy and 80% (95% CI, 20-97) in those receiving combination therapy.

Figure 3.

OS in MDS patients by disease state. (A) Overall TN patients vs overall R/R patients. The median OS in patients with TN MDS was NR (95% CI, 7.3 to NR) and represented 2 patients (29%) with an event and 5 (71%) who were censored. The median OS in patients with R/R MDS was 16.3 months (95% CI, 3.1-36.6) and represented 12 (80%) with an event and 3 (20%) who were censored. (B) Combination therapy group TN patients vs combination therapy R/R patients. The median OS in patients with TN MDS who received combination OLU + AZA was 26.9 months (95% CI, 7.3 to NR) and represented 2 (40%) with an event and 3 (60%) who were censored. The median OS in patients with R/R MDS who received combination OLU + AZA was 27.5 months (95% CI, 1.5-36.6) and represented 9 (82%) with an event and 2 (18%) who were censored.

EFS followed similar patterns as OS and are summarized in Table 3 and supplemental Table 4.

Correlative analysis

Responder analysis according to baseline MDS risk type (per IPSS-R definition) demonstrated alignment with the prognostic scoring (Figure 4A; supplemental Table 5). Among patients with a response assessment and known number of comutations at baseline (n = 17), the median number of comutated genes detected was 3 (range, 1-7). The patients who achieved CR (n = 6) had fewer comutations, with a median of 1.5 comutations (range, 1-3; Figure 4B), than patients with marrow CR (n = 5) and nonresponders (n = 6; 5 SD and 1 PD), both with a median of 3 comutations (range, 1-7). This numerical difference was not statistically significant, possibly due to relatively low numbers per response category or other confounding factors. For example, the 1 patient with PD had only 2 comutations, but one of which was TP53 associated with a complex karyotype, indicating an extremely poor prognosis.

Figure 4.

Baseline risk factors and response outcomes. (A) Oncoprint heat map depicting per patient baseline characteristics that may affect outcome, panel of baseline comutations, IPSS-R category, cytogenetics, and cytogenetic risk classification presented by best overall response. (B) Box and whisker plot of number of comutated genes by best overall response. Due to the small patient numbers, there was insufficient power to demonstrate a statistical difference between response categories. P > .05, for all comparisons including CR vs non-CR (ordinary 1-way analysis of variance) and responders vs nonresponders (unpaired t test). (C) IDH1 mutation type, based on local laboratory results. BOR, best overall response.

The most common comutated genes were ASXL1 (8/22 [36%]) and SRSF2 (8/22 [36%]), and the ORR by mutation is reported in supplemental Table 5. Three patients in the study had known comutations in receptor tyrosine kinase pathways; 1 had a FLT3 mutation and achieved CR on combination therapy; and 2 had NRAS mutations, including 1 who achieved marrow CR on combination therapy and 1 who had SD on monotherapy and later achieved CR when AZA was added to olutasidenib (allowed in the protocol; n = 1). This patient was only included in the TN monotherapy cohort for both response and survival analysis. There were 3 patients classified as MDS under the WHO 2008 criteria who would have been diagnosed with AML by WHO 2022 guidelines. They included 2 patients with NPM1 mutations and 1 patient with peripheral blasts >20%. Supplemental Table 6 summarizes the best overall response in patients who would meet WHO 2022 diagnostic criteria for MDS.

The most common IDH1 mutation type among patients in this analysis was R132H (12/22 [55%]), followed by R132C (7/22 [32%]; Figure 4C). Mutation types were evenly distributed across response categories (Figure 4A).

Discussion

Treatment of intermediate- to very high-risk TN and R/R mIDH1 MDS with olutasidenib with or without AZA demonstrated an expected and tolerable safety profile with meaningful clinical activity, consistent with prior reports of olutasidenib with or without AZA in patients with AML.^29,31 Currently, there are very limited data regarding the treatment of TN or R/R MDS with an mIDH1 inhibitor plus HMA combination. Therefore, we explored olutasidenib plus AZA and olutasidenib monotherapy in an exclusively higher-risk TN and R/R MDS population.

This is, to our knowledge, the first full report of safety and clinical activity of an mIDH1 inhibitor plus HMA combination in TN and R/R patients with MDS. There have been 2 prior reports of ivosidenib (an mIDH1 inhibitor) monotherapy in R/R MDS: the MDS substudy from the original phase 1b AML trial, reporting a median OS of 35.7 months in R/R MDS (n = 18)³²; and the phase 2 IDIOME trial, which included patients with both lower- and higher-risk TN MDS and higher-risk R/R MDS and reported a median OS of 8.9 months in the R/R MDS cohort (n = 22).³³ The phase 1 ivosidenib study was conducted in a R/R MDS population with 63% low to intermediate risk, 32% high to very high risk, and 5% unknown, whereas IDIOME included only patients with higher-risk R/R MDS, which may explain the difference in the median OS between these 2 studies.

Our study was conducted in a strictly higher-risk MDS population, with no low-risk patients enrolled, and 86% (19/22) had high/very high risk. The median OS in all patients was 27.2 months, and the median OS in all patients treated with combination therapy was 27.5 months (n = 16). Not surprisingly, the median OS in TN patients with MDS treated with olutasidenib (±AZA) was NR (n = 7). In R/R patients (n = 15; 93% failed prior HMA) treated with olutasidenib ± AZA, the median OS was 16.3 months (95% CI, 3.1-36.6). Survival estimates for R/R MDS are generally poor, with 1 study reporting a median OS of 5.6 months in high-risk MDS after HMA failure.⁷ In this pooled analysis of 22 patients, olutasidenib alone or in combination with AZA improved survival in both TN and R/R MDS compared to historical benchmarks. TN patients had particularly long OS with olutasidenib monotherapy or in combination with AZA, as did R/R patients receiving combination therapy, suggesting that not only can olutasidenib salvage HMA failure but continuing an HMA in combination with olutasidenib may benefit higher-risk R/R patients with MDS.⁷

Olutasidenib with or without AZA also demonstrated encouraging response rates. All response-evaluable patients treated with monotherapy at the recommended phase 2 dose of olutasidenib were responders, and response rates were higher with combination therapy. Olutasidenib 150 mg twice daily was affirmed as the clinically effective dose with or without AZA in higher-risk MDS. In this primarily high- to very high-risk MDS population, 3 patients would have been reclassified as AML by current WHO 2022 criteria. Among 14 response-evaluable patients classified as MDS under 2022 criteria who were treated with olutasidenib 150 mg twice daily ± AZA, all but 1 patient achieved a response (ORR, 13/14 [93%]; CR, 6/14 [43%]; marrow CR, 7/14 [50%]). The 1 patient with MDS with PD harbored a TP53 comutation associated with a complex karyotype (likely MDS-biallelic TP53 per WHO 2022 criteria), indicating an extremely poor prognosis that is functionally equivalent, if not more adverse, to having a high/very high comutational burden.³⁴

We demonstrated high rates of conversion from transfusion dependence to TI, as well as maintenance of TI, with olutasidenib. Frequent and long-term transfusion dependence is a significant burden in MDS, associated with risk of complications (eg, transfusion-related acute lung injury, infection, and hemolytic transfusion reactions), and represents an economic and health-related quality of life detriment to patients.³⁵ Furthermore, iron overload secondary to chronic transfusion therapy is associated with reduced OS in MDS.³⁶ Therefore, reducing or eliminating transfusion dependence is an impactful clinical outcome.

Cytopenias were among the most commonly reported TEAEs; however, this was likely due to increased risk of hematologic toxicities associated with AZA, an HMA.³⁷ In this study, differentiation syndrome was reported in 3 patients (14%), with only 1 occurrence (5%) of grade 3 severity. The low rate of differentiation syndrome in MDS is consistent with previous reports for commercially available IDH inhibitors.^11,29,31,38, 39, 40 No treatment-related deaths occurred. Two patients (9%) discontinued the study drug for the primary reason of an AE, and 1 patient (5%) discontinued due to a TRAE.

In the age of precision medicine and with the availability of several tools to predict prognosis and risk in MDS, we examined patient characteristics for response patterns. The rate of CR aligned with a lower baseline IPSS-R MDS risk score and lower comutational burden. However, several patients with very high risk and poor cytogenetic risk classification were able to achieve marrow CR and/or HI with the combination of olutasidenib plus AZA. Higher rates of CR and greater probability of 24-month OS (86% vs 43%) were observed in TN than R/R MDS. Combination therapy may offer shorter time to CR + marrow CR than monotherapy (2 months vs 4.7 months). Given the impressive outcomes with combination therapy in TN MDS and that patients with prior therapy were nearly all R/R to HMAs, our results suggest that adding olutasidenib to AZA earlier in the MDS disease course may improve survival in patients with mIDH1. Olutasidenib was recently added to the National Comprehensive Cancer Network Guidelines Myelodysplastic Syndromes recommendations as monotherapy for patients with R/R mIDH1 MDS in certain settings, such as higher-risk (MDS-7) patients ineligible for transplant with no prior IDH1 inhibitor exposure (National Comprehensive Cancer Network category 2B).⁴¹ In addition, olutasidenib in combination with AZA was recommended as a frontline treatment option for higher-risk MDS with mIDH1 regardless of transplant eligibility (category 2B).⁴¹

Study limitations include that this was an open-label phase 1 of 2 trial with multiple MDS subsets and the relatively small sample size, which could affect statistical power and hinder detection of rare events. However, these challenges were inherently expected given the rarity of mIDH1 in MDS. Additionally, this study included patients from the MDS dose-escalation cohort who received suboptimal doses of olutasidenib (at or below the minimum effective dose), which may have affected clinical activity in the monotherapy group.

In conclusion, olutasidenib, alone or in combination with AZA, induced a high rate of clinically meaningful CR and HI responses with durable remission duration, accompanied by low rates of serious TEAEs in patients with intermediate-, high-, or very high-risk MDS harboring mIDH1. With limited treatment options available for MDS and even fewer effective treatment options for patients with high- to very high-risk MDS who have failed a prior HMA, these results are highly encouraging.⁴² Taken together, there is a strong rationale for continued evaluation of olutasidenib in patients with both TN and R/R MDS harboring mIDH1. A new phase 2 clinical trial investigating olutasidenib combined with an HMA in patients with TN or R/R mIDH1 higher-risk MDS (as well as chronic myelomonocytic leukemia or advanced myeloproliferative neoplasm) is currently underway (ClinicalTrials.gov identifier, NCT06597734).

Conflict-of-interest disclosure: J.E.C. reports research funding for his current or former institution from and is a consultant to Astellas, Amphivena, Bristol Myers Squibb (BMS), Novartis, Pfizer, Takeda, Daiichi-Sankyo, Jazz Pharmaceuticals, Merus, and Forma Therapeutics; and is a consultant to Rigel, BiolineRx, Bioptah, Sun Pharma, Syndax, and Tern Pharma. J.Y. reports research funding from Janssen, Pfizer, Forma, and PureTech Health. G.J.R. has served as a consultant for Agios, Amgen, Amphivena, Astex, Celator, Celgene, Clovis Oncology, CTI BioPharma, Genoptix, Immune Pharmaceuticals, Janssen Pharmaceuticals, Juno, MedImmune, MEI Pharma, Onconova, Pfizer, Roche, and Sunesis; reports research funding from AbbVie, BMS, Teva, and Karyopharm; is an advisory board member or consultant for Novartis, AbbVie, BeiGene, BerGenBio, Arcellx, Jazz Pharmaceuticals, Syros, BMS, Genentech, ImmunoGen, AstraZeneca, Kura, Ryvu, Magenta, Qihan Biotech, and Zentalis Pharmaceuticals; and has provided research support to Janssen. S.N.D. reports research funding from Pfizer, Rigel, BMS, Novartis, and Kite/Gilead. E.S.W. reports consultancy fees from AbbVie, Astellas, BMS, Gilead, GSK, Jazz, Kite, Novartis, Pfizer, PharmaEssentia, and Takeda; and speakers’ bureau fees from Astellas, Kite, Novartis, Pfizer, Dava Oncology, and Kura Oncology. A.H.W. reports honoraria, research funding, consultancy fees, speakers’ bureau fees, and board/advisory committee membership fees from Astellas; honoraria, research funding, speakers’ bureau fees, and board/advisory committee membership fees from AbbVie/Genentech, Amgen, Celgene/BMS, and Novartis; research funding, honoraria, consultancy fees, and board/advisory committee membership fees from Servier and Syndax; honoraria, consultancy fees, and board/advisory committee membership fees from Janssen and Gilead; honoraria and board/advisory committee membership fees from MacroGenics and Pfizer; honoraria, research funding, and board/advisory committee membership fees from AstraZeneca; research funding from Astex; is an employee of the Walter and Eliza Hall Institute; and is eligible for a fraction of the royalty stream related to venetoclax. H.T. and F.d.T. are employees of and hold stock in Rigel Pharmaceuticals, Inc. M.R.B. reports institutional research funding from AbbVie, Ascentage, Astellas, Curadev, Forma, and Takeda. J.M.W. reports research funding and board/advisory committee membership fees from Takeda and Rigel; research funding from Immune System Key, Ltd; and is a consultant or board/advisory committee member for Genentech, Rafael Pharma, Reven Pharma, Celgene/BMS, Servier, Aptose, Astellas, and Daiichi-Sankyo. W.D. declares no competing financial interests.