Glasdegib with intensive/nonintensive chemotherapy in Japanese patients with untreated acute myeloid leukemia or high‐risk myelodysplastic syndromes

Abstract

Glasdegib is a potent, selective, oral inhibitor of the hedgehog signaling pathway. In this phase I study, previously untreated Japanese patients with acute myeloid leukemia (AML) or high‐risk myelodysplastic syndromes were treated with glasdegib (100 mg once daily) combinations: low‐dose cytarabine (20 mg twice daily; cohort 1, n = 6; expansion cohort, n = 15); daunorubicin and cytarabine (60 mg/m² i.v.; cohort 2, n = 6); or azacitidine (100 mg/m² i.v.; cohort 3, n = 6). Patients, except cohort 2, were ineligible for intensive chemotherapy. The primary end‐point was dose‐limiting toxicity in cohorts 1–3 and disease‐modifying response in the expansion cohort. Disease‐modifying response rate was tested with the null hypothesis of 6.8%, which was set based on the results from the phase II BRIGHT AML 1003 study (NCT01546038). No dose‐limiting toxicities were observed in cohorts 1 or 3; one patient in cohort 2 experienced a dose‐limiting toxicity of grade 3 erythroderma. The most common grade ≥3 treatment‐related adverse events were neutropenia and thrombocytopenia (66.7% each) in cohort 1 and thrombocytopenia (60.0%) in the expansion cohort. In the expansion cohort, the disease‐modifying response rate was 46.7% (90% confidence interval, 24.4–70.0; p < 0.0001), with all patients achieving either a complete response or complete response with incomplete blood count recovery. Median overall survival was 13.9 months. In this study, the primary disease‐modifying response end‐point with glasdegib plus low‐dose cytarabine was met. The study confirms the safety and efficacy of glasdegib plus low‐dose cytarabine in Japanese patients with AML ineligible for intensive chemotherapy.

Glasdegib is a potent, selective, oral inhibitor of the hedgehog signaling pathway. This phase I study confirms the safety and efficacy of glasdegib plus low‐dose cytarabine in Japanese patients with acute myeloid leukemia ineligible for intensive chemotherapy.

Abbreviations

AE

adverse event

AML

acute myeloid leukemia

AUC

area under the plasma concentration–time curve

AUC_tau

AUC from time zero to time tau, the dosing intervals, where tau = 24 h

CI

confidence interval

C _max

maximum plasma concentration

CR

complete remission

CRi

CR with incomplete blood count recovery

CYP

cytochrome P450

DLT

dose‐limiting toxicity

DMR

disease‐modifying response

Hh

hedgehog

IPSS

International Prognostic Scoring System

LDAC

low‐dose cytarabine

MDS

myelodysplastic syndromes

NE

not evaluable

OS

overall survival

PD

pharmacodynamics

PK

pharmacokinetics

SAE

serious adverse event

TEAE

treatment‐emergent adverse event

1. INTRODUCTION

Acute myeloid leukemia and MDS comprise a group of heterogeneous myeloid neoplasms that primarily affect older adults. ¹ Due to the presence of comorbidities and unfavorable disease‐related factors, older patients are often ineligible for treatment with intensive induction chemotherapy and are instead treated with less intensive chemotherapy, such as LDAC or a hypomethylating agent (e.g., azacitidine or decitabine), with poor outcomes. ¹ , ² More recently, low‐intensity chemotherapy combinations with the Hh inhibitor glasdegib or the B‐cell leukemia/lymphoma‐2 inhibitor venetoclax have emerged as treatment options in patients ineligible for intensive chemotherapy. ³ , ⁴ , ⁵ , ⁶

Aberrant activation of the Hh pathway has been mechanistically linked to the development and progression of various malignancies, including leukemias and expansion of leukemic stem cells. ⁷ , ⁸ Overexpression of the pathway has been associated with a chemoresistance phenotype in myeloid leukemia cells, which can be reverted through inhibition of Hh pathway components. ⁹ , ¹⁰ Glasdegib is an oral inhibitor of the Hh pathway component smoothened, which has been shown to attenuate the leukemia initiation potential of AML cells and sensitize cells to cytarabine in preclinical models. ¹¹ , ¹² In clinical studies, treatment with single‐agent glasdegib provided preliminary evidence of efficacy in patients with hematologic malignancies, with comparable results in Japanese and non‐Japanese patients. ¹³ , ¹⁴ The clinical activity of single‐agent glasdegib, together with preclinical evidence of potential synergy with chemotherapy, provided rationale for evaluating glasdegib in combination with chemotherapy in clinical studies.

Glasdegib in combination with LDAC, decitabine, azacitidine, or standard cytarabine/daunorubicin induction has shown clinical benefit and acceptable tolerability in non‐Japanese patients with newly diagnosed AML or high‐risk MDS in phase Ib/II clinical trials. ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ In the phase II BRIGHT AML 1003 trial, glasdegib in combination with LDAC significantly improved OS compared with LDAC alone (median OS, 8.8 months vs. 4.9 months, respectively; hazard ratio, 0.51; 80% CI, 0.39–0.67; p = 0.0004) in non‐Japanese patients with newly diagnosed AML or high‐risk MDS ineligible for intensive chemotherapy. ¹⁵ Based on these results, glasdegib was approved in Europe and the United States in combination with LDAC for the treatment of patients with newly diagnosed AML who are unable to receive intensive chemotherapy due to comorbidities or older age (≥75 years). ³ , ⁴

We aimed to evaluate the safety, PK, PD, and efficacy of glasdegib in combination with LDAC as well as cytarabine/daunorubicin or azacitidine in Japanese patients with AML or high‐risk MDS.

2. MATERIALS AND METHODS

2.1. Study design

This was an open‐label, multicenter, phase I study (ClinicalTrials.gov, NCT02038777) of glasdegib in previously untreated Japanese patients with AML or high‐risk MDS (Table 1). First, the safety of glasdegib was evaluated in combination with LDAC (cohort 1) or cytarabine and daunorubicin (cohort 2). The safety of glasdegib was then evaluated in combination with azacitidine (cohort 3, AML only). After that, the efficacy of glasdegib plus LDAC was evaluated in the expansion cohort.

TABLE 1.

Study design and disposition of Japanese patients with acute myeloid leukemia (AML) or high‐risk myelodysplastic syndromes (MDS)

Treatment regimen	Cohort 1	Cohort 2	Cohort 3	Expansion cohort
	Glasdegib + LDAC	Glasdegib + cytarabine and daunorubicin	Glasdegib + azacitidine	Glasdegib + LDAC
Eligibility for intensive chemotherapy	Ineligible	Eligible	Ineligible	Ineligible
Primary end‐point	Cycle 1 DLTs			DMR rate
Patients enrolled and treated, n	6	6	6	15
AML	4	6	6	15
MDS	2	0	0	0

Abbreviations: DLT, dose‐limiting toxicity; DMR, disease‐modifying response; LDAC, low‐dose cytarabine.

The primary end‐points of cohorts 1–3 were first‐cycle DLTs, AEs, and vital signs and laboratory abnormalities. Secondary end‐points were PK, PD, and objective disease response of each combination. Additional secondary end‐points included OS (cohorts 1 and 3), duration of response, and time to response (cohort 3).

The primary end‐point of the expansion cohort was DMR rate, which includes CR, CRi, morphologic leukemia‐free state, marrow CR, and partial remission. Secondary end‐points were AEs, vital signs and laboratory abnormalities, OS, objective disease response, CR rate, duration of response, time to response, PK, and PD.

The study was carried out in compliance with the Declaration of Helsinki and followed the International Council for Harmonization Good Clinical Practice guidelines. The protocol was approved by the institutional review boards of the participating institutions, and all patients provided signed informed consent.

2.2. Patients

Patients eligible for the LDAC combination (cohort 1 and expansion cohort) were: age ≥55 years with newly diagnosed, previously untreated AML (including de novo AML and secondary AML [AML evolving from MDS or an antecedent hematologic disorder, or AML after previous cytotoxic therapy or radiation]) or refractory anemia with excess blast‐2 (high‐risk MDS) according to the WHO 2008 classification ¹⁹ ; considered ineligible for intensive chemotherapy, with ECOG performance status ≤2; and with adequate organ function. Patients with one or more of the following criteria were considered ineligible for intensive chemotherapy: age ≥75 years, ECOG performance status of 2, serum creatinine >1.3 mg/dL, or severe cardiac disease (e.g., left ventricular ejection fraction <45% by multigated acquisition or echocardiography at screening). Patients were allowed one prior regimen with commercially available agents for prior hematologic disease but no prior therapy for their AML. Patients were excluded from the study if they had acute promyelocytic leukemia, t(9;22) translocation, hyperleukocytosis at study entry, active and uncontrolled leukemia in the central nervous system, or cardiac disease in the previous 6 months.

Eligible patients in cohort 2 were age ≥20 years and considered eligible for intensive chemotherapy. In cohort 3, patients were age ≥20 years with AML only and considered ineligible for intensive chemotherapy per investigator's judgment. Other eligibility criteria were the same as for the LDAC combination.

2.3. Treatment

Glasdegib was given orally, q.d., continuously, at a starting dose of 100 mg from day 1 of each cycle (unless otherwise stated), in 28‐day cycles. In cohort 1 and the expansion cohort, glasdegib was given in combination with LDAC 20 mg s.c. b.i.d. on days 1–10 of each 28‐day cycle. In cohort 1, glasdegib dosing started on day 3 of cycle 1. Treatment continued for up to 12 cycles (cohort 1 only) or until disease progression or relapse, patient refusal, or unacceptable toxicity.

In cohort 2, glasdegib was started on day −3 of cycle 1 in combination with daunorubicin 60 mg/m² i.v. on days 1–3 and cytarabine 100 mg/m² i.v. on days 1–7 of induction cycle. Consolidation for patients who achieved a CR/CRi consisted of glasdegib in combination with cytarabine 1 g/m² i.v. bid on days 1, 3, and 5 for two to four cycles unless the patient relapsed, unacceptable toxicity, or patient withdrawal. Patients who had completed the induction phase and two to four cycles of consolidation and maintained a CR/CRi were eligible for glasdegib maintenance therapy (monotherapy q.d. for a maximum of six cycles). Treatment continued until disease progression or relapse, patient refusal, or unacceptable toxicity.

In cohort 3, glasdegib was started on day 2 of cycle 1 and combined with azacitidine 75 mg/m² i.v. or s.c. on days 1–7 of each 28‐day cycle. Treatment was continued for at least six cycles or until disease progression or relapse, death, unacceptable toxicity, or patient refusal.

2.4. Assessments

2.4.1. Safety

Adverse events were graded based on the NCI Common Terminology Criteria for Adverse Events version 4.0.

Dose‐limiting toxicities were evaluated in cohorts 1–3 during the first cycle of treatment (days 1–28 in cohorts 1 and 3; day −3 to days 21–28 in cohort 2). The following AEs, when considered possibly related to glasdegib combination therapy, met the definition of a DLT: grade ≥3 nonhematologic toxicity excluding infection, fever, infusion‐related AEs, electrolyte abnormalities, and alanine aminotransferase/aspartate aminotransferase elevation returning to grade ≤1 or baseline within 7 days; grade ≥3 corrected QT prolongation (corrected QT interval ≥501 ms); and prolonged (≥42 days) myelosuppression (absolute neutrophil count <500/μL or platelet count <10 × 10⁹/L with normal bone marrow). Patients were not evaluable for DLT if they had not received at least 80% of the planned study doses for all agents due to nonhematologic toxicities or if they had experienced a delay of more than 28 days in receiving the next scheduled cycle due to persisting nonhematologic toxicities.

2.4.2. Efficacy

Response to treatment was assessed based on the International Working Group response criteria and WHO guidelines for AML and MDS. ²⁰ , ²¹

Cohort 1 included follow‐up for survival. In the expansion cohort and cohort 3, patients were followed for survival every 8 weeks for up to 2 years (expansion) or 12 weeks for up to 3 years (cohort 3) or until death, end of study, or patient withdrawal.

2.4.3. Pharmacokinetics

Serial blood samples were collected for PK assessments of glasdegib and its combination partners at protocol‐defined time points throughout treatment. Standard PK parameters including the observed C _max, time to C _max, and AUC were estimated for each drug using noncompartmental analysis.

2.4.4. Pharmacodynamics

RNA sequencing was undertaken centrally using bone marrow aspirate samples collected at screening from patients who were evaluable for response. Variant analysis of the whole transcriptome was carried out by ACE ImmunoID NeXT (Personalis, Inc.) tumor alone configuration. Transcripts per million values were log₂‐transformed for further analysis of individual genes or standardized gene pathway signature scores. For each gene, the mean expression and SD were calculated across samples and the mean subtracted and divided by the SD to standardize the gene score, which was then centered to zero with units of SD (z‐score). The pathway score for each sample was calculated as the average of the standardized values for the set of genes within the pathway. For longitudinal analysis, a linear mixed model with fixed effects and random effects was used to test differences in longitudinal change of a variable between two groups. ²² The fixed effects capture the overall population‐level effects of the predictor variables (group membership and time) on the outcome variable (gene expression). The random effects account for individual variability in the intercepts, allowing each subject to have their own baseline level. The p values were calculated from interaction terms between the group membership variable and the time variable in the fixed‐effects part of the model, which determines whether the rate of change in the outcome variable over time is significantly different between the two groups. The Molecular Signatures Database (MSigDB) gene sets were used for pathway level gene expression analysis. ²³

Two additional analyses were undertaken for the Hh signaling pathway signature. For the first analysis, a linear mixed model using compound symmetry for the covariance pattern and treating Hh signaling at the screening visit as a covariate and an on‐treatment visit as a categorical variable was carried out. Given that only one non‐DMR patient had data for the last two visits (day 1 in cycles 6 and 9), the second analysis only included the visit on day 1 in cycle 3 as the outcome measure. A linear model that treated the screening visit as a covariate was undertaken for the second analysis. Both analyses showed similar results—that is, after controlling for baseline difference, the DMR patients showed significant Hh signaling pathway modulation by glasdegib at the initial on‐treatment visits in comparison to the non‐DMR patients.

2.5. Statistical analysis

No statistical sample size determination was performed for cohorts 1–3. The glasdegib combination was considered tolerable if one or fewer of the six patients experienced a DLT by the end of cycle 1. In the expansion cohort, a sample size of 15 patients was estimated to give 80% power with a one‐sided type I error rate of 0.05 to test the null hypothesis that the true investigator‐reported DMR rate was 6.8% versus the alternative hypothesis of 34.1%, with four responses required for a positive study. The null DMR rate and alternative DMR rate were set based on the results from the phase II BRIGHT AML 1003 trial (data cut‐off: January 3, 2017). The DMR rate was tested using the exact test for a single proportion (one‐sided significance level, 0.05).

Safety, PK, PD, and clinical activity were summarized with descriptive statistics. Overall survival, defined as the time from the first dose to death of any cause, was estimated using the Kaplan–Meier method.

This analysis was based on the April 13, 2021, database lock.

3. RESULTS

3.1. Patients

A total of 33 patients were enrolled in the study and treated with glasdegib in combination with LDAC (cohort 1, n = 6; expansion cohort, n = 15), cytarabine and daunorubicin (cohort 2, n = 6), or azacitidine (cohort 3, n = 6) (Table 1). Patient demographics and clinical characteristics of cohort 1 and the expansion cohort are presented in Table 2 and those of cohorts 2 and 3 in Table S1. All patients had AML, except for two patients with MDS in cohort 1.

TABLE 2.

Baseline demographics and clinical characteristics for patients with acute myeloid leukemia (AML) or high‐risk myelodysplastic syndromes (MDS) treated with glasdegib plus low‐dose cytarabine (LDAC) (cohort 1 and expansion cohort)

	Cohort 1	Expansion cohort
	Glasdegib + LDAC (n = 6)	Glasdegib + LDAC (n = 15)
Age, years
≥65, n (%)	5 (83.3)	15 (100)
Median (range)	71.5 (59–81)	76.0 (68–87)
Sex, n (%)
Male	5 (83.3)	8 (53.3)
Female	1 (16.7)	7 (46.7)
ECOG performance status, n (%)
0	1 (16.7)	5 (33.3)
1	2 (33.3)	6 (40.0)
2	3 (50.0)	4 (26.7)
AML	n = 4	n = 15
Disease history, n
De novo	1	6
Secondary	3	9
AML risk category, 34 n (%)
Intermediate‐II	2 (50.0)	12 (80.0)
Adverse	1 (25.0)	3 (20.0)
Not evaluated	1 (25.0)	0 (0.0)
MDS	n = 2	n = 0
Disease history, n
De novo	2	NA
Risk category, a n (%)
Poor risk	2 (100)	NA
IPSS score, 35 n (%)
1.5–2 (intermediate‐II)	2 (100)	NA
Karyotype, n (%)
Normal	1 (16.7)	5 (33.3)
Abnormal	5 b (83.3)	10 c (66.7)
RUNX1‐RUNX1T1	0	–
INV(16)(P13.1Q22)	0	–
t(16;16)(P13.1;Q22); CBFB‐MYH11	0	–
t(9;22)(Q34;Q11)	0	–
Other	5	–
Not classified as favorable or adverse	–	7 d
Complex karyotype/monosomal karyotype	–	3 d
INV(3)(Q21.3;Q26.2); GATA2, MECOM(EV11) DEL(5Q); −7; −17/ABN(17P)	–	1

Abbreviations: IPSS, International Prognostic Scoring System; NA, not applicable.

^a MDS risk was assessed by cytogenetic abnormalities that were known at the time the study was initiated; poor cytogenetic risk = poor risk group.

^b Included RUNX1‐RUNX1T1, INV(16)(P13.1Q22), t(16;16)(P13.1;Q22); CBFB‐MYH11, t(9;22)(Q34;Q11) (per case report form completion requirement). Five patients had “other” abnormal characteristics.

^c Included not classified as favorable or adverse, complex karyotype/monosomal karyotype, and INV(3)(Q21.3;Q26.2); GATA2, MECOM(EV11) DEL(5Q); −7; −17/ABN(17P). ^d

^d One patient had chromosomal abnormalities classified as ‘cytogenetics not classified as favorable or adverse’ and ‘complex/monosomal karyotype’.

As of the data cut‐off, two patients in the expansion cohort were still receiving treatment with glasdegib plus LDAC.

3.2. Safety and tolerability

3.2.1. Cohort 1 and the expansion cohort

The median duration of treatment with glasdegib plus LDAC was 51 days (range, 29–486 days) in cohort 1 and 288 days (range, 34–736 days) in the expansion cohort. The mean relative dose intensity for glasdegib and LDAC, respectively, was 88.4% and 98.7% in cohort 1, and 96.3% and 96.1% in the expansion cohort.

No DLTs were observed with glasdegib plus LDAC in cohort 1. All patients treated with glasdegib plus LDAC in both cohorts experienced at least one all‐causality TEAE (Table S2). The TEAEs that were considered treatment‐related were experienced by all patients in cohort 1 and 93.3% (n = 14) in the expansion cohort (Table 3). In cohort 1, the most common treatment‐related TEAEs were dysgeusia, neutropenia, and thrombocytopenia (66.7%, n = 4 each), and anemia, febrile neutropenia, and leukopenia (50%, n = 3 each); in the expansion cohort, these were thrombocytopenia, anemia, and nausea (60%, n = 9 each), and leukopenia (53.3%, n = 8). Grade ≥3 treatment‐related TEAEs occurred in 83.3% (n = 5) of patients in cohort 1 and 86.7% (n = 13) of patients in the expansion cohort. The most frequent were neutropenia and thrombocytopenia (66.7%, n = 4 each) in cohort 1 and thrombocytopenia (60.0%, n = 9), anemia, and leukopenia (53.3%, n = 8 each) in the expansion cohort. One patient in cohort 1 and three patients in the expansion cohort experienced a grade 5 TEAE, none of which were considered treatment‐related. In cohort 1, one patient experienced an SAE (leukemia progression), which was not related to study treatment. In the expansion cohort, nine patients experienced 13 SAEs, including sepsis, compression fracture, pneumonia, leukemia progression, cellulitis, recurrent lung adenocarcinoma, secondary primary malignancy, bacteremia, Clostridium difficile colitis, and febrile neutropenia. All except cellulitis were considered unrelated to study treatment. No patient in cohort 1 permanently discontinued study treatments due to TEAEs, while two patients in the expansion cohort permanently discontinued treatment due to TEAEs (grade 4 pneumonia and grade 3 recurrent lung adenocarcinoma), neither of which was considered related to treatment.

TABLE 3.

Treatment‐emergent, treatment‐related adverse events (AEs) occurring in ≥20% of patients with acute myeloid leukemia or high‐risk myelodysplastic syndromes treated with glasdegib plus low‐dose cytarabine (cohort 1 and expansion cohort)

n (%)	Cohort 1		Expansion cohort
	Glasdegib + LDAC (n = 6)		Glasdegib + LDAC (n = 15)
	All grades	Grade ≥3	All grades	Grade ≥3
Any AE	6 (100)	5 (83.3)	14 (93.3)	13 (86.7)
Dysgeusia	4 (66.7)	0 (0.0)	6 (40.0)	0 (0.0)
Neutropenia a	4 (66.7)	4 (66.7)	6 (40.0)	6 (40.0)
Thrombocytopenia b	4 (66.7)	4 (66.7)	9 (60.0)	9 (60.0)
Anemia	3 (50.0)	3 (50.0)	9 (60.0)	8 (53.3)
Febrile neutropenia	3 (50.0)	3 (50.0)	7 (46.7)	7 (46.7)
Leukopenia c	3 (50.0)	3 (50.0)	8 (53.3)	8 (53.3)
Constipation	2 (33.3)	0 (0.0)	1 (6.7)	0 (0.0)
Malaise	2 (33.3)	0 (0.0)	2 (13.3)	0 (0.0)
Nausea	2 (33.3)	0 (0.0)	9 (60.0)	0 (0.0)
Pyrexia	2 (33.3)	0 (0.0)	6 (40.0)	0 (0.0)
Weight decreased	2 (33.3)	0 (0.0)	3 (20.0)	2 (13.3)
ALT increased	1 (16.7)	0 (0.0)	3 (20.0)	0 (0.0)
Decreased appetite	1 (16.7)	0 (0.0)	7 (46.7)	1 (6.7)
Alopecia	0 (0.0)	0 (0.0)	3 (20.0)	0 (0.0)
Injection site reaction	0 (0.0)	0 (0.0)	3 (20.0)	0 (0.0)
Vomiting	0 (0.0)	0 (0.0)	3 (20.0)	0 (0.0)

Abbreviations: ALT, alanine aminotransferase.

^a Includes AEs reported as neutropenia and neutrophil count decreased.

^b Includes AEs reported as thrombocytopenia and platelet count decreased.

^c Includes AEs reported as leukopenia and white blood cell count decreased.

3.2.2. Cohorts 2 and 3

As of data cut‐off, the median duration of treatment with glasdegib in combination with cytarabine and daunorubicin in cohort 2 was 75.5 days (range, 32–343 days) and 286.5 days (range, 23–841 days) in combination with azacitidine in cohort 3.

One patient receiving glasdegib plus cytarabine and daunorubicin in cohort 2 experienced a DLT of grade 3 erythroderma, which resolved with treatment discontinuation. No DLTs were observed with glasdegib plus azacitidine in cohort 3. The most common treatment‐related TEAEs with glasdegib plus cytarabine/daunorubicin (cohort 2) or azacitidine (cohort 3) are presented in Table S3. In cohort 2, three patients experienced treatment‐related SAEs, including cytomegalovirus pneumonia, erythroderma, and gingival bleeding. In addition, three patients discontinued treatment due to treatment‐related TEAEs, and one patient had a grade 5 event that was considered treatment‐related (cytomegalovirus pneumonia). In cohort 3, no treatment‐related SAEs, discontinuations due to treatment‐related AEs, or treatment‐related deaths were observed.

3.3. Clinical activity

The DMR rate in patients treated with glasdegib plus LDAC in the expansion cohort was 46.7% (90% CI, 24.4–70.0; p < 0.0001). The best overall responses for patients in the expansion cohort are shown in Table 4; all patients with DMR achieved CR/CRi. Among the patients who achieved DMR or CR/CRi, the median duration of response was 10.1 months (95% CI, 3.9–NE) and 9.5 months (95% CI, 3.9–NE), respectively. The median time to response was 2.3 months (95% CI, 0.9–5.0) for patients with DMR and 5.0 months (95% CI, 0.9–5.9) for patients with CR/CRi. The median follow‐up for OS was 15.4 months, and the probability of being event‐free at 12 months was 53.3% (95% CI, 26.3–74.4). The median OS was 13.9 months (95% CI, 3.8–18.8) (Figure 1A), with 10 deaths (66.7%). The median OS in DMR responders (n = 7) was 18.8 months versus 4.9 months in DMR nonresponders (n = 8) (Figure 1B).

TABLE 4.

Best overall response in patients with acute myeloid leukemia treated with glasdegib plus low‐dose cytarabine (LDAC) (expansion cohort)

	Expansion cohort
	Glasdegib + LDAC (n = 15)
DMR, n (%), [90% CI]	7 (46.7) [24.4–70.0]
Best overall response, n (%)
CR	6 (40.0)
CRi	1 (6.7)
MLFS	0 (0.0)
PR	0 (0.0)
PRi	1 (6.7)
MR	2 (13.3)
SD	2 (13.3)
Treatment failure	2 (13.3)
Indeterminate	0 (0.0)
Not evaluable	1 (6.7)

Abbreviations: CI, confidence interval; CR, complete remission; CRi, CR with incomplete blood count recovery; DMR, disease‐modifying response; MLFS, morphologic leukemia‐free state; MR, minor response; PR, partial remission; PRi, PR with incomplete blood count recovery; SD, stable disease.

FIGURE 1.

Kaplan–Meier estimates of overall survival for patients with acute myeloid leukemia treated with glasdegib plus low‐dose cytarabine (expansion cohort). (A) Overall population (N = 15). (B) Disease‐modifying response (DMR) responders (n = 7) vs. DMR nonresponders (n = 8). CI, confidence interval; NE, not evaluable; OS, overall survival.

Of the six patients treated with glasdegib plus LDAC in cohort 1, one patient (16.7%) with AML achieved a CR, with duration of response of 13.9 months. Best responses for the other three patients with AML were SD (n = 1) and treatment failure (n = 2), while the two patients with MDS achieved a best response of SD. The median OS was 18.0 months (1.9–NE) in the AML group, with a median follow‐up of 41.9 months, and 7.1 months (95% CI, NE–NE) in the MDS groups.

The best overall responses in patients treated with glasdegib plus cytarabine/daunorubicin (cohort 2) or azacitidine (cohort 3) are presented in Table S4.

3.4. Pharmacokinetics

The PK parameters after treatment with glasdegib plus LDAC (day 10 of cycle 1) and glasdegib alone (day 21 of cycle 1) in cohort 1 are summarized in Table 5. The ratios (glasdegib plus LDAC vs. glasdegib alone) of adjusted geometric means for glasdegib AUC_tau and C _max were 96.8% (90% CI, 48.2–194.7) and 89.0% (90% CI, 50.1–158.0), respectively (Table 6). The ratios (glasdegib plus LDAC vs. LDAC alone) of adjusted geometric means for cytarabine AUC_tau and C _max were 139.3% (90% CI, 73.4–264.4) and 128.7% (90% CI, 72.5–228.6), respectively (Table S5). Using a previously reported population PK model, the estimated PK parameters and exposures in Japanese patients were similar to those of non‐Japanese patients (Figure S1). ²⁴ The median plasma glasdegib concentration–time profiles in cohort 1 are presented in Figure S2. No drug–drug interactions were observed in cohorts 2 and 3 (Tables S6 and S7; Figures S3 and S4).

TABLE 5.

Summary of glasdegib pharmacokinetic parameters when given alone (day 21 of cycle 1) and in combination with low‐dose cytarabine (LDAC) (day 10 of cycle 1; cohort 1)

Parameter (units) a	Cohort 1 glasdegib + LDAC (n = 6)
Day 21 of cycle 1 (glasdegib alone)
AUCtau (ng·h/mL)	16,070 (113)
AUC4 (ng·h/mL)	4014 (85)
C max (ng/mL)	1317 (86)
C min (ng/mL)	341.6 (172)
C av (ng/mL)	670.5 (113)
C trough (ng/mL)	376.2 (142)
CL/F (L/h)	6.223 (113)
T max (h)	1.935 (0.97–2.00)
Day 10 of cycle 1 (coadministration)
AUCtau (ng·h/mL)	15,560 (58)
AUC4 (ng·h/mL)	3065 (57)
C max (ng/mL)	1172 (38)
C min (ng/mL)	317.6 (91)
C av (ng/mL)	648.3 (58)
C trough (ng/mL)	330.7 (92)
CL/F (L/h)	6.428 (58)
T max (h)	3.950 (1.00–4.05)

Abbreviations: AUC, area under the plasma concentration–time curve; AUC₄, AUC from time zero to 4 h; AUC_tau, AUC from time zero to time tau, the dosing intervals, where tau = 24 h; C _av, average plasma concentration; CL/F, oral plasma clearance; C _max, maximum plasma concentration; C _min, minimum plasma concentration; C _trough, lowest plasma concentration before the next dose; T _max, time to maximum plasma concentration.

^a Data are geometric mean (geometric % coefficient of variation) for all parameters except median (range) for T _max.

TABLE 6.

Summary statistics of glasdegib pharmacokinetic parameters when given alone (day 21 of cycle 1) and in combination with low‐dose cytarabine (LDAC) (day 10 of cycle 1; cohort 1)

Parameter (units)	Comparison (test vs. reference)	N (test)/N (reference)	Adjusted geometric means		Ratio, %	90% CI
			Test	Reference	Test/reference
AUCtau (ng·h/mL)	Glasdegib + LDAC vs. glasdegib alone	6/6	15,560	16,070	96.83	48.15–194.74
C max (ng/mL)			1172	1317	89.00	50.14–157.96

Abbreviations: AUC_tau, area under the plasma concentration–time curve from time zero to 24 h; CI, confidence interval; C _max, maximum plasma concentration.

3.5. Pharmacodynamics

Gene expression was analyzed in bone marrow aspirate samples collected from 12 patients while on‐treatment to determine whether expression changes in Hh signaling could be detected compared with baseline samples (n = 14). Previously, GLI1 transcript levels were found to decrease in skin punch biopsy samples collected from patients after glasdegib treatment. ¹⁴ Unfortunately, GLI1 transcript levels were too low in bone marrow aspirates to be detected by bulk RNA sequencing analysis. A linear mixed model analysis revealed significant suppression of the Hh signaling pathway with glasdegib treatment in the DMR population versus the non‐DMR population (p = 0.0496) at early time points from screening to cycle 9 day 1 (Figure 2). Consistent results were observed when using screening samples as a covariate while treating each visit as a categorical variable or only including the first two time points (i.e., baseline and cycle 3 day 1). The trend of Hh signaling in DMR/non‐DMR patients is similar, though not identical, to a plasma cell gene expression signature (data not shown), suggesting the reduction in Hh signaling might not be solely due to a reduction in leukemic blast counts, although this possibility cannot be eliminated.

FIGURE 2.

Linear mixed model analysis of hedgehog (Hh) signaling mRNA abundance in patients who achieved a disease modifying response (DMR) versus patients with no DMR. Participants who achieve DMR have lower Hh pathway signaling than participants who do not. Longitudinal analysis of transcriptomics data from bone marrow aspirate samples collected at screening, cycle 3 day 1, cycle 6 day 1, and cycle 9 day 1 revealed a reduction in the Hh signaling Molecular Signatures Database hallmark gene set in participants who experienced DMR compared with participants who did not. Data for each participant are shown, and the vertical bars represent the mean ± SE at each visit.

4. DISCUSSION

This phase I study determined the safety and efficacy of glasdegib in combination with LDAC in a Japanese population of patients with AML or high‐risk MDS. A previous study reported that the 100 mg dose of single‐agent glasdegib was safe and well tolerated in Japanese patients, with signs of preliminary clinical activity. ¹⁴ The results presented here extend these findings to glasdegib combinations and confirm the results observed in non‐Japanese patients with AML or high‐risk MDS. ¹⁵ , ¹⁶

Treatment with glasdegib in combination with LDAC appeared safe and well tolerated, with a similar safety profile to those reported in non‐Japanese patients. ¹⁵ No DLTs were observed, which is consistent with the phase I results in a non‐Japanese cohort. ¹⁶ As expected, some of the most common treatment‐related TEAEs were hematologic events, such as neutropenia and thrombocytopenia, and gastrointestinal events, such as nausea. Another frequently reported treatment‐related TEAE in this study was dysgeusia, which has (along with alopecia and muscle spasms) been previously reported with glasdegib and other Hh pathway inhibitors and could be related to their mechanism of action. ¹⁴ , ¹⁵ , ²⁵ , ²⁶ The dysgeusia and alopecia events observed in this study were all grade 1–2 in severity. None of the deaths or treatment discontinuations observed with glasdegib plus LDAC in this study were considered related to treatment.

The PK analysis did not reveal any drug–drug interactions between glasdegib and LDAC. The PK profile was generally consistent with that of 100 mg glasdegib monotherapy in Japanese patients and that of glasdegib monotherapy and LDAC combination in non‐Japanese patients. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ²⁴ Preclinical studies have shown that glasdegib is primarily metabolized by CYP3A4/5, and therefore, the effects of a CYP3A4 inhibitor (ketoconazole) and inducer (rifampicin) were previously investigated. ²⁷ , ²⁸ The coadministration of ketoconazole increased plasma exposure in healthy individuals, with rifampicin decreasing exposure. Together these studies indicate that concomitant use with strong and moderate CYP3A4 inhibitors or inducers should be avoided. ³ , ⁴ In this study, one patient in cohort 1 and five patients in the expansion cohort received medications (e.g., moderate CYP3A4 inhibitor) that may alter glasdegib exposure (data not shown), although these were not considered to have any effect on the planned PK analysis due to the timing.

The primary efficacy end‐point in the expansion cohort was met, with glasdegib plus LDAC showing statistically significant DMR of 46.7% (7/15 patients; p < 0.0001) in patients with AML, thereby exceeding the significance threshold of 26.7% (4/15 patients) required for a positive study. All seven patients achieved either a CR or CRi. The median OS was 13.9 months, which increased to 18.8 months in DMR responders. These results are consistent with those reported in a non‐Japanese population of patients. ¹⁵ In the BRIGHT AML 1003 trial, the CR/CRi rate was 24.4% in patients with AML treated with glasdegib plus LDAC, and the median OS was 8.3 months. ¹⁵ A post hoc analysis reported a median OS of 26.1 months in patients with AML who achieved CR. ²⁹ Importantly, this analysis also showed clinical benefits with glasdegib in patients who did not achieve CR. A long‐term analysis of the BRIGHT AML 1003 trial showed that the OS benefit of glasdegib plus LDAC was more pronounced among patients with secondary AML (9.1 months) compared with de novo AML (6.6 months). ³⁰ In our study, 9 of the 15 patients (60%) had secondary AML, which is relatively high and could partly impact the results.

This phase I study included an analysis of biomarkers that could help identify patients more likely to benefit from glasdegib. Although the analysis was confounded to some extent by the small sample size (n = 14 at baseline, n = 12 on‐treatment), an analysis of hallmark gene signatures revealed that Hh pathway signaling decreased in patients who achieved DMR versus patients who did not achieve a DMR (p = 0.0496), suggesting that patients exposed to glasdegib who achieved a DMR have reduced Hh pathway signaling.

Patients with AML ineligible for intensive chemotherapy who receive treatment with LDAC have poor outcomes, with low response rates (<20%) and short OS (<5 months). ¹⁵ , ³¹ , ³² In the global phase III VIALE‐C trial, the addition of venetoclax to LDAC in patients ineligible for intensive chemotherapy improved the median OS from 4.1 months to 7.2 months, but the difference failed to reach statistical significance (p = 0.11). ³² The CR/CRi rate was significantly improved from 13% with LDAC to 48% with venetoclax plus LDAC (p < 0.001).

This study also confirms the safety and preliminary efficacy of glasdegib plus intensive chemotherapy and azacitidine in patients with AML. Overall, the safety and efficacy results are consistent with those reported with glasdegib in these combinations in a non‐Japanese population of patients. ¹⁶ , ¹⁷ , ¹⁸ The most common treatment‐related TEAEs consisted of hematologic and gastrointestinal events, as well as class‐effect events associated with treatment with an Hh pathway inhibitor, including dysgeusia (both cohorts), alopecia (cohort 2), and muscle spasms (cohort 3). One DLT of grade 3 erythroderma was observed in cohort 2; this resolved with treatment discontinuation. Combination with azacitidine was well tolerated, with no DLTs, treatment‐related SAEs, or treatment‐related deaths.

As is common with phase I studies, the limitations of this study include the small number of patients in each treatment cohort, which can influence outcomes. In addition, the absence of a comparator arm limits the interpretability of the data. In the biomarker analysis, the small sample size limited the conclusions that could be drawn and emphasized the need for further confirmatory studies.

In conclusion, glasdegib in combination with LDAC was well tolerated and associated with clinical activity in untreated Japanese patients with AML, confirming previously reported results in a global cohort of patients. ¹⁵ The results of this study are important in an effort to expand the available treatment options in Japan. The results presented here are encouraging and warrant further studies of glasdegib plus LDAC in Japanese patients with AML. Glasdegib was evaluated in combination with cytarabine/daunorubicin or azacitidine in the phase III BRIGHT 1019 trial, which included both Japanese and non‐Japanese patients with AML. ³³

AUTHOR CONTRIBUTIONS

Koji Izutsu: Investigation; writing – review and editing. Kumi Ubukawa: Investigation; writing – review and editing. Takanobu Morishita: Investigation; writing – review and editing. Yasushi Onishi: Investigation; writing – review and editing. Kenichi Ishizawa: Investigation; writing – review and editing. Yosuke Fujii: Formal analysis; writing – review and editing. Nobuyuki Kimura: Formal analysis; writing – review and editing. Miyuu Yokochi: Formal analysis; writing – review and editing. Tomoki Naoe: Conceptualization; investigation; writing – review and editing.

FUNDING INFORMATION

This study was funded by Pfizer.

CONFLICT OF INTEREST STATEMENT

KIzutsu has received honoraria from Pfizer, SymBio, Sanofi, Janssen, Ono, Chugai, Takeda, BMS, Eisai, and Kyowa Kirin; served in a consulting or advisory role for Micron, IQVIA, Kyowa Kirin, and Sawai; and received research funding from Otsuka, AbbVie, Novartis, IQVIA, and Zenyaku Kogyo. YO has received honoraria from Novartis and Pfizer, and received research funding from Incyte, Janssen, Meiji Seika Pharma, Pfizer, and Takara Bio. KIshizawa has received lecture fees from BMS, Chugai, Eisai, Janssen, Ono, and Novartis, and received research funding from AbbVie, Novartis, Pfizer, and SymBio. YF, NK, and MY were full‐time employees of Pfizer during the conduct of this study. The other authors have no conflicts of interest.

ETHICS STATEMENT

Approval of research protocol by an Institutional Review Board: The study protocol was approved by the institutional review boards of the participating institutions. The study was conducted in compliance with the Declaration of Helsinki and followed the International Conference on Harmonization Good Clinical Practices guidelines.

Informed consent: All patients provided signed informed consent.

Registry and the registration no. of the study/trial: ClinicalTrials.gov identifier: NCT02038777.

Animal studies: N/A.

Supporting information

Figure S1.

Figure S2.

Figure S3.

Figure S4.

Table S1.

Table S2.

Table S3.

Table S4.

Table S5.

Table S6.

Table S7.

Izutsu K, Ubukawa K, Morishita T, et al. Glasdegib with intensive/nonintensive chemotherapy in Japanese patients with untreated acute myeloid leukemia or high‐risk myelodysplastic syndromes. Cancer Sci. 2024;115:1250‐1260. doi: 10.1111/cas.16054

DATA AVAILABILITY STATEMENT

Upon request, and subject to review, Pfizer will provide the data that support the findings of this study. Subject to certain criteria, conditions, and exceptions, Pfizer may also provide access to the related individual deidentified participant data. See https://www.pfizer.com/science/clinical‐trials/trial‐data‐and‐results for more information.