Ceralasertib, an ATR kinase inhibitor, as monotherapy in Japanese patients with advanced solid malignancies: Results from a phase 1 study

Abstract

Background

Ceralasertib is an oral, selective, and potent inhibitor of ATR serine/threonine kinase, a key protein involved in cell cycle checkpoint regulation and DNA damage response. We report preliminary safety, tolerability, and pharmacokinetic data of ceralasertib monotherapy in Japanese patients with advanced solid tumors.

Methods

In this phase 1, open-label study, patients orally received ceralasertib twice daily 240 mg on days 1–7 (Cohort 1) or 160 mg on days 1–14 (Cohort 2) of a 28-day cycle, respectively; both cohorts also received a single ceralasertib dose 4 days before Cycle 1. Patients were aged ≥ 18 years with solid malignancies refractory to standard therapies or for which no standard therapy exists. The primary objective was to determine ceralasertib safety; secondary objectives included assessing antitumor activity and pharmacokinetics. Exploratory objectives included biomarker analysis.

Results

Twelve of 14 patients screened received ceralasertib. At data cutoff (August 17, 2023), all 12 patients had experienced at least one treatment-emergent adverse event (AE; grade ≥ 3, n = 3). One patient in each cohort had a dose-limiting toxicity; no AE-related deaths were reported. In total, 6 patients had a best objective response of stable disease (Cohort 1, n = 2; Cohort 2, n = 4). A trend suggesting dose-proportional increases in exposure following single and multiple administration of ceralasertib was observed.

Conclusion

Ceralasertib monotherapy was generally well tolerated in Japanese patients with advanced solid tumors. The small number of patients enrolled prevents definitive conclusions on the efficacy of ceralasertib monotherapy to be made.

Trial registration

ClinicalTrials.gov, NCT05469919. Registration date: May 18, 2022.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10637-025-01592-x.

Introduction

In 2022, approximately one million new cancer cases and 420,000 cancer-related deaths occurred in Japan; the three most common cancers were lung, colon, and gastric, together accounting for approximately 40% of all new cases and over 40% of all deaths [1, 2].

Anti-programmed cell death protein 1 (PD-1) and programmed cell death ligand-1 (PD-L1) antibodies have revolutionized cancer treatment, and they are now standard of care for several cancers including lung, skin, and breast [3, 4]. However, resistance to anti-PD-(L)1 therapies is common and occurs through multiple mechanisms intrinsic or extrinsic to the tumor, including T-cell exhaustion and immunosuppression within the tumor microenvironment (TME) [5, 6]. Survival of patients who progress on anti-PD-(L)1 therapies remains low, with 2-year survival rates around 15% for patients with metastatic non-small-cell lung cancer in a real-world setting [7]. With low survival rates and potential for evolving resistance to anti-PD-(L)1 therapies, new therapeutic options are needed.

Ceralasertib is an oral, selective, and potent inhibitor of ATR serine/threonine kinase, a key protein involved in cell cycle checkpoint regulation and cellular response to DNA damage and replication stress [8]. Ceralasertib inhibits the repair of stalled replication forks during cell replication, leading to accumulation of double strand DNA breaks, genomic instability, and cell death [9]. Ceralasertib also activates the ATM-dependent signaling pathway and suppresses recombination-mediated DNA repair [10].

In preclinical studies, ceralasertib inhibited tumor cell proliferation and activated immune responses in the TME by reprogramming different cell types and initiating interferon-mediated cross talk [9]. Intermittent treatment with ceralasertib resulted in a reduction of T cells with an exhausted phenotype during the ‘on’ phase of treatment, and recovery of T cells with a non-activated phenotype during the ‘off’ phase in the TME. Ceralasertib also activated dendritic cells by inducting signaling by Type 1 interferons. Furthermore, ceralasertib has also shown enhanced activity when combined with poly(ADP-ribose) polymerase inhibitors in patient-derived xenograft models from patients with BRCA-mutation-positive, triple-negative breast cancer, demonstrating its potential as a combination partner [10].

In the phase 1 PATRIOT study (NCT02223923), ceralasertib monotherapy demonstrated antitumor activity in patients with advanced solid tumors with no standard treatment options [11], with 39 of 66 evaluable patients (59%) having disease control (confirmed partial response [PR], n = 5; stable disease or unconfirmed PR, n = 34). In phase 2 studies, ceralasertib showed promising antitumor activity and manageable safety in combination with durvalumab in patients with advanced gastric cancer [12], metastatic melanoma [13], and non-small-cell lung cancer [14], with objective response rates (ORR) ranging from 14 to 31%.

These findings contribute to the growing body of preclinical and clinical evidence supporting the safety and potential efficacy of ceralasertib monotherapy in patients with advanced solid malignancies who had exhausted current standard of care therapies. Accordingly, this study (NCT05469919) reports preliminary data on safety, tolerability, and pharmacokinetics (PK) of ceralasertib monotherapy in Japanese patients with advanced solid tumors refractory to standard therapies or for which no standard therapy exists. The study aims to evaluate whether the safety, PK and efficacy of ceralasertib in Japanese patients with advanced solid malignancies are consistent in those observed in previous global studies. Two dosing regimens were explored to assess the balance between efficacy and tolerability: 160 mg BID (with a dosing schedule of 2 weeks on, 2 weeks off) which was previously shown to be well tolerated in the global phase 1 PATRIOT study [11] and 240 mg BID (1 week on, 3 weeks off) which was supported by preliminary safety data and PK modeling. Simulations predicted that the 240 mg dose would maintain plasma concentrations above the estimated IC90 for ATR inhibition in > 95% of patients, with exposure–response modelling indicating an optimal pharmacodynamic/toxicity balance (data on file). These regimens aimed to identify an optimal schedule that maintains therapeutic plasma concentrations while minimizing dose-limiting toxicities (DLT).

Methods

Study design

This phase 1, open-label study (NCT05469919) enrolled patients from two centers in Japan. This study followed a sequential dose escalation design consisting of two cohorts. In Cohort 1, patients received ceralasertib 240 mg orally as a single dose on Day 1 of Cycle 0 (C0D1) and twice daily (BID) on Days 1–7 of a 28-day cycle; in Cohort 2, patients received ceralasertib 160 mg orally as a single dose on C0D1 and BID on Days 1–14 of a 28-day cycle (Supplementary Fig. 1). Cycle 0 duration for both cohorts was 4 days.

All patients were assessed for DLTs in Cycles 0 and 1. Three to six evaluable patients were enrolled in each cohort using a “rolling 6” design. If no DLT was observed in 3–6 evaluable patients or only one DLT was observed in 6 evaluable patients in the cohort, the dose was deemed tolerable and transition to the following cohort occurred. Patients received treatment until disease progression, unacceptable toxicity, or withdrawal from the study.

Patients

Key inclusion criteria included: adults ≥ 18 years of age with histologically or cytologically confirmed, malignant solid tumors refractory to standard therapies or for which no standard therapy exists; Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1 with no deterioration between screening and the first dose of study treatment; and no cancer-associated cachexia. The full eligibility criteria are available in the Supplementary Methods.

Study endpoints and assessments

The primary endpoints were safety and tolerability of ceralasertib, including the incidence of adverse events (AEs), serious AEs (SAEs), and DLTs. The number of patients experiencing each AE was summarized according to the Medical Dictionary for Regulatory Activities (MedDRA) System Organ Class and Preferred Terms. All AEs were graded according to Common Terminology Criteria for Adverse Events (CTCAE) version 4.03. Treatment-emergent AEs (TEAEs) were defined as any AE occurring or worsening after the first administration of study treatment until the defined 30-day follow-up period after completing or discontinuing the study treatment. Treatment-related AEs (TRAEs) were defined as any AEs determined to be causally related to ceralasertib by the investigator. DLTs were defined as any toxicity not attributable to the disease or disease-related processes under investigation that occurred before the end of Cycle 1 and met the criteria listed in the Supplementary Methods. Dose modifications for ceralasertib were implemented in response to clinically significant or unacceptable toxicities. Treatment was interrupted and supportive care provided until the toxicity resolved to ≤ CTCAE Grade 1 or 2, per predefined criteria. Dose re-escalation or continuation at the lowest dose level was permitted based on clinical response and investigator discretion. Repeat interruptions were allowed for up to 28 days and longer interruptions required consultation with the study physician.

Secondary endpoints included the antitumor activity and PK of ceralasertib. Antitumor activity was assessed by evaluating the ORR, duration of response (DoR), percentage change in tumor size, and progression-free survival (PFS); tumors were evaluated at baseline using computed tomography or magnetic resonance imaging. The method of assessment used at baseline was used at each subsequent follow-up assessment until radiologically confirmed disease progression as determined by the investigator and defined by Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Tumor assessments were performed every 8 weeks (± 1 week) thereafter until objective disease progression. ORR was defined as the proportion of patients with a best overall response (BOR) of confirmed complete response (CR) or confirmed PR and was restricted to patients with measurable disease at baseline. DoR was defined as the duration from the date of the first documentation of response (CR or PR), which was subsequently confirmed, to the date of documented disease progression or death due to any cause in the absence of disease progression, and was listed for patients with a BOR of CR or PR. PFS was defined as the time from the start of treatment until the date of objective disease progression or death (by any cause in the absence of progression), regardless of patient withdrawal from therapy or receipt of another anticancer therapy prior to progression.

PK was assessed by determining plasma ceralasertib concentrations and derived PK parameters, including maximum plasma concentration (C_max), time to maximum concentration (t_max), terminal half-life (t_1/2), area under the concentration–time curve up to the last quantifiable sample (AUC_0–t), apparent clearance (CL/F), and apparent volume of distribution during the terminal phase (V_z/F). Venous blood samples for determining plasma concentrations of ceralasertib were collected before dosing on C0D1 and at 0.25, 0.5, 1, 2, 3, 4, 6, 12, 24, 48, and 72 h post-dose as well as before dosing on C1D1 for both cohorts, to determine single-dose PK. For multiple-dose PK, samples were collected before dosing on Cycle 1 Days 7 (Cohort 1) or Day 8 (Cohort 2), and at 0.25, 0.5, 1, 2, 3, 4, 6, 12 h post-dose; in both cohorts, the 12-h sample was taken before the evening dose of ceralasertib.

Biomarker analysis from baseline tissue and blood samples

Tumor tissue samples collected during screening were analyzed using the FoundationOne CDx assay, which detects mutations, copy number variations, and selected rearrangements across 324 cancer-related genes and provides information on microsatellite instability (MSI) and tumor mutational burden (TMB) [15].

Baseline blood samples were collected from all 12 patients and analyzed using the Guardant Health OMNI assay, a next-generation sequencing platform designed to detect single nucleotide variants (SNVs), insertions and deletions (InDels), fusions, amplifications, MSI-high (MSI-H) status, and TMB score reported as mutations per megabase (mut/Mb) [16].

Statistical analyses

Safety data were analyzed in the safety analysis set and summarized using descriptive statistics; the safety analysis set included all patients who received at least one dose of study treatment.

Efficacy endpoints were analyzed in the response evaluable set, which included all patients who had measurable disease at baseline and received at least one dose of ceralasertib. The BOR of CR, PR, stable disease (SD), progressive disease (PD), and not evaluable (NE) were summarized descriptively. Stable disease needed to be maintained for ≥ 7 weeks from Cycle 1 Day 1. The Clopper-Pearson exact method for binomial proportion was used to calculate the 95% confidence intervals (CI) for ORR. The percentage changes in target lesion tumor size from baseline at 8 and 16 weeks were summarized descriptively. PFS data were analyzed in the safety analysis set. PFS was summarized by cohort and assessed using the Kaplan‒Meier method.

The PK analysis set consisted of patients who had a quantifiable plasma concentration of ceralasertib, and without protocol deviations or events which may have affected PK evaluations. PK parameters were derived by standard non-compartmental analysis using the individual plasma concentrations of ceralasertib. C_max and t_max were determined by inspection of concentration–time profiles. The terminal elimination rate constant (λ_z) was calculated by log-linear regression of the terminal portion of the concentration–time profiles. AUC_0–12 was calculated using linear up/down trapezoidal rule.

Study oversight

This study was conducted in accordance with the principles outlined in the Declaration of Helsinki, the International Council for Harmonization, Good Clinical Practice guidelines, and relevant regulatory authorities. All patients provided written informed consent.

Results

Baseline characteristics and demographics

Of 14 patients screened, 12 received ceralasertib. Six patients were assigned to each cohort. At data cutoff (August 17, 2023), 1 (16.7%) patient in Cohort 1 and 2 (33.3%) patients in Cohort 2 were still participating in the study. Five (83.3%) patients in Cohort 1 and 4 (66.7%) patients in Cohort 2 had discontinued treatment due to disease progression. The median total time on study was 109.5 days (range, 54–393 days) for Cohort 1 and 134.5 days (78–253 days) for Cohort 2.

Demographics and baseline characteristics are shown in Table 1. All patients were Asian with a median age of 57 years (range, 38–73); 8 (66.7%) patients were female. The most common primary tumor locations were colon (3 [25%] patients), breast, and lung (2 [16.7%] patients each). All patients had metastatic disease. In total, 11 (91.7%) patients had received prior anticancer therapy, including targeted therapy (11 [91.7%] patients) and immunotherapy (4 [33.3%] patients); 1 patient with primary tumor of the testis in Cohort 1 did not receive prior anticancer therapy as no standard-of-care treatment is currently available. Ten (83.3%) patients had received ≥ 2 previous chemotherapy regimens; the median number of previous chemotherapy regimens was four (range, 1–16; ranges for Cohort 1 and 2 were 3–10 and 1–16, respectively). Although patients in each cohort presented with variable tumor types, there were no major differences between the treatment cohorts in demographics and baseline characteristics.

Table 1.

Patient demographics and baseline characteristics

	Cohort 1 240 mg BID (n = 6*)	Cohort 2 160 mg BID (n = 6*)	Total (N = 12)
Median age, (range)	57.5 (38–73)	55.5 (48–62)	57.0 (38–73)
Sex, n (%)
Male	2 (33.3)	2 (33.3)	4 (33.3)
Female	4 (66.7)	4 (66.7)	8 (66.7)
Race, n (%)
Asian	6 (100)	6 (100)	12 (100)
ECOG PS, n (%)
0	4 (66.7)	3 (50.0)	7 (58.3)
1	2 (33.3)	3 (50.0)	5 (41.7)
Primary tumor location, n (%)
Breast	2 (33.3)	0	2 (16.7)
Cervix uteri	1 (16.7)	0	1 (8.3)
Colon	2 (33.3)	1 (16.7)	3 (25.0)
Lung	0	2 (33.3)	2 (16.7)
Oropharynx	0	1 (16.7)	1 (8.3)
Ovary	0	1 (16.7)	1 (8.3)
Testis	1 (16.7)	0	1 (8.3)
Unknown	0	1 (16.7)	1 (8.3)
Stage at screening,† n (%)
IB	1 (16.7)	0	1 (8.3)
III	1 (16.7)	0	1 (8.3)
IIIA	0	2 (33.3)	2 (16.7)
IIIB	1 (16.7)	0	1 (8.3)
IV	2 (33.3)	2 (33.3)	4 (33.3)
Missing	1 (16.7)	2 (33.3)	3 (25.0)
Disease classification, n (%)
Locally advanced‡	1 (16.7)	3 (50.0)	4 (33.3)
Metastatic§	6 (100)	6 (100)	12 (100)
Previous treatment modalities,¶ n (%)
Cancer therapy	5 (83.3)	6 (100)	11 (91.7)
Immunotherapy	1 (16.7)	3 (50.0)	4 (33.3)
Hormonal therapy	1 (16.7)	0	1 (8.3)
Targeted therapy	5 (83.3)	6 (100)	11 (91.7)
Other	5 (83.3)	6 (100)	11 (91.7)
Radiotherapy	2 (33.3)	1 (16.7)	3 (25.0)
Adjuvant	0	1 (16.7)	1 (8.3)
Palliative	2 (33.3)	0	2 (16.7)
Not applicable	0	1 (16.7)	1 (8.3)
Number of previous regimens, n (%)
1	0	1 (16.7)	1 (8.3)
2	0	1 (16.7)	1 (8.3)
3	1 (16.7)	1 (16.7)	2 (16.7)
4	1 (16.7)	1 (16.7)	2 (16.7)
5	1 (16.7)	0	1 (8.3)
≥ 6	2 (33.3)	2 (33.3)	4 (33.3)
No cancer therapy	1 (16.7)	0	1 (8.3)
Median (range)	5.0 (3–10)	3.5 (1–16)	4.0 (1–16)

Data cutoff: August 17, 2023

^*Number of patients in the safety analysis set within each cohort. Percentages are calculated from the number of patients in the safety analysis set in each cohort

^†Staged according to AJCC, if applicable

^‡Patients who had at least one locally advanced site of disease

^§Patients who had at least one metastatic site of disease

^¶The same patient may have had more than one previous disease-related treatment

AJCC, American Joint Committee on Cancer; BID, twice daily; ECOG PS, Eastern Cooperative Oncology Group performance status

Safety

The median total duration of exposure and actual exposure to ceralasertib was 54.0 days (range, 11–393) and 18.5 days (8–93) respectively, for Cohort 1, and 90.0 days (18–223) and 50.0 days (15–113), respectively, for Cohort 2.

At data cutoff, all patients had experienced at least one TEAE (Table 2). The most common TEAEs in Cohort 1 were anemia and nausea (3 [50%] patients each), followed by constipation, neutrophil count decreased, and platelet count decreased (2 [33.3%] patients each) (Supplementary Table 1). In Cohort 2, the most common TEAEs were anemia, nausea, and platelet count decreased (4 [66.7%] patients each), followed by white blood cell (WBC) count decreased (3 [50%] patients). In total, 5 (83.3%) patients in Cohort 1 and all patients in Cohort 2 experienced TRAEs (Table 2). There were no TEAEs or TRAEs leading to death.

Table 2.

Safety summary

AE category, n (%)	Cohort 1 240 mg BID (n = 6)	Cohort 2 160 mg BID (n = 6)
Any TEAE	6 (100)	6 (100)
Grade ≥ 3	1 (16.7)*	2 (33.3)
Any TRAE	5 (83.3)	6 (100)
Grade ≥ 3	1 (16.7)*	1 (16.7)†
Any TEAE leading to death	0	0
Any TRAE leading to death	0	0
Any SAE	0	2 (33.3)
Grade ≥ 3	0	2 (33.3)
Any SAE related to treatment	0	1 (16.7)†
Grade ≥ 3	0	1 (16.7)†
Any TEAE leading to discontinuation of ceralasertib	1 (16.7)*	0
Related to treatment	1 (16.7)*	0
Any TEAE leading to dose interruption of ceralasertib	0	1 (16.7)†
Related to treatment	0	1 (16.7)†
Any TEAE leading to dose reduction of ceralasertib	1 (16.7)	1 (16.7)†
Related to treatment	1 (16.7)	1 (16.7)†
Any DLT	1 (16.7)*	1 (16.7)†

Data cutoff: August 17, 2023

^*The same patient appears in these categories

^†The same patient appears in these categories

Patients with multiple events in the same category are counted only once in that category. Patients with events in more than one category are counted once in each of those categories

Includes AEs with an onset date on or after the date of the first dose, up to and including 30 (+ 7) days following the date of the last dose or 30 (+ 7) days after study drug discontinuation, whichever occurred last

AE, adverse event; BID, twice daily; DLT, dose limiting toxicity; SAE, serious adverse event; TEAE, treatment-emergent adverse event; TRAE, treatment-related adverse event

Of 12 patients enrolled, 9 patients had low grade (i.e., grade ≤ 2) events only; the recorded grade ≥ 3 TEAEs, grade ≥ 3 TRAEs, SAEs, treatment discontinuations, dose interruptions, and DLT occurred in the remaining 3 patients (Table 2).

One patient in Cohort 1 experienced a grade 3 TEAE of liver disorder. This patient had metastatic liver disease at screening with elevated alanine aminotransferase (ALT: 0.400 µkat/L), aspartate aminotransferase (AST: 0.550 µkat/L), and gamma glutamyl transferase (GGT: 2.584 µkat/L). On Cycle 1 Day 8, this patient’s laboratory investigations showed elevated ALT (1.784 μkat/L), AST (3.417 μkat/L) and GGT (3.651 μkat/L). The liver disorder, deemed to be treatment related by the investigator, resulted in discontinuation of ceralasertib on Cycle 1 Day 17 and was identified as a DLT by the Safety Review Committee. The patient did not receive any treatment for liver disorder. Concurrent with the liver disorder event, this patient also experienced non-serious adverse events, including constipation, anemia and pruritus. This patient was the only patient from both cohorts to discontinue ceralasertib due to a TRAE, and the only patient in Cohort 1 to experience a DLT. Another (16.7%) patient in Cohort 1 with grade 2 TRAE of neutrophil count decreased also experienced a dose reduction of ceralasertib to 160 mg BID from Cycle 3 Day 1.

In Cohort 2, 1 patient had a grade 3 TEAE of anaphylactic reaction, an SAE which required hospitalization. This event was considered unrelated to ceralasertib by the investigator and study sponsor.

Another patient in Cohort 2 had grade ≥ 3 TEAEs of anemia and WBC count decreased (both grade 3), and platelet count and neutrophil count decreased (both grade 4). This patient experienced SAEs of anemia (grade 1 which worsened to grade 3) and platelet count decreased (a grade 2 event and a grade 3 event, both worsening to grade 4), which were important medical events requiring transfusion; all three SAEs were deemed to be causally related to ceralasertib by the investigator and study sponsor. This patient was the only patient from both cohorts to experience dose interruption, due to TEAEs of grade 2 neutrophil count decreased (which worsened to grade 4 and lasted longer than 4 consecutive days, thus classified as a DLT) and WBC count decreased. This patient also had a dose reduction of ceralasertib to 120 mg BID on Cycle 2 Day 1 due to the grade 3 SAEs of anemia and platelet count decreased (which worsened to grade 4 and was considered to be a DLT), grade 4 neutrophil count decreased and grade 3 WBC count decreased.

Antitumor activity

In the response evaluable set, no patients had a CR or PR (Fig. 1A, B; Table 3). Best objective response of SD was recorded in 2 (33.3%) patients in Cohort 1 and 4 (66.7%) patients in Cohort 2. Disease progression occurred in 3 (50%) patients in Cohort 1 and 2 (33.3%) patients in Cohort 2.

Fig. 1.

Best percentage change in target lesion size from baseline (A) and percentage change in target lesion size from baseline over time (B). Data cutoff: August 17, 2023. Best percentage change in target lesion size is the maximum reduction from baseline or the minimum increase from baseline in the absence of a reduction BID, twice daily

Table 3.

Best overall response

Best overall response	Number of patients, n (%)
	Cohort 1 240 mg BID (n = 6)	Cohort 2 160 mg BID (n = 6)
Objective response	0	0
Complete response*	0	0
Partial response*	0	0
Stable disease†	2 (33.3)	4 (66.7)
Progressive disease	3 (50.0)	2 (33.3)
Not evaluable	1 (16.7)	0
No post-baseline assessment	1 (16.7)	0

Data cutoff: August 17, 2023

Best overall response was determined for each patient based on the best response recorded from start of study treatment to end of treatment, including any assessments for confirmation after the end of treatment

^*Response required confirmation at least 4 weeks after previous complete response/partial response. Patients with a confirmed response were counted

^†Stable disease needed to be maintained for ≥ 7 weeks from Cycle 1 Day 1

BID, twice daily

The data cutoff for PFS was May 25, 2023. Median PFS was 2.1 months (95% CI, 1.64–not calculable [NC]) in Cohort 1 and 3.3 months (95% CI, 0.92–NC) in Cohort 2 (Supplementary Fig. 2).

Pharmacokinetics

Following single-dose administration of ceralasertib at 240 mg (Cohort 1) or 160 mg (Cohort 2) on C0D1, ceralasertib was rapidly absorbed with a median t_max of 2.0 and 1.9 h, respectively, independent of dose (Table 4). Plasma concentrations of ceralasertib declined in a generally biphasic manner after reaching C_max with a mean t_1/2 of 12.2 and 13.9 h in Cohorts 1 and 2, respectively (Supplementary Fig. 3). The inter-patient variability in ceralasertib exposure was generally moderate for C_max and AUC parameters (i.e., AUC_[0–12], AUC_last, or AUC_inf) in the two cohorts, with a coefficient of variance (CV) of 18.0% in Cohort 1 and 16.9% in Cohort 2 for C_max and ranging from 12.8% to 29.2% across both cohorts for the AUC parameters. CL/F was low, with mean values of 2.4 L/h and 2.3 L/h for ceralasertib doses of 240 mg (Cohort 1) and 160 mg (Cohort 2); V_z/F was moderate, with mean values of 40.5 L and 43.0 L, respectively.

Table 4.

PK parameters following single and multiple dose administration of ceralasertib

	Cohort 1 240 mg BID (n = 6)	Cohort 2 160 mg BID (n = 6)
Single-dose administration
tmax (h)
Median (range)	2.04 (1.98–3.97)	1.92 (0.85–2.13)
t1/2 (h)
Arithmetic mean (SD)	12.24 (3.472)	13.90 (5.733)
Cmax (ng/mL)
Geometric mean (CV%)	7670 (18)	6553 (16.89)
AUC(0–12) (ng x h/mL)
Geometric mean (CV%)	54,260 (12.8)	39,160 (16.33)
AUClast (ng x h/mL)
Geometric mean (CV%)	100,700 (19.14)	69,120 (25.54)
AUCinf (ng x h/mL)
Geometric mean (CV%)	102,800 (20.67)	71,440 (29.19)
CL/F (L/h)
Arithmetic mean (SD)	2.375 (0.4926)	2.318 (0.6707)
Vz/F (L)
Arithmetic mean (SD)	40.47 (7.307)	43.03 (8.284)
Multiple-dose administration
tss,max (h)
Median (range)	1.97 (0.95–2.83)	1.87 (0.52–2.92)
t1/2λz (h)
Arithmetic mean (SD)	9.335 (2.138)	9.136 (4.613)
Css,max (ng/mL)
Geometric mean (CV%)	11,970 (16.26)	8132 (16.01)
AUCss,last (ng x h/mL)
Geometric mean (CV%)	92,140 (20.55)	56,020 (17.05)
AUCss (ng x h/mL)
Geometric mean (CV%)	94,970 (20.38)	57,400 (17.41)
CLss/F (L/h)
Arithmetic mean (SD)	2.568 (0.4725)	2.822 (0.4864)
Vss/F (L)
Arithmetic mean (SD)	33.78 (6.163)	35.26 (11.95)
RacCmax
Arithmetic mean (SD)	1.586 (0.3195)	1.271 (0.2857)
RacAUC
Arithmetic mean (SD)	1.787 (0.4046)	1.480 (0.2318)
TCP
Arithmetic mean (SD)	0.9398 (0.1917)	0.816 (0.1564)

Data cutoff: August 17, 2023

AUC, area under the plasma concentration–time curve; AUC_(0–12), AUC up to 12 h; AUC_inf, AUC extrapolated to infinity; AUC_last, AUC from dosing to the time of last measured concentration; AUC_ss, AUC at steady state; AUC_ss,last, AUC from steady state to last measured concentration; BID, twice daily; CL/F, apparent clearance; CL_ss/F, apparent total clearance at steady state; C_max, maximum plasma concentration; C_ss,max, maximum plasma concentration at steady state; CV, coefficient of variation; PK, pharmacokinetic; R_ac, ratio of accumulation; SD, standard deviation; t_1/2, half-life; t_1/2λz, terminal half-life; TCP, temporal change parameter; t_max, time to maximum drug concentration; t_ss,max, time to reach maximum drug concentration at steady state; V_ss/F, apparent volume of distribution at steady state; V_z/F, apparent volume of distribution

Following multiple administrations of ceralasertib at 240 mg (Cohort 1) or 160 mg (Cohort 2), ceralasertib was rapidly absorbed with a median t_max at steady state (t_ss,max) of 2.0 and 1.9 h, respectively (Table 4). Mean t_1/2λz was 9.3 (Cohort 1) and 9.1 h (Cohort 2); these values are an underestimate due to insufficient duration to detect the terminal phase within a dosing interval of 12 h. The inter-patient variability in ceralasertib exposure was generally moderate for C_ss,max and AUC parameters (i.e., AUC_ss,last or AUC_ss) in the two cohorts, with a CV of 16.3% and 16.0% in Cohorts 1 and 2, respectively, for C_ss,max and ranging from 17.1% to 20.6% across both cohorts for the AUC parameters. Accumulation of ceralasertib in plasma was observed following BID multiple dosing with an arithmetic mean accumulation ratio (R_ac)C_max of 1.6 and 1.3 and R_acAUC of 1.8 and 1.5 in Cohorts 1 and 2, respectively (Supplementary Fig. 3). Minimal time-dependent changes in AUC were detected with multiple dosing for the two dose levels, with arithmetic mean temporal change parameter (TCP) values of 0.94 in Cohort 1 and 0.82 in Cohort 2.

Biomarker analysis

Tissue samples were collected from 10 of 12 patients enrolled. Of the 10 tissue samples available, one was of insufficient quality, resulting in evaluable data from nine patients. Pathogenic mutations that may serve as predictive biomarkers for the efficacy of ATR inhibitors were identified in all nine patients (Supplementary Table 2). In addition, genetic alterations in the ATM molecular pathway were identified in 6 patients using either tissue or blood samples (Supplementary Table 2).

Discussion

To our knowledge, this is the first study evaluating safety, tolerability, and PK of ceralasertib in Japanese patients with advanced solid tumors.

Ceralasertib monotherapy, given as 240 mg BID (Days 1–7 of a 28-day cycle; Cohort 1) or 160 mg BID (Days 1–14 of a 28-day cycle; Cohort 2), demonstrated acceptable safety with only 1 patient in each cohort having a DLT and no patients experiencing AEs with an outcome of death.

The most common AEs were mostly hematologic with the top three TEAEs across both cohorts being anemia, nausea, and platelet count decreased, consistent with the reported safety profile of ceralasertib monotherapy in the phase 1 PATRIOT study [11].

In our study, 2 (33.3%) patients in Cohort 1 and 4 (66.7%) patients in Cohort 2 had a BOR of SD; no patient had an objective response. Definitive conclusions could not be drawn regarding the efficacy of ceralasertib monotherapy due to the small number of enrolled patients. Ceralasertib in combination with durvalumab, an anti-PD-L1 therapy, has shown encouraging efficacy signals [12–14] and the phase 3 LATIFY study (NCT05450692) of ceralasertib in combination with durvalumab in patients with non-small-cell lung cancer is ongoing [17].

Following single and multiple oral dosing BID, ceralasertib was rapidly absorbed with median t_max and t_ss,max ranging from 1.9 to 2.0 h post dose. After single dosing, ceralasertib was eliminated from plasma with a mean t_1/2 ranging from 12.2 to 13.9 h. Both absorption and elimination kinetics were consistent with the expected PK of ceralasertib from previously reported data in Western patient populations [11, 18, 19]. Although a formal post-hoc analysis has not been performed, comparison of PK between Japanese patients in this study and Caucasian patients from the PATRIOT study showed broadly similar profiles. Median t_max was 1.9–2.0 h in Japanese patients and 0.5–4 h in Caucasian patients. Mean t₁/₂ after single dosing was slightly longer in Japanese patients (12.2–13.9 h) compared to the PATRIOT studies (11.2–12.8 h at higher doses), but steady-state values were comparable (9.1–9.3 h) [11]. Accumulation ratios for C_max and AUC were 1.3–1.6 and 1.5–1.8 in Japanese patients versus 1.6–2.2 in the PATRIOT study [11]. These findings suggest absorption, elimination and exposure of ceralasertib were not different across ethnicities.

Moderate inter-patient variability was observed in ceralasertib exposure across the two cohorts. Previous in vitro data suggest that the metabolic clearance of ceralasertib was mainly via CYP3A4 and partially via CYP2C8 [20, 21]. A relatively wide inter-patient variability of CYP3A4 expression and activity was previously reported [22], which could explain the moderate inter-patient variability seen for ceralasertib in this study.

A trend suggesting a dose-proportional increase in exposure following single and multiple dosing of ceralasertib was observed. However, caution should be applied when interpreting dose proportionality due to limited data availability from the two dose levels investigated and the moderate inter-patient variability of PK observed.

Accumulation of ceralasertib in plasma of up to 1.8-fold was seen following multiple administrations of ceralasertib at 160 mg or 240 mg, as expected from the t_1/2 observed following a single dose. Arithmetic mean TCP was 0.82–0.94, suggesting no marked time dependency of PK of ceralasertib after multiple dosing.

DNA damage response (DDR) gene mutations (excluding TP53) have been reported in up to 30% of patients with solid cancers [23–25]; however, in our study, they were identified in 9 out of 12 (75%) patients. Previous studies have reported that the presence of certain genetic mutations can predict sensitivity to ATR inhibitors [26, 27]. In our phase 1 cohort, no objective responses were observed. Correlating genetic mutation patterns with clinical responses was challenging, given the small sample size, the lack of selection based solely on DDR mutation status, and incomplete sample availability from all patients. Therefore, drawing definitive conclusions from the current analysis may be premature.

Several factors may explain the lack of observed efficacy. First, not all DDR gene alterations result in a complete loss of function or heightened replication stress; some mutations may result in cells retaining partial repair capacity or may activate alternative pathways, potentially reducing dependence on ATR-mediated repair [27, 28]. Second, intratumoral heterogeneity may lead to the presence of subclones without DDR mutations, thereby diluting the overall impact of ATR inhibition [29, 30]. Third, compensatory mechanisms—such as reversion mutations or upregulation of parallel signaling pathways—could restore some degree of gene function, further limiting sensitivity to ATR blockade [26].

While these considerations offer possible explanations, further studies involving larger and more appropriately selected cohorts are needed to clarify the role of DDR mutations as predictive biomarkers for ATR inhibitor responsiveness.

For example, across multiple clinical studies, molecular characterization of responders to ceralasertib has consistently highlighted DDR deficiencies and immune activation as a key determinant of therapeutic benefit. In the PATRIOT study, monotherapy responses were observed in tumors harboring ARID1A loss and other DDR defects, with responders showing elevated immune inflammation signatures [11]. Ceralasertib in combination with durvalumab has shown clinical activity across multiple tumor types, particularly in biomarker-selected populations. In gastric cancer, melanoma, and non-small cell lung cancer, therapeutic benefit was associated with features such as ATM loss, homologous recombination deficiency, high PD-L1 expression, and immune-inflamed tumor microenvironments [12–14]. These studies also highlighted immune activation and tumor microenvironment remodelling, such as enhanced interferon signalling and preserved CD8⁺ T-cell function, as key contributors to response. The tumor microenvironment plays a critical role in shaping antitumor immunity and therapeutic outcomes, influencing both drug sensitivity and resistance mechanisms. Collectively, these findings support the use of DDR deficiency and immune activation markers as predictive biomarkers for ATR inhibition, guiding patient selection in future studies. The main limitation of this study is the small number of patients enrolled; tumor types of patients in the cohorts were variable and definitive conclusions cannot be drawn regarding efficacy. Caution is needed when interpreting the dose proportionality of ceralasertib, as limited data were available due to only two dose levels being investigated.

Overall, the results from this study demonstrate that ceralasertib monotherapy was well tolerated with an acceptable safety profile in Japanese patients with advanced solid malignancies.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the writing, review and editing of the manuscript. In addition, authors contributed as follows. Yasutoshi Kuboki: investigation, supervision; Nobuaki Matsubara: validation, investigation, data curation; Hiromichi Nakajima: investigation; Takao Fujisawa: resources; Takafumi Koyama: data curation; Jun Sato: investigation; Yuki Katsuya, investigation; Aleksandra Kmieciak: supervision; Daniel Slade: methodology, supervision; Kiyomi Iwata: conceptualization, writing – original draft, visualization, supervision; Yusuke Takahashi: conceptualization, methodology, resources, writing – original draft, visualization, supervision, project administration; Masahiro Nii: formal analysis; Kosho Murayama: conceptualization, formal analysis, investigation, data curation, writing – original draft, visualization; Hisashi Kawasumi: conceptualization, methodology, validation, investigation, resources, supervision, project administration, funding acquisition; Noboru Yamamoto: investigation, resources.

Funding

The funder of the study participated in study design, data collection, data analysis, data interpretation, and writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Data availability

Data underlying the findings described in this manuscript may be obtained in accordance with AstraZeneca’s data sharing policy described at: https://www.astrazenecaclinicaltrials.com/our-transparency-commitments/. Data for studies directly listed on Vivli can be requested through Vivli at www.vivli.org. Data for studies not listed on Vivli could be requested through Vivli at https://vivli.org/members/enquiries-about-studies-not-listed-on-the-vivli-platform/. The AstraZeneca Vivli member page is also available outlining further details: https://vivli.org/ourmember/astrazeneca/.

Declarations

Competing interests

Jun Sato and Yuki Katsuya have no conflicts to disclose. Yasutoshi Kuboki has received grants or contracts (to the institution) from Taiho, Astellas, Eli Lilly, Takeda, Daiichi Sankyo, Astra-Zeneca, Boehringer Ingelheim, Chugai, Genmab, Incyte, AbbVie, Amgen, Merck, Hengrui, Novartis, Ono Pharmaceutical, Noile-Immune Biotech Inc, Kyowa Kirin, and BMS; consulting fees from Incyte, Takeda, Noile-Immune Biotech Inc, Amgen, and AbbVie; honoraria from Taiho, Eli Lilly, Takeda, Kyowa Kirin, and Amgen; and support for attending meetings and/or travel from Amgen. Nobuaki Matsubara participated in an advisory council or committee for Janssen, Pfizer, Novartis, Astellas and AbbVie; received honoraria from Sanofi and Janssen; and received grants or funds from Janssen, Astra-Zeneca, Bayer, Roche, MSD, Taiho, Astellas, Amgen, Eisai, Eli Lilly, Takeda, Pfizer, Chugai, Seagen, AbbVie, Telix and Novartis. Hiromichi Nakajima has received honoraria from Chugai, Takeda, Pfizer, Eli Lilly, Astra-Zeneca, Ono Pharmaceutical, Sanofi/Regeneron, and MSD; consulting fees from Terumo Corporation; grants or funds from Merck BioPharma; and notes that an immediate family member is employed by Otsuka Pharmaceutical Co., Ltd.. Takao Fujisawa has received honoraria from Amelieff and Astra-Zeneca. Takafumi Koyama has received honoraria from Chugai Pharma and Sysmex; and grants or funds from Chugai Pharma, Daiichi Sankyo RD Novare, Eli Lilly, Novartis, PACT Pharma, Boehringer Ingelheim, Astra-Zeneca, Pfizer, Takeda and Zymeworks. Aleksandra Kmieciak and Daniel Slade are employed by Astra-Zeneca, and own stocks/shares in Astra-Zeneca. Kiyomi Iwata, Yusuke Takahashi, Masahiro Nii, Toshio Kawata, Hisashi Kawasumi are employed by Astra-Zeneca K.K.. Kosho Murayama is employed by Astra-Zeneca K.K. and owns stocks/shares in Eisai, Astellas and Pfizer. Noboru Yamamoto has participated in an advisory council or committee for Eisai, Boehringer Ingelheim, Cmic, Chugai, Merck, Healios K. K., Mitsubishi Tanabe, Rakuten Medical, IQVIA, Noile-Immune Biotech, and Janssen Pharma; and received honoraria from Chugai, Daiichi Sankyo, and Eisai; and grants or funds from Astellas, Chugai, Eisai, Taiho, BMS, Pfizer, Novartis, Eli Lilly, AbbVie, Daiichi Sankyo, Bayer, Boehringer Ingelheim, Kyowa Kirin, Takeda, Ono Pharmaceutical, Janssen Pharma, MSD, Merck, GlaxoSmithKline, Chiome Bioscience, Otsuka, Carna Biosciences, Genmab, Shionogi, Toray, Kaken, Astra-Zeneca, InventisBio, Rakuten Medical, Amgen, Bicycle Therapeutics, and Zymeworks.