Assessment of ethnic differences in pharmacokinetics and clinical responses of acalabrutinib between Chinese and White patients with B‐cell malignancies

Abstract

Aims

The aim of this study was to assess differences between Chinese and White patients in pharmacokinetics (PK) of, and clinical response to, acalabrutinib and its pharmacologically active major metabolite, ACP‐5862, to support recommended dosing in Chinese patients with B‐cell malignancies.

Methods

A population pharmacokinetic (pop‐PK) analysis was conducted to integrate data from a Chinese and a Japanese study into the existing model. The effect of race (East Asian vs. non‐East Asian) on acalabrutinib and ACP‐5862 PK parameters was assessed. The relationships between model‐predicted exposures (e.g., acalabrutinib area under the plasma concentration–time curve for 24 h at steady‐state [AUC_24h,ss]) and best overall response and safety outcomes were investigated in Chinese patients and the overall population.

Results

The pop‐PK analysis included 686 patients with B‐cell malignancies (Chinese, n = 105; Japanese, n = 6). A two‐compartment model adequately described the PK profiles of acalabrutinib and ACP‐5862 for Chinese patients. East Asian race was a statistically significant covariate on relative bioavailability and apparent volume of the peripheral compartment of acalabrutinib, showing 28% higher PK exposures in Chinese patients. The exposure–efficacy analysis showed that efficacy plateaued in Chinese patients with 100 mg twice‐daily dosing while the exposure–safety analysis indicated a flat relationship of acalabrutinib AUC_24h,ss with grade ≥2 or ≥3 adverse events in both Chinese patients and the overall population with 100–400 mg daily doses.

Conclusions

Higher PK exposure was observed in Chinese patients vs. White patients but did not indicate a safety concern based on exposure–response relationships. The results supported using the 100 mg twice‐daily dosing regimen in Chinese patients.

What is already known about this subject

• Pop‐PK and E‐R relationships of acalabrutinib have been established in a White population.

• Higher PK exposures of acalabrutinib were observed in Chinese patients (n = 12) compared with White patients, pointing to potential safety concerns.

• Comprehensive assessment of PK and clinical response of acalabrutinib in a larger Chinese patient population was warranted.

What this study adds

• A Pop‐PK analysis demonstrated 28% higher PK exposures of acalabrutinib in Chinese patients at 100 mg twice daily.

• However, E‐R analyses showed minimal safety risk, and efficacy response plateaued over the exposure range of 100 mg twice daily.

• 100 mg twice daily was the optimal dosing regimen in Chinese patients.

1. INTRODUCTION

Non‐Hodgkin's lymphoma (NHL) is a heterogeneous group of lymphoproliferative disorders originating in B lymphocytes, T lymphocytes or natural killer cells. Mantle cell lymphoma (MCL) is a distinct subtype of NHL comprising approximately 7% of all adult NHL cases in the United States (US) and approximately 3% ¹ of all NHL cases in China, with a moderately aggressive clinical course and poor prognosis. Chronic lymphocytic leukaemia (CLL), another subtype of NHL, is the most common adult leukaemia in the US, with an estimated incidence of 4.2 cases per 100 000/year, ² and <1 case per 100 000/year ³ , ⁴ in Asia including China. Treatments for MCL and CLL have progressed significantly over the past decades, with promising clinical response. ⁵ , ⁶ However, most patients eventually relapse and require multiple lines of therapy. Newer therapies with less toxicity and more durable responses are therefore needed for the treatment of relapsed or refractory (R/R) MCL and R/R CLL, especially in China.

Bruton tyrosine kinase (BTK) inhibitors have dramatically transformed the treatment landscape of B‐cell malignancies such as CLL and MCL. ⁷ , ⁸ The first‐in‐class BTK inhibitor, ibrutinib, approved by the US Food and Drug Administration (FDA) in 2013, opened a new era of chemotherapy‐free medications to patients with B‐cell malignancies. However, a recent publication reported that ~24% of patients discontinued ibrutinib within 4 years in clinical practice due to adverse events (AEs) such as atrial fibrillation and haemorrhage. ⁹ Acalabrutinib, one of the second‐generation BTK inhibitors, had a lower incidence of atrial fibrillation and haemorrhage compared with ibrutinib ⁹ while retaining a high response rate in patients with R/R MCL and CLL. ¹⁰ , ¹¹ , ¹² Acalabrutinib was initially approved by the US FDA in 2017 for the treatment of patients with R/R MCL, with subsequent approvals in more than 50 countries.

Acalabrutinib demonstrates desirable pharmacokinetic (PK) properties in humans. It is rapidly absorbed after oral administration, with a median time to peak plasma concentration (T _max) of 0.75 h. ¹³ Its absolute bioavailability is 25%. ¹³ Acalabrutinib is extensively distributed throughout the body, with a mean steady‐state volume of distribution (V_ss) of approximately 34 L. ¹³ Acalabrutinib is predominantly metabolized by CYP3A enzymes into its major active metabolite, ACP‐5862. ¹⁴ A radio‐labelled human mass balance study indicated 84% of the dose is recovered in the faeces and 12% of the dose is recovered in the urine, with less than 2% of the dose excreted as unchanged acalabrutinib in urine and faeces. ¹⁴ Acalabrutinib eliminates rapidly from circulation, with a mean terminal half‐life of 1.4 h, whereas ACP‐5862 eliminates more slowly, with a mean terminal half‐life of 6.6 h. ¹⁵ The existing global population PK model has well characterized the PK of acalabrutinib and its active metabolite ACP‐5862. ¹³ , ¹⁴ , ¹⁶

To support the new drug application filings for acalabrutinib in China, a single‐arm pivotal study was conducted in Chinese patients. The PK data from this study in patients with intensive PK sampling, using a noncompartmental analysis approach, showed that, in Chinese participants, PK exposure measured with the area under the plasma concentration–time curve (AUC) of acalabrutinib was 48% higher than that in White patients. Higher exposures may affect the safety profile of acalabrutinib in Chinese patients, potentially necessitating different dosing regimens. Indeed, there are reports of higher PK exposures in East Asian patients, including Chinese and Japanese patients, that have led to different medication dosing regimens compared with the dosing regimens in other populations in other diseases. ¹⁷ , ¹⁸

To determine the optimal dosing regimen for acalabrutinib in Chinese patients with B‐cell malignancies, we conducted a population PK (pop‐PK) analysis in this study based on the existing global model with pooled data to estimate more precisely the magnitude of acalabrutinib and ACP‐5862 PK exposure differences between Chinese patients and White patients and to study covariates that may lead to such differences. We also further conducted an exposure–response (E–R) analysis to investigate the relationships between systemic drug exposure of acalabrutinib, ACP‐5862, and the total active moiety and key efficacy endpoints and safety outcomes in Chinese patients and the overall population with B‐cell malignancies.

2. METHODS

2.1. Data

A pivotal, phase 1/2, open‐label, two‐part, multicentre, single‐arm study (NCT03932331) assessed the safety, tolerability, PK and clinical efficacy of acalabrutinib in Chinese adult patients with R/R MCL, R/R CLL or other B‐cell malignancies. Twelve patients had intensive PK sampling (Cycle 0, Day 1: pre‐dose and 0.25, 0.5, 0.75, 1, 2, 4, 6, 8, 12 and 24 h post‐dose; Cycle 1, Day 8: pre‐dose and 0.25, 0.5, 0.75, 1, 2, 4, 6, 8 and 12 h post‐dose; Cycle 1, Day 28: 1, 2 and 4 h post‐dose) and 93 patients had only sparse PK sampling (Cycle 1, Day 8 and Day 15 or 22: 1, 2 and 4 h post‐dose).

Clinical data from this study (n = 105) were combined with a phase 1 study (NCT03198650) conducted in Japanese patients (n = 6), and global clinical studies ¹⁹ for the analyses. In total, plasma concentrations of acalabrutinib and ACP‐5862 and relevant covariate data from 14 clinical trials were included in the pop‐PK analysis: four phase 1 trials in healthy participants and 10 phase 1/2/3 trials in patients with B‐cell malignancies, including CLL, MCL, follicular lymphoma (FL), diffuse large B‐cell lymphoma (DLBCL), small lymphocytic lymphoma (SLL), multiple myeloma (MM), or Waldenström macroglobulinaemia (WM).

Analyses assessing the relationships between individual predicted exposures and the efficacy endpoints included patients from the pivotal phase 1/2 China study with R/R MCL (n = 34, in which n = 1 from part 1 and n = 33 from part 2 MCL cohort) and R/R CLL (n = 60 from part 2 CLL cohort), respectively.

For the exposure–safety (E–S) analysis, patients from seven clinical studies, including six global studies and the China study (n = 476, studies ACE‐CL‐001, ACE‐CL‐003, ACE‐CL‐007, ACE‐LY‐002, ACE‐LY‐003, ACE‐LY‐004 and D8220C00007) were included. This safety data pooling strategy best captured the relevant pooled population to facilitate review of the proposed R/R MCL and R/R CLL indications while adding the other broader haematological safety data (e.g., non‐CLL).

2.2. Pop‐PK analysis

In the existing global model, ¹⁹ the acalabrutinib structural model (Figure 1) was a two‐compartment model with five transit compartment, and sequential first‐order describing absorption as well as between‐occasion variability (BOV) on mean transit time (MTT) and relative bioavailability (F1), and between‐subject variability (BSV) on apparent clearance of parent drug (CL/F), apparent volume of the central compartment for parent drug (V_c/F), and apparent volume of the peripheral compartment for parent drug (V_p/F). The ACP‐5862 structural model (Figure 1) was a two‐compartment disposition model with a first‐order production rate of 0.4 CL/F (the fraction of the drug metabolized having been fixed to 0.4 based on a clinical absorption, distribution, metabolism and excretion study ¹⁴ ) and first‐order elimination with BSV on apparent clearance of the metabolite (CLM/F), apparent volume of the central compartment for the metabolite (V_cM/F), apparent volume of the peripheral compartment for the metabolite (V_pM/F), and apparent intercompartmental clearance for the metabolite (QM/F). The covariates evaluated for their impact on PK of acalabrutinib and ACP‐5862 included body weight, concomitant use of acid‐reducing agents (proton pump inhibitors [PPIs] and H2‐receptor antagonists [H2RAs]), race (White, Black, others), renal function (creatinine clearance), hepatic function (National Cancer Institute Organ Dysfunction Working Group criteria), health status (healthy volunteers vs. patients), Eastern Cooperative Oncology Group (ECOG) performance status score, age, indication, line of therapy, sex, etc. The covariates identified to explain variabilities included health status, ECOG performance status and concomitant administration of PPIs. In the current analysis, a covariate analysis was conducted to confirm the impact of PPI, ECOG score and health status on the PK parameters of acalabrutinib and ACP‐5862, along with the additional evaluation of the race effect (which is of special interest). Other covariates were not re‐tested because the small fraction of the new Chinese data (12% of the overall population) was unlikely to affect the analysis results for these covariates. In the existing global model, the additive residual error (residual unexplained variability [RUV]) models on natural log‐transformation concentrations for both parent drug and metabolite were used, which approximated exponential residual error on the original scale. Additionally, separate residual error parameters were included for samples obtained on occasions for which rich profiles vs. sparse samples had been collected.

FIGURE 1.

Diagram of the existing final structural model of acalabrutinib and ACP‐5862. CL/F, apparent clearance; CLM/F, apparent clearance of the metabolite; ECOG, Eastern Cooperative Oncology Group; F1, relative bioavailability; Fm, fraction metabolized; ka, first‐order absorption rate constant; K_tr, transit absorption rate constant [MTT (mean transit time) = (N _tr  + 1)/K_tr]; PPI, proton pump inhibitor; Q/F, apparent intercompartment clearance; QM/F, apparent intercompartment clearance of the metabolite.

The same model structure, random effect (BSV and RUV) models and covariate structure from the existing global model were used for the current analysis. Regarding the parameterization, the categorical variable was evaluated as a proportional shift as follows:

$$P_{x,i}={\mathit{TVP}}_x\times \left[1+P_{x,i}^{\mathit{cov}}\right]\times \exp \left({\eta }_{x,i}\right)$$

where P _x,i is the estimated parameter x in the ith patient, P _x ^cov is the increase or decrease in P _x in the ith patient compared to the typical value for patients with the respective covariate, TVP _x is the typical parameter value for the reference group, and ${\eta }_{x,i}$ represents log‐normally distributed random deviation from the typical parameter value with covariate effects accounted.

In the current analysis, missing PK concentrations were ignored, while missing dosing/PK sampling times were imputed with their scheduled times. For covariates, any covariate missing in more than 10% of patients was excluded from the analysis. Missing continuous covariates were imputed using the median or earlier or later assessment values. Missing values of categorical covariates were replaced by a separate category or combined with one of the other categories, as appropriate, depending on the number of patients in the missing category.

The fraction below the limit of quantitation (BLQ) was 4.2% for acalabrutinib in Chinese patients. We included all BLQ data and utilized a maximum likelihood estimation approach (M3 approach), as proposed by Beal, to fit the data. ²⁰

2.3. E–R analyses

Individual systemic exposure metrics at steady‐state such as AUC for 24 h at steady‐state (AUC_24h,ss) and maximum plasma concentration at steady state (C _max,ss) for acalabrutinib and ACP‐5862, derived from post hoc estimates, were used to investigate the relationships between PK and efficacy and safety responses. In addition, total active AUC_24h,ss or C _max,ss (exposure metric for the total active moiety) adjusted with respective BTK potency and protein binding ¹⁶ (shown below) was also included in the analysis.

$$\mathit{\text{Total active concentration}}=C_{\mathit{\text{parent}}}*{\mathit{fu}}_{\mathit{\text{parent}}}+C_{\mathit{\text{metabolite}}}*{\mathit{fu}}_{\mathit{\text{metabolite}}}*0.5$$

where C _parent and C _metabolite are molar concentrations of acalabrutinib and ACP‐5862, respectively; fu _parent (free fraction of acalabrutinib) = 0.025; fu _metabolite (free fraction of ACP‐5862) = 0.013; and where ACP‐5862 exhibits approximately 0.5‐fold potency for BTK inhibition compared with acalabrutinib. Acalabrutinib and ACP‐5862 concentrations in molar units were predicted using the pop‐PK model and converted to total active concentrations (ng/mL scale) using acalabrutinib's molecular weight (MW _{acalabrutinib}  = 465.5).

Efficacy endpoints evaluated in the analysis included best overall response (BOR) assessed by an independent review committee, categorized as complete response (CR), partial response (PR), stable disease, progressive disease (PD) and ‘not evaluated’ for R/R MCL and as PR, partial response with lymphocytosis (PRL), stable disease and ‘not evaluated’ for R/R CLL/SLL, and tumour regression assessed by change from baseline in the sum of the products of the greatest diameters (SPD) of index lesions for both R/R MCL and R/R CLL/SLL.

The safety endpoints that were evaluated included incidence of any grade ≥2 adverse events (AEs), incidence of any grade ≥3 AEs, and incidence of grade ≥2 AEs of clinical interest experienced by ≥5% of the overall population. AEs of clinical interest included anaemia, cardiac events, diarrhoea, headache, haemorrhage, hepatic events, hypertension, infections, neutropenia, and thrombocytopenia, following Medical Dictionary for Regulatory Activities version 21.1 terminology.

The graphical visualization and logistic‐regression modelling analyses were used to investigate the relationships of systemic drug exposures (acalabrutinib ACP‐5862 and total active moiety) with efficacy (exposure−efficacy [E–E]) response and safety (E–S) outcomes.

2.4. Model evaluation

Model evaluation included goodness‐of‐fit (GOF) and prediction‐corrected visual predictive check (pcVPC) plots. GOF plots included the standard set of dependent variables (DV) vs. population predictions (PRED), DV vs. individual predictions (IPRED), conditional weighted residuals (CWRES) vs. PRED, CWRES vs. time after dose, and evaluations of CWRES distributions.

2.5. Hardware and software details

The software package NONMEM, version 7.3.0 (ICON Development Solutions, Ellicott City, MD, USA) was used in the pop‐PK analysis. Model fitting was performed in a Linux environment (CentOS 7) with GFortran Compiler, version 5.2 (Gnu Compiler Collection, GCC). The Perl‐Speaks‐NONMEM version 4.4.8 (PSN, https://uupharmacometrics.github.io/PsN/) was used for executing NONMEM runs. The Stochastic Approximation Expectation Maximization (SAEM) algorithm was used to estimate model parameters. The objective function, condition number, empirical Bayes estimates (EBEs), shrinkage and relative standard error (RSE) were determined using importance sampling (IMP) with setting EONLY = 1 and MAPITER = 0. The software R (version 3.5.1, R‐project, www.r-project.org) was used for the exploratory analysis and post‐processing of NONMEM output (e.g., to assess GOF, to generate the individual exposure parameters, for E–R graphical visualization evaluation, and for logistic‐regression model analysis).

2.6. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, and are permanently archived in the Concise Guide to PHARMACOLOGY 2019/20. ²¹

3. RESULTS

3.1. Pop‐PK analysis

3.1.1. Dataset

The integrated analysis dataset comprised 11152 acalabrutinib and 3367 ACP‐5862 plasma concentrations and relevant covariates from 138 healthy participants and 686 patients with B‐cell malignancies. The East Asian data from the China and Japan studies contained 972 acalabrutinib and ACP‐5862 samples from 111 patients (105 Chinese patients and six Japanese patients), accounting for 13.5% of the overall population, among which 18 patients had both sparse and intensive PK data compared with 93 patients with only sparse PK data.

Baseline demographics and characteristics for the Chinese population are shown in Tables 1 and 2 for categorial and continuous variables, respectively. Two‐thirds of the population were male. Most patients had CLL (n = 64, 61.0%), followed by MCL (n = 34, 32.4%). The majority of patients (n = 101, 96.2%) had an ECOG performance status ≤1. Most patients had normal hepatic function (n = 83, 79.0%); the rest had mild or moderate impairment (n = 22, 21.0%). With respect to renal function, most patients had mild renal impairment (n = 57, 54.3%) followed by normal renal function (n = 27, 25.7%) and moderate impairment (n = 21, 20.0%); none had severe impairment or end‐stage disease. Mean (standard deviation [SD]) age was 60.5 (10.1) years and mean (SD) body weight was 65.9 (10.2) kg.

TABLE 1.

Demographics and clinical characteristics (categorical variables) – Chinese population (n = 105).

Characteristic	Variable	Patients (n = 105)
Sex	Male	75
	Female	30
Indication	Chronic lymphocytic leukaemia	64
	Mantle cell lymphoma	34
	Follicular lymphoma	4
	R/R diffuse large B‐cell lymphoma	1
	Small lymphocytic lymphoma	2
Hepatic function (NCI‐ODWG)	Normal	83
	Mild	19
	Moderate	3
	Severe	0
Renal function	Normal	27
	Mild	57
	Moderate	21
	Severe	0
	End stage	0
ECOG performance status	0	66
	1	35
	2	4
	3	0
Use of PPI	No	104
	Yes	1 a
Use of H2RA	No	103
	Yes	2

Abbreviations: ECOG, Eastern Cooperative Oncology Group; H2RA, H2‐receptor antagonists; NCI‐ODWG, National Cancer Institute Organ Dysfunction Working Group; PPI, proton pump inhibitor; R/R, relapsed/refractory.

^a This patient took the PPI omeprazole.

TABLE 2.

Demographics and patient characteristics (continuous variables) – Chinese population (n = 105).

Characteristics, mean (SD)	Patients (n = 105)
Age, years	60.5 (10.1)
Body weight, kg	65.9 (10.2)
Body mass index, kg/m2	23.7 (2.8)
Body surface area, %	1.7 (0.2)
Creatinine clearance, a mL/min	74.7 (17.5)
ALT, IU/L	17.25 (10.7)
AST, IU/L	21.39 (7.0)
Bilirubin, μmol/L	14.5 (6.9)

Abbreviations: ALT, alanine transaminase; AST, aspartate transaminase; SD, standard deviation.

^a Creatinine clearance was calculated based on the Cockcroft‐Gault equation.

3.1.2. Model result and evaluation

The East Asian PK data were accurately described by the existing pop‐PK model; both acalabrutinib and ACP‐5862 parameters (Table 3) were similar to those previously reported. ¹⁹ The plots of goodness of fit (Supplementary Figures S1 and S2) and prediction‐corrected visual predictive check (Figure 2) indicated proper data fitness and model performance for East Asian patients.

TABLE 3.

Final model parameter estimates.

Run: 525bCLL3, OFV: 8731, condition number: 26.9
Parameter (unit)	Estimate (RSE %)	Shrinkage a (%)
Fixed effects
CL/F (L/h)	130.3 (0.4)	–
Vc/F (L)	28.8 (4.5)	–
Q/F (L/h)	19.5 (1.3)	–
Vp/F (L)	105.6 (0.7)	–
Ka (h−1)	1.4 (6.6)	–
MTT (h)	0.4 (5)	–
CLM/F (L/h)	21.5 (0.6)	–
VcM/F (L)	24.3 (1.7)	–
QM/F (L/h)	21.1 (2.9)	–
VpM/F (L)	90.9 (1.3)	–
Covariate effects		–
F1_PPI b	−0.307 (6.3)	–
CL/F_HEALTHGP b	0.645 (17.1)	–
Vp/F_HEALTHGP b	−0.424 (5.9)	–
CL/F_ECOG2+ b	−0.128 (43.9)	–
Vp/F_EASTASIA	−0.381 (15)	–
F1_EASTASIA	0.226 (44.3)	–
Random effects		–
BSV CL/F (L/h)	23.2 (5.6)	37.8
BSV Vc/F (L)	306 (6.3)	26.7
BSV Vp/F (L)	35.9 (6.9)	50.8
BSV CLM/F (L/h)	13.2 (18.7)	70.4
BSV VcM/F (L)	37.4 (13.7)	69.6
BSV QM/F (L/h)	62.4 (13.8)	64.9
BSV VpM/F (L)	29.3 (18.8)	77.2
BOV MTT (h)	125 (2.2)	45.3
BOV F1	52.9 (1.7)	44.9
RE parent (SD)	0.641 (0.5)	12.2
RE parent sparse samples (SD)	0.925 (0.9)	7.9
RE metabolite (SD)	0.36 (1.3)	13.1
RE metabolite sparse samples (SD)	0.401 (1.5)	13.2

Abbreviations: BOV, between‐occasion variability; BSV, between‐subject variability; CL/F, apparent clearance of parent drug (acalabrutinib); CLM/F, apparent clearance of metabolite (ACP‐5862); ECOG, Eastern Cooperative Oncology Group; EASTASIA, Chinese and Japanese patients; F1, relative bioavailability; HEALTHGP, healthy individuals; ka, first‐order absorption rate constant; MTT, mean transit time; OFV, objective function value; PPI, proton pump inhibitor; Q/F, apparent intercompartmental clearance for parent drug (acalabrutinib); QM/F, apparent intercompartmental clearance for metabolite (ACP = 5862); RE, residual error; RSE, relative standard error; SD, standard deviation; V_c/F, apparent volume of the central compartment for parent drug (acalabrutinib); V_cM/F, apparent volume of the central compartment for metabolite (ACP‐5862); V_p/F, apparent volume of the peripheral compartment for parent drug (acalabrutinib); V_pM/F, apparent volume of the peripheral compartment for metabolite (ACP‐5862).

^a Shrinkage for BOVs is given as the mean of the four occasions.

^b Relative change (1 + estimate).

FIGURE 2.

Prediction‐corrected visual predictive check for the final model in East Asian patients for acalabrutinib and ACP‐5862. The solid and dashed lines are the median and the 5th and 95th percentiles of the observations, respectively. The shaded areas are the 90% confidence intervals of the median and the 5th and 95th percentiles predicted by the model.

3.1.3. Covariates analysis result

The previous significant covariates were confirmed in the current analysis. Results showed that CL/F was 64.5% higher and V_p/F was 42.4% lower in healthy participants than in patients with B‐cell malignancies. Patients with an ECOG performance status ≥2 had a 12.8% decrease in CL/F compared with patients with ECOG ≤1. PPIs were shown to reduce F1 by 30.7%.

3.1.4. Effect of race covariate

The effect of race on V_p/F and F, both individually and combined, was statistically significant (P < 0.05). As the smallest objective function value was observed when race was applied to V_p/F and F combined, the effect of race on V_p/F and F combined was evaluated in the final analysis; compared with non‐East Asian individuals, V_p/F was 38% lower and F was 23% higher in East Asian individuals.

The exposures of AUC_24h,ss and C _max,ss for acalabrutinib and ACP‐5862 derived from the empirical Bayes estimates at the intended dosing regimen of 100 mg twice daily (BID) were used to assess differences according to race. Results indicated that the geometric means of acalabrutinib and ACP‐5862 exposures for both AUC_24h,ss and C _max,ss in Chinese patients (n = 105) were 28% higher than in White patients (n = 512) (Figure 3).

FIGURE 3.

Effects of race on acalabrutinib and ACP‐5862 AUC_24h,ss and C _max,ss (derived from the most prevalent dose of 100 mg BID). The black line within the box shows the median and the box's upper and lower edges show the IQR. Whiskers extend to the highest value within 1.5·IQR. Data outside whiskers are shown as circles. Sample size for Black or African American, Chinese, Japanese, other, and White patients was 47, 105, 6, 44 and 622, respectively. AUC_24h,ss, area under the plasma concentration–time curve for 24 h at steady‐state; C _max,ss, maximum plasma concentration at steady state; BID, twice daily; IQR, interquartile range.

3.2. E–E analysis

Following doses of 100 mg BID, the E–E analysis indicated flat relationships between acalabrutinib, ACP‐5862 and total active moiety exposures and selected efficacy endpoints (BOR and SPD) in Chinese patients with R/R MCL (n = 34; Figure 4) and R/R CLL/SLL (n = 60; Figure 5).

FIGURE 4.

Box plot of AUC_24h,ss and C _max,ss for total active moiety, acalabrutinib, and ACP‐5862 by best overall response for patients with R/R MCL. Total active AUC_24h,ss is area under the plasma concentration–time curve from time 0 to 24 h (two dosing intervals) at steady‐state for the total active moiety (i.e., acalabrutinib + ACP‐5862 adjusted for molecular weight, potency and protein binding). Total active C _max,ss is the maximum plasma concentration at steady‐state for the total active moiety. Open circles represent estimated AUC_24h,ss and C _max,ss in each response category. The ends of the boxes are the lower and upper quartiles of AUC_24h,ss, with the middle line showing the median. The whiskers outside the box indicate the 1.5·IQR (interquartile range) of the AUC_24h,ss and C _max,ss range. Sample sizes for CR, PR, SD, PD and not evaluated were 12, 16, 1, 4 and 1, respectively. CR, complete response; MCL, mantle cell lymphoma; PD, progressive disease; PR, partial response; R/R, relapsed or refractory; SD, stable disease.

FIGURE 5.

Box plot of AUC_24h,ss and C _max,ss for total active moiety, acalabrutinib, and ACP‐5862 by best overall response for patients with R/R CLL/SLL. Total active AUC_24h,ss is the area under the plasma concentration–time curve from time 0 to 24 h (two dosing intervals) at steady‐state for the total active moiety (i.e., acalabrutinib + ACP‐5862 adjusted for molecular weight, potency and protein binding). Total active C _max,ss is the maximum plasma concentration at steady‐state for the total active moiety. Open circles represent estimated AUC_24h,ss and C _max,ss in each response category. The ends of the boxes are the lower and upper quartiles of AUC_24h,ss, with the middle line showing the median. The whiskers outside the boxes indicate the 1.5·IQR (interquartile range) of the AUC_24h,ss and C _max,ss range. Sample sizes for PR, PRL, SD and not evaluated were 50, 2, 6 and 2, respectively. CLL, chronic lymphocytic leukaemia; PR, partial response; PRL, partial response with lymphocytosis; R/R, relapsed or refractory; SD, stable disease; SLL, small lymphocytic lymphoma.

3.3. E–S analysis

In total, 476 patients from seven clinical studies, including six global studies ¹⁶ and the China single‐arm study, were included in the analysis, including 66 Chinese patients with R/R CLL/SLL and 34 with R/R MCL (Supplementary Table S1).

Because ACP‐5862 plasma concentrations were measured only in a subset of the overall population (n = 337), acalabrutinib AUC_24h,ss is the primary PK exposure metric discussed here. The plots of AUC_24h,ss and C _max,ss for acalabrutinib, ACP‐5862 and total active moiety stratified by grade ≥2 and grade ≥3 AEs are presented for the global and Chinese populations in Supplementary Figures S3–S8.

The logistic regression analysis indicated no statistically significant relationship between acalabrutinib AUC_24h,ss and grade ≥2 and grade ≥3 AEs in the overall population (Figure 6), indicating the absence of an E–S relationship across daily doses ranging from 100 to 400 mg. The majority of patients received 100 mg BID, for which results indicated a flat relationship between acalabrutinib AUC_24h,ss and grade ≥3 AEs in the overall population.

FIGURE 6.

Logistic‐regression model analysis of the relationship between grade ≥2 and grade ≥3 AEs and acalabrutinib AUC_24h,ss. The purple line is the regression line. The shaded region represents the 95% confidence interval for the regression line. AEs, adverse events; AUC_24h,ss, area under the plasma concentration–time curve from time 0 to 24 h (two dosing intervals) at steady‐state; BID, twice daily; QD, once daily.

4. DISCUSSION

In China, acalabrutinib has been approved for the treatment of patients with R/R MCL and R/R CLL/SLL. The current pop‐PK and E–R analyses using pooled data from global studies, a single‐arm China study, and a Japan study, assessed the differences in PK and clinical responses of acalabrutinib and ACP‐5862 between East Asian and non‐East Asian patients. Results suggested that race (East Asian vs. non‐East Asian) was a statistically significant covariate on the PK disposition of acalabrutinib, with up to 28% higher PK exposure in Chinese patients compared with White patients. However, the higher exposure did not raise safety concerns, as a flat relationship between acalabrutinib exposure and AEs was observed in patients with acalabrutinib daily doses ranging from 100–400 mg in the logistic regression analysis evaluating the relationship between acalabrutinib AUC_24h,ss and grade ≥3 AEs. Based on the E–E analysis, the efficacy response following acalabrutinib 100 mg BID reached a plateau, supporting robust clinical outcomes across populations. The 100 mg BID regimen of acalabrutinib is the recommended regimen for use in White patients with R/R MCL and CLL/SLL. In addition, this regimen has been implemented in multiple ongoing clinical trials with acalabrutinib monotherapy or combination therapy, in multiple indications. In the current analysis, although PK exposure of acalabrutinib was slightly higher compared to non‐Chinese patients, the flat exposure–response relationships indicated that dose adjustment in Chinese patients is not warranted. Additionally, acalabrutinib has a wide therapeutic window, with the maximum tolerated dose not observed at a 400 mg daily dose. The clinical data demonstrated favourable benefit–risk profiles in Chinese and non‐Chinese patients with R/R MCL, CLL/SLL, and other indications at the 100 mg BID dose regimen. Clinical pharmacodynamic data suggested that the proposed 100 mg BID dose regimen achieved the maximal BTK occupancy over 24 h. We believe the current proposed dosing, 100 mg BID, is appropriate for Chinese patients.

The reason for the higher exposure of acalabrutinib in Chinese patients is unclear. Of note, body weight was not a statistically significant covariate; the higher F (~23%) may explain this difference. The mechanism of the higher bioavailability observed in Chinese patients is not fully understood. For an oral drug, the elimination and absorption profiles are mainly influenced by the physicochemical properties of the drug and its formulation, as well as by the physiology of the gastrointestinal (GI) tract. The main GI parameters that may differ in race are epithelial transporter expressions and metabolic enzymes. ²² Gastric pH also needs to be considered, as acalabrutinib is pH‐dependent, with solubility decreasing with increasing pH. Moreover, acalabrutinib is the substrate of breast cancer‐resistance protein (BCRP). There have been some reports of the potential racial differences in some polymorphism genes of ABCG2 (the gene coding BCRP). ²² , ²³ , ²⁴ Zhang et al. ²⁵ reported that in East Asians, the reduced‐function single nucleotide polymorphism ABCG2 c.421C>A was more prevalent than in Whites, which has been shown to cause the interethnic difference in rosuvastatin PK. However, other reports have shown inconsistent and ambiguous results regarding the relationship of polymorphic genes with systemic drug concentrations; for example, in the retrospective analysis conducted by Tomita et al., ²⁶ results indicated that ABCG2 c.421C>A (with another gene SLCO1B1 c.521T>C) cannot explain the observed higher PK concentrations in Asians taking statins. Therefore, further studies are needed to confirm the correlation between BCRP allele frequency and drug concentrations in circulation.

A major limitation of this model is the large shrinkage (>60%) for the metabolite, ACP‐5862. The estimate of shrinkage of interindividual variability (IIV) for ACP‐5862 was relatively high, similar to a previous global model. ¹³ This was most likely due to the sparse sampling schedule for ACP‐5862. In the current analysis, we used the PK exposures of parent drug, metabolite and total active moiety to evaluate E–E and E–S relationships, all of which indicated flat relationships. Although large shrinkage was observed for IIV of the PK parameters of ACP‐5862, we believe this should not affect the overall results and conclusion regarding flat E–R relationships of acalabrutinib.

5. CONCLUSIONS

A higher PK exposure was observed in Chinese patients vs. White patients but did not point to a safety concern based on the E–R relationships. Current modelling analyses support using the 100 mg BID dosing regimen recommended for White patients in Chinese patients.

AUTHOR CONTRIBUTIONS

Conceptualization: Peiming Ma. Data curation: Tingting Yao and Qianwei Xu. Formal analysis: Tingting Yao, Junjie Ding, Shringi Sharma, Yunfei Li and Qianwei Xu. Funding acquisition: Not applicable. Investigation: Tingting Yao, Junjie Ding, Shringi Sharma, Yunfei Li and Qianwei Xu. Methodology: Peiming Ma, Junjie Ding and Shringi Sharma. Project administration: Peiming Ma, Tingting Yao and Junjie Ding. Resources: Shringi Sharma, Tingting Yao and Qianwei Xu. Software: Tingting Yao and Qianwei Xu. Supervision: Peiming Ma. Validation: Tingting Yao, Yunfei Li and Qianwei Xu. Visualization: Tingting Yao, Junjie Ding, Shringi Sharma and Yunfei Li. Writing—original draft: Tingting Yao and Junjie Ding. Writing—review and editing: Tingting Yao, Junjie Ding, Shringi Sharma, Yunfei Li, Qianwei Xu and Peiming Ma.

CONFLICT OF INTEREST STATEMENT

Tingting Yao, Sharma Shringi, Qianwei Xu and Peiming Ma are employees of AstraZeneca and may own stock. Junjie Ding and Yunfei Li are former employees of AstraZeneca.

Supporting information

Table S1 Number of patients with different dosing stratified by treatment group

Figure S1 Basic goodness‐of‐fit plots for the final model:

observations vs predictions (East Asian patients; acalabrutinib)

Figure S2 Basic goodness‐of‐fit plots for the final model:

observations vs predictions (East Asian patients; ACP‐5862)

Figure S3 Box plot of acalabrutinib AUC_24h,ss stratified by any grade ≥2 and grade ≥3 safety outcomes (overall population)

Figure S4 Box plot of acalabrutinib C _max,ss stratified by any grade ≥2 and grade ≥3 AEs (overall population)

Figure S5 Box plot of ACP‐5862 AUC_24h,ss stratified by any grade ≥2 and grade ≥3 AEs (overall population)

Figure S6 Box plot of ACP‐5862 C _max,ss stratified by any grade ≥2 and grade ≥3 AEs (overall population)

Figure S7 Box plot of total active AUC_24h,ss stratified by any grade ≥2 and grade ≥3 AEs (overall population)

Figure S8 Box plot of total active C _max,ss stratified by any grade ≥2 and grade ≥3 AEs (overall population)

Yao T, Ding J, Sharma S, Li Y, Xu Q, Ma P. Assessment of ethnic differences in pharmacokinetics and clinical responses of acalabrutinib between Chinese and White patients with B‐cell malignancies. Br J Clin Pharmacol. 2025;91(7):2008‐2019. doi: 10.1002/bcp.70018

DATA AVAILABILITY STATEMENT

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