Pharmacokinetics, Bioequivalence, and Safety Studies of Crisaborole Ointment in Healthy Chinese Subjects

Abstract

The aim of this study was to evaluate the bioequivalence and safety of a single application of crisaborole ointment to test formulation and reference formulation in healthy subjects under fasting conditions. A total of 32 subjects were included and divided into 2 groups (test‐reference; reference‐test). A single‐center, single‐dose, 2‐formulation, 2‐period, 2‐sequence, open, randomized, and crossover trial design was adopted. Eligible healthy subjects were applied with the test preparation (domestic crisaborole ointment) or the reference preparation (original crisaborole ointment), followed by a 7‐day washout period. Blood samples were collected at predetermined time points before and after administration. After the development and verification of the blood concentration detection method for this study, the third‐party supplier used liquid chromatography‐tandem mass spectrometry to determine the plasma concentration of crisaborole and used SAS Version 9.4 software to calculate the pharmacokinetic parameters and evaluate the bioequivalence. In this study, the 90% confidence intervals of the geometric mean ratios of maximum concentration, area under the concentration‐time curve over the dosing interval, and area under the concentration‐time curve from time 0 to infinity were within the acceptable range (80%‐125%). During the study, 5 subjects had 8 adverse events, and no serious adverse events were reported. In this study, the tested formulation of crisaborole ointment is bioequivalent to the reference formulation, and the safety is comparable.

Atopic dermatitis (AD), a chronic inflammatory skin disorder, exhibits substantial global prevalence, with estimates ranging from 15% to 30% in pediatric populations and from 2% to 10% among adults. ¹ Current therapeutic algorithms prioritize topical interventions as first‐line management, principally employing corticosteroids and calcineurin inhibitors. However, the well‐documented limitations of these agents—including corticosteroid‐associated cutaneous atrophy and calcineurin inhibitor‐induced burning sensations—have necessitated the development of novel nonsteroidal topical alternatives with improved safety profiles. ² , ³ , ⁴

The intracellular enzyme phosphodiesterase 4 (PDE4), a critical modulator of inflammatory cascades through its hydrolysis of cyclic adenosine monophosphate, has emerged as a promising therapeutic target. Pathological overexpression and dysregulated activation of PDE4 in immune cells, particularly in inflammatory dermatoses like AD, have catalyzed clinical investigations into PDE4 inhibitors for cutaneous inflammation management. Mechanistically, PDE4 inhibition elevates intracellular cyclic adenosine monophosphate levels, subsequently activating protein kinase A–mediated suppression of proinflammatory cytokine production—a therapeutic pathway of particular relevance in AD pathogenesis characterized by Th2‐mediated inflammation. ⁴

Crisaborole (C₁₄H₁₀BNO₃), a low‐molecular‐weight (251.1 g/mol), boron‐containing compound, demonstrates favorable transdermal permeability attributed to its lipophilic properties and compact molecular structure. Following topical application, the drug exhibits limited systemic exposure due to rapid hydrolysis into inactive metabolites (metabolites 1 and 2) through boronate ester cleavage. Pharmacokinetic analysis revealed steady‐state plasma concentrations of these metabolites by Day 8, with mean accumulation ratios (Day 8/Day 1 area under the concentration–time curve [AUC] from time 0 to 12 h) of 1.7 and 6.3 for metabolites 1 and 2, respectively. Renal elimination accounts for >90% of total metabolite clearance, predominantly via glomerular filtration. ⁵ , ⁶ Approved for mild‐to‐moderate AD management in patients aged 2 years or older, crisaborole ointment has demonstrated an optimal safety profiles in postmarketing surveillance. ⁷

While substantial pharmacological characterization exists for crisaborole ointment, including comprehensive pharmacokinetic profiling and well‐established safety/efficacy parameters, bioequivalence data remain insufficiently characterized in the current literature. This randomized crossover study implemented a head‐to‐head comparative pharmacokinetic assessment of test versus reference crisaborole formulations in a healthy adult cohort under standardized application protocols.

Methods

Study Participants

This study adopted a single‐center, open‐label, randomized, single‐dose, 2‐period, 2‐sequence crossover design under fasting conditions. A total of 32 eligible participants meeting the following inclusion criteria were planned for enrollment: age 18 years or older, with male and female participants weighing 50.0 kg or more and 45.0 kg or more, respectively, and body mass index ranging from 19.0 to 28.0 kg/m² (inclusive). All subjects did not meet the following exclusion criteria:

1. Clinically significant abnormalities in physical examination, vital signs, or clinical assessments including laboratory tests, alcohol breath analysis, urine drug screening, or 12‐lead electrocardiogram findings.

2. Significant medical history involving major organ systems (cardiovascular, metabolic, gastrointestinal, neurological, endocrine, or respiratory) or documented hypersensitivity, including active allergic disorders (urticaria, eczema, etc.), history of specific/food/drug allergies and hereditary angioedema or idiopathic angioedema.

3. Dermatological conditions potentially interfering with administration site evaluation: inflammatory skin disorders (dermatitis, acne vulgaris, erythema, etc.) and preexisting lesions/abnormalities (edema, desquamation, ulceration) at potential application sites.

4. Use of topical dermatologic medications (corticosteroids, retinoids, immunomodulators) at administration sites within 28 days before screening.

5. Pregnancy (confirmed by β‐human chorionic gonadotropin testing), lactation, or women of childbearing potential refusing contraceptive measures.

6. Substance use disorder meeting Diagostic and Statistical Manual of Mental Disorders, Fifth Edition, diagnostic criteria within 12 months.

7. The average alcohol consumption in the 3 months before screening, the number of cigarettes smoked that did not meet protocol requirements, or the inability to stop alcohol/tobacco intake during the trial.

8. Blood/plasma donation (>400 mL) or acute blood loss (>200 mL) within 3 months before enrollment.

9. Participation in investigational drug trials with therapeutic exposure within 90 days preceding baseline.

10. Recent use (28 days prescreening) of vasoactive agents (calcium channel blockers, nitrates) and hemorheological modifiers (pentoxifylline, cilostazol).

11. Medication washout violations: prescription drugs: <14‐day discontinuation; over‐the‐counter medications/herbal products: <7‐day discontinuation.

12. Live vaccine administration within 7 days predosing or anticipated vaccination during trial.

13. Major surgical procedures within 90 days prescreening or planned operative interventions during the study period.

14. Caffeine/xanthine intake exceeding 300 mg/day within 48 hours before pharmacodynamic assessments.

15. Contraindications to study procedures: History of vasovagal syncope during phlebotomy. Investigator‐assessed unsuitable venous access. Any condition compromising protocol compliance/safety per the investigator judgment.

Study Design

This clinical trial was conducted in full compliance with the Declaration of Helsinki, China's Good Clinical Practice guidelines, and applicable local regulations. Before study initiation, all essential trial documentation—including the study protocol, informed consent forms, recruitment materials, and investigator's brochure—received formal approval from the Institutional Review Board of Hebei PetroChina Central Hospital (Ethics Review Approval Number: IRB2022‐105‐01). All enrolled participants provided written informed consent using the Institutional Review Board–approved informed consent form version before undergoing standardized screening procedures per protocol requirements. This trial was registered at the clinical trial register platform of the National Medical Products Administration (http://www.chinadrugtrials.org.cn) (No. CTR20223119).

Qualified subjects who passed preliminary screening arrived at the Phase I Clinical Trial Unit on Day 1 (the day before dosing) for admission‐related examinations and medical interviews. Investigators determined eligibility based on screening results and Day –1 assessments, with qualified subjects proceeding to randomization. Subjects were stratified into 2 sequence groups using a computer‐generated randomization schedule: Group A first applied the test formulation, followed by the reference formulation in subsequent periods, while Group B followed the reverse sequence. To ensure protocol compliance, subjects maintained an upright seated position under continuous supervision during application and for 4 hours after dosing (except for essential movements). Research personnel monitored subjects' hands to prevent inadvertent contact with application areas. At the 12‐hour time point, trained staff sequentially cleansed dosing sites according to administration timelines using standardized techniques.

Blood Sampling and Pretreatment

Biological samples were collected before dosing (0 hour, within 60 minutes before administration) and at 1, 2, 2.5, 3, 3.5, 4, 4.5, 5, 5.5, 6, 6.5, 7, 7.5, 8, 9, 10, 12, 14, 16, 24, and 36 hours after dosing. Approximately 4 mL of whole blood was collected per time point into vacuum blood collection tubes containing dipotassium ethylenediaminetetraacetate as anticoagulant. The samples were gently inverted 8‐10 times for proper mixing and subsequently centrifuged at 1700 × g for 10 minutes under controlled temperature conditions (4°C ± 2°C) within 1 hour of collection. Following centrifugation, approximately 0.8 mL of the upper plasma layer was aliquoted into primary cryovials for analysis, with the remaining plasma transferred to backup storage tubes. All processed plasma samples were required to be transferred to ultra‐low temperature freezers (−70°C or below) within 2 hours after blood collection. Plasma concentrations of crisaborole were quantified using validated liquid chromatography‐tandem mass spectrometry methods, with the resultant pharmacokinetic parameters being utilized for bioequivalence assessment.

Sample Assay Method

Plasma samples were processed via liquid‐liquid extraction, followed by quantification of crisaborole concentrations using a validated high‐performance liquid chromatography‐tandem mass spectrometry method. The chromatographic system consisted of an LC‐20ADXR HPLC system coupled with an Ultimate XB C18 column (2.1 × 50.0 mm, 3.0 µm). Mobile phases A and B were prepared in accordance with the standard operating procedures established by a third‐party analytical laboratory. The system interfaced with a Triple Quad 4500 mass spectrometer equipped with an electrospray ionization source operating in negative ion mode. Quantification employed multiple reaction monitoring of precursor → product ion transitions at m/z 250.1 → 117.9 for crisaborole and m/z 254.1 → 122.1 for the internal standard crisaborole‐d4.

The calibration standards for crisaborole were established at concentrations of 0.200, 0.400, 1.00, 5.00, 20.0, 50.0, 90.0, and 100 ng/mL, demonstrating a validated linear range of 0.200‐100 ng/mL, with a lower limit of quantification (LLOQ) at 0.200 ng/mL and upper limit of quantification at 100 ng/mL. Quality control (QC) samples were prepared at 5 concentration levels: LLOQ QC (0.200 ng/mL), low‐level QC (0.600 ng/mL), low‐mid‐level (3.0 ng/mL), mid‐level (30.0 ng/mL), and high‐level (75.0 ng/mL) to comprehensively evaluate intra‐ and interbatch precision and accuracy. The analytical method demonstrated tight precision control across calibration standards (excluding LLOQ), with maximum intrabatch coefficients of variation (CV) of 2.8% and mean accuracy deviations ranging from −3.8% to +2.8% for QC samples. LLOQ samples exhibited exceptional precision performance, showing a maximum intra‐batch CV of 1.4% with a mean accuracy deviation of −1.0%. Non‐LLOQ QCs maintained rigorous analytical consistency, achieving maximum intrabatch CV values of 3.8% and mean accuracy deviations between −2.7% and −0.7%. These precision and accuracy metrics confirm the method's robustness across the analytical range while demonstrating particular strength in low‐concentration quantitation, with all parameters remaining well within accepted bioanalytical criteria for regulated pharmaceutical analysis.

Pharmacokinetic Analysis

The pharmacokinetic parameters of crisaborole, including time to maximum concentration, maximum concentration (C_max), AUC over the dosing interval (AUC_0‐t), AUC from time 0 to infinity (AUC_0‐∞), and terminal elimination half‐life, were determined through noncompartmental analysis using WinNonlin 8.0 (Certara) or SAS 9.4 software (SAS Institute), with actual sampling times incorporated for subjects receiving both test and reference formulations. Of note, all values below the limit of quantification were entered as 0 and included as such in the calculation of means. Concentrations below the limit of quantification were assigned a value of 0 and incorporated as such in all pharmacokinetic calculations, including determination of mean plasma concentration values. For bioequivalence assessment, the primary pharmacokinetic parameters (C_max, AUC_0‐t, and AUC_0‐∞) were logarithmically transformed and subjected to analysis of variance employing a mixed‐effects model. The statistical model incorporated treatment sequence, formulation, and period as fixed effects, with subjects (nested within sequence) designated as random effects. Geometric mean ratios (GMRs) with 90% confidence intervals (CIs) were calculated between the test and reference formulations. Bioequivalence was concluded if the 90% CIs for the GMRs of all 3 primary pharmacokinetic parameters (C_max, AUC_0‐t, and AUC_0‐∞) fell entirely within the predefined acceptance range of 80%‐125%.

Safety Assessment

All randomized subjects who received at least 1 dose of the investigational product were included in the safety analysis set. Adverse events (AEs) were coded using the Medical Dictionary for Regulatory Activities (Version 25.0 or higher) and summarized by incidence rates stratified by treatment group (test vs. reference formulations). All reported AEs were systematically tabulated with corresponding severity grades and investigator‐assessed causality relationships. Pre‐ versus posttreatment clinical evaluations included comparative analyses of vital signs, physical examinations, dermal reactions, and electrocardiogram findings. Laboratory assessments encompassed hematology, serum chemistry, and urinalysis parameters. Clinically significant deviations from baseline measurements were documented with detailed shift analyses.

Results

The study enrolled 32 subjects, comprising 24 men (75%) and 8 women (25%), with an age range of 21‐50 years. Thirty‐one participants completed the trial per protocol, while 1 subject discontinued participation due to nonadministration of the test formulation during Period 2. This discontinuation did not affect subsequent pharmacokinetic analyses as the withdrawal occurred before critical end point assessments.

Pharmacokinetics

Pharmacokinetic parameters, including time to maximum concentration, C_max, AUC_0‐t, AUC_0‐∞, and terminal elimination half‐life, were calculated using noncompartmental analysis implemented in SAS Version 9.4 for both test and reference crisaborole ointment formulations. Figure 1 illustrates the mean plasma concentration‐time profiles following application of the test and reference formulations, while Figure 2 presents the corresponding semilogarithmic plots. Comprehensive pharmacokinetic parameter estimates are summarized in Table 1, demonstrating comparable systemic exposure characteristics between the 2 formulations.

Figure 1.

The average drug‐time curve of crisaborole in subjects after application of the test formulation (n = 31) and the reference formulation (n = 32). T indicates test formulation; R, reference formulation.

Figure 2.

The average concentration‐time semilogarithmic curves of crisaborole in subjects after application of the test formulation (n = 31) and the reference formulation (n = 32). T indicates test formulation; R, reference formulation.

Table 1.

Main Pharmacokinetic Parameters of Crisaborole in Subjects After Application of the Test Formulation (n = 31) and the Reference Formulation (n = 32)

	Test formulation		Reference formulation
Pharmacokinetic parameters (unit)	N	Mean ± SD (CV%)	N	Mean ± SD (CV%)
Cmax (ng/mL)	31	41.5 ± 15.0 (36.0)	32	40.5 ± 16.5 (40.8)
AUC0‐t (ng•h/mL)	31	394 ± 102 (25.9)	32	391 ± 123 (31.4)
AUC0‐∞ (ng•h/mL)	31	396 ± 102 (25.7)	32	394 ± 123 (31.2)
tmax (hour) a	31	5.50 (2.00, 10.0)	32	8.00 (2.00, 12.0)
t1/2 (hour)	31	4.34 ± 1.07 (24.7)	32	4.47 ± 1.05 (23.5)

AUC_0‐∞, area under the concentration‐time curve from time 0 to infinity; AUC_0‐t, area under the concentration‐time curve over the dosing interval; C_max, maximum concentration; CV, coefficient of variation; SD, standard deviation; t_1/2, terminal elimination half‐life; t_max, time to maximum concentration.

^a T_max (hour) is expressed as median (minimum, maximum).

The test and reference formulations demonstrated comparable pharmacokinetic profiles, with GMRs and corresponding 90% CIs for key parameters as follows: C_max (97.5%‐110%), AUC_0‐t (96.6%‐109%), and AUC_0‐∞ (96.6%‐109%). All 90% CIs for these primary bioequivalence metrics remained entirely within the predefined 80%‐125% equivalence range. These pharmacokinetic comparisons confirm the bioequivalence between the test formulation and reference crisaborole ointment under the study conditions, satisfying current regulatory requirements for topical drug product equivalence assessments. The bioequivalence assessment outcomes for crisaborole are summarized in Table 2.

Table 2.

Average Bioequivalence Evaluation Results of Crisaborole

	Test		Reference		Test/reference
Parameters (unit)	N	GM	N	GM	GMR (%)	90% CI (%)	Intrasubject CV (%)
Cmax (ng/mL)	31	39.0	32	37.7	103	97.5‐110	13.7
AUC0‐t (ng•h/mL)	31	380	32	371	103	96.6‐109	14.0
AUC0‐∞ (ng•h/mL)	31	381	32	373	103	96.6‐109	13.8

AUC_0‐∞, area under the concentration‐time curve from time 0 to infinity; AUC_0‐t, area under the concentration‐time curve over the dosing interval; C_max, maximum concentration; CV, coefficient of variation; GM, geometric mean; GMR, geometric mean ratio.

Safety

The safety analysis set comprised 32 enrolled subjects, with 8 AEs reported across 5 participants (15.6% incidence rate). The observed AEs consisted of decreased neutrophil count (2 events in 2 subjects) and elevated urinary leukocytes (2 events in 2 subjects), collectively representing 50% of all reported AEs. No serious AEs or treatment discontinuations occurred. AE severity grading revealed 7 Grade 1 events in 4 subjects and 1 Grade 2 event in a single participant. All events were assessed as “unlikely related” to study medications per investigator causality evaluation. Notably, no cutaneous reactions—including pain (burning/stinging sensations), pruritus, edema, erythema, contact urticaria, or allergic contact dermatitis—were observed following application of either test or reference formulations. This safety profile demonstrates favorable tolerability of both crisaborole ointment formulations under clinical investigation conditions.

Discussion

For optimized disease monitoring and therapeutic safety, a step‐down therapeutic approach is recommended across acute, subacute, and chronic cases. Posttreatment follow‐up evaluations should guide therapeutic modifications, including (1) transition to lower‐potency corticosteroids upon significant lesion resolution and pruritus alleviation, (2) reduction in corticosteroid application frequency, or (3) implementation of nonsteroidal maintenance therapy. Crisaborole ointment, as a steroid‐free phosphodiesterase‐4 inhibitor, has emerged as a prominent candidate for such step‐down regimens due to its extensive clinical adoption and substantial market potential, particularly evidenced by its current status as a priority target for generic drug development in dermatological therapeutics.

Current guidance for topical generic drug development acknowledges that while full pharmaceutical equivalence in excipient composition may not always be achievable, clinical equivalence studies remain mandatory when quality equivalence is established. The demonstration of bioequivalence in human trials necessitates rigorous control of critical parameters, including application site topography, administration technique uniformity, dose quantification methodology, and comprehensive adherence monitoring protocols. These operational considerations are particularly crucial for topical formulations where localized drug delivery and patient compliance significantly influence therapeutic outcomes.

In this study, the drug administration protocol was carefully designed to account for substantial drug exposure levels and interindividual variations in body surface area. A standardized dose of 4 g (equivalent to 5 mg/cm²) was topically applied over an 800‐cm² demarcated area on subjects' dorsal skin surfaces. Preadministration skin preparation involved standardized cleansing procedures at the application site to minimize sebum interference with drug absorption. Before dosing operations, all trial drug aliquots underwent precise gravimetric measurement and documentation. The drug was portioned using analytical balances with tare function to ensure measurement accuracy, with strict protocols implemented to prevent pharmaceutical contamination or mechanical loss between weighing and application. Administration procedures mandated cross‐verification of 3 critical identifiers: dosing spatula code, glove serial number, and subject ID. Operators donned double‐layer gloves with preweighed and numbered gloves worn as the outer layer. Using a standardized digital spreading technique (employing index, middle, and ring finger pads), operators systematically distributed the measured drug quantity across the designated area. The 4‐minute administration protocol required continuous circular spreading motions until achieving uniform transparent film formation without visible accumulation. Precise temporal documentation included both initiation and completion time stamps. Postadministration protocols involved inverted removal of contaminated outer gloves and collection of all drug‐contact materials (spatulas and gloves) for residual weight quantification. Final delivered doses were calculated through gravimetric differential analysis between initial drug aliquots and postapplication residues, with all mass measurements recorded in controlled laboratory records.

This investigation prioritized comprehensive participant protection measures beyond pharmaceutical control, with particular emphasis on safeguarding volunteer rights and ensuring protocol compliance. Sex‐segregated administration and observation environments were strictly maintained throughout the study period. Before trial commencement, all participants underwent comprehensive pretrial counseling sessions addressing both procedural details and emotional considerations.

A Phase I randomized, single‐center, vehicle‐controlled, parallel‐group study evaluated the safety, tolerability, and pharmacokinetic profile of single and multiple crisaborole ointment applications in patients with AD (Cohort 2). Participants were randomized in a 5:1 ratio to receive either crisaborole ointment or vehicle control, administered twice daily via nonocclusive application over 8 consecutive days. Safety outcomes demonstrated treatment‐emergent adverse events (TEAEs) in 9 patients (90%) from the crisaborole group compared with 2 patients (100%) in the vehicle group. No severe TEAEs, treatment‐related serious AEs, or dose adjustments/interruptions were documented in either treatment arm. The crisaborole cohort reported 13 TEAE instances, predominantly application site irritation (7 events) and application site pain (4 events). Patient‐reported outcomes characterized application site irritation as generating milder discomfort relative to pain perception. All TEAEs in the active treatment group were classified as mild severity. Comprehensive monitoring revealed no clinically significant abnormalities in laboratory parameters, vital signs, or electrocardiographic findings across both groups, with no mortality events recorded. ³ In contrast, the current investigation documented 8 AEs, primarily consisting of minor laboratory deviations. Notably absent were protocol‐defined AEs of special interest, including pain‐related manifestations (burning/stinging sensations), pruritus, edema, erythema, contact urticaria, or allergic contact dermatitis. Given this study's design as a single‐dose bioequivalence trial in healthy volunteers with limited treatment duration, the safety database remains constrained. Postmarketing surveillance following regulatory approval will facilitate the comprehensive characterization of the product's safety profile through large‐scale pharmacovigilance data collection.

Conclusion

In this study, the tested formulation of crisaborole ointment is bioequivalent to the reference formulation, and the safety is comparable. This comprehensive evaluation demonstrates that the successful implementation of this bioequivalence study for topical dermatological formulations holds significant implications amid the expanding landscape of cutaneous bioequivalence trials. The established protocol and outcomes provide valuable methodological references for subsequent generic drug development initiatives, while simultaneously offering clinicians and patients evidence‐based therapeutic alternatives through rigorous demonstration of pharmaceutical equivalence. These findings contribute to the progressive refinement of regulatory standards for topical product evaluation, ultimately enhancing patient access to safe and effective dermatological treatments.

Author Contributions

All authors contributed and reviewed all content of this manuscript.

Conflicts of Interest

The authors declare no potential conflicts of interest.

Funding

This study was funded by Qilu Pharmaceutical Co., Ltd, and S&T Program of Langfang (2023013234).

Data Availability Statement

The data sets generated or analyzed during this study are available from the corresponding author on reasonable request.