This phase 1, single-center, open-label, randomized, two-period, two-sequence, crossover trial conducted in China between 28th February 2023 and 19th May 2023 compared the PK and safety of bisoprolol 5-mg tablets manufactured in China with those of bisoprolol 5-mg tablets manufactured in Germany. The study duration was approximately 4 weeks and consisted of a 2-week screening period, a 3-day dosing/sampling period, a washout period of at least 7 days, and a 3-day second dosing/sampling period, and the study ended on day 10 of the second dosing period.

Healthy participants were allocated into either the fasted or the fed group, with each sex representing at least one-quarter of the participants in each group, and assigned a computer-generated randomization number. Allocated participants were then randomly assigned in a 1:1 ratio to Sequence A or Sequence B. Participants in Sequence A received the test product {one bisoprolol 5-mg tablet manufactured in China [Merck Pharmaceutical Manufacturing (Jiangsu) Co., Ltd., Nantong, China]} first and the reference product [one bisoprolol 5-mg tablet manufactured in Germany (Merck Healthcare KGaA, Darmstadt, Germany)] next, and those in Sequence B received the reference product first and the test product next (Figure 1).

The fasted group refrained from all food and drink, except water, from the evening (after dinner) of the day before the intervention for a total fasting time of at least 10 h. Water was allowed until 1 h before receiving the intervention, and then it was allowed again from 2 h after the intervention. The fed group consumed a standard high-fat, high-calorie breakfast, as per the United States Food and Drug Administration guidelines (U.S. Food and Drug Administration, 2002), within 30 min prior to receiving the intervention.

The study was conducted in compliance with the ethical principles outlined in the Declaration of Helsinki, the International Council on Harmonization Guidelines for Good Clinical Practice Council, and all applicable local laws and regulations. The study protocol and other relevant documents were approved by the Beijing Friendship Hospital Institutional Review Board and Ethics Committee (approval number: 2022-P1-drug-071-01). Participants provided written informed consent, and the clinical trial was registered in the Chinese Clinical Trials Register under the identifier CTR20230391.

Eligible participants were aged ≥18 years, considered healthy by medical evaluation (including medical history), had a body weight between 50 and 90 kg, and had a body mass index (BMI) between 19 and 26 kg/m² (inclusive). The key exclusion criteria were positive screening for hepatitis B-soluble antigens or hepatitis C antibodies, human immunodeficiency virus, or Treponema pallidum antibodies; abnormal chest x-ray or computed tomography finding at screening; blood donation ≥400 mL or significant blood loss within 90 days prior to the administration of the first intervention; non-acceptance of the study’s high-fat breakfast; receipt of prescription or non-prescription medication within 28 days before the administration of the first intervention (including multivitamins and herbal products); participation in another clinical trial within 90 days prior to the first intervention administration; and any condition that was considered an inappropriate risk or contraindication for participation in the study.

The primary endpoints were the maximum observed concentration (C_max), the area under the plasma concentration–time curve (AUC) from time 0 to the time of the last measurable concentration (AUC_0–tlast), and the AUC from time 0 to infinity (AUC_{0– ∞} ). PK parameters were evaluated using 3-mL blood samples that were collected via an indwelling cannula or direct venipuncture prior to the intervention and at 0.5, 1, 1.5, 2, 3, 4, 6, 8, 10, 12, 15, 24, 30, 36, and 48 h after the intervention. Blood samples were centrifuged at 4°C for 10 min at 2,000 g to separate the plasma supernatant, and the plasma was divided into two aliquots and stored at −20°C or lower.

The secondary endpoints included the time of the maximum observed concentration (t_max), elimination half-life (t_1/2), AUC from the time of the last quantifiable concentration extrapolated to infinity as a percentage (%AUC_extra), apparent terminal elimination rate constant (λ_z), apparent plasma clearance of the drug following extravascular administration (CL/F), and apparent volume of distribution during the terminal phase following extravascular administration (V_z/F).

Baseline characteristics included weight, height, BMI, body surface area, alcohol consumption, nicotine use, drug screening, ethanol test, and viral serology. The dissolution profiles of the two products were determined using high-performance liquid chromatography, and the peak areas in the chromatogram were identified using the external standard method. The dissolution medium was sodium chloride buffer prepared in accordance with the European Pharmacopeia recommendations on dissolution testing (Ph. Eur. 5.17.1.), adjusted to pH 1.2, pH 4.5, and pH 6.8. Samples were evaluated using ultraviolet spectroscopic analysis at 5, 10, 15, 30, and 45 min.

Plasma concentrations of bisoprolol were analyzed using a fully validated liquid chromatography–tandem mass spectrometry (LC–MS/MS) method at Labcorp (Shanghai, China). Briefly, bisoprolol and its deuterium-labeled internal standard (bisoprolol-d5) were isolated from plasma (50 μL) using a 96-well plate following a protein precipitation extraction procedure with acetonitrile. The chromatographic separation was achieved using an ACQIUITY UPLC BEH C18 column (2.1 mm × 50 mm, 1.7 μm particle size; Waters Corporation) at a flow rate of 0.5 mL/min with a 3-min gradient profile, where the HPLC run began with the mobile phase B at 40% for the first 0.1 min, then increased to 95% at 1.5 min, held at 95% B until 2 min, and then returned to the starting composition at 2.1 min for column re-equilibration. Mobile phase A was 20 mM ammonium acetate, and mobile phase B was methanol, both containing 0.1% formic acid. The analytes were detected using a triple–quadruple mass detector (Sciex API 5500) in the positive ion mode, with turbo ion spray in multiple-reaction monitoring (m/z 326.2 → 116.2 for bisoprolol and m/z 331.3 → 121.1 for bisoprolol-d5). The method showed good linearity in the range of 0.500–75 ng/mL, with a lower limit of quantification set at 0.5 ng/mL. The intra- and inter-assay precision values (≤4.4% and ≤4.8%, respectively) met the acceptance criteria as per the regulatory guidelines. A battery of stability studies (i.e., bench-top, freeze–thaw, and long-term stability) were performed, and bisoprolol was stable in plasma for 24 h at room temperature and for 135 days at −30∼−10°C after five freeze/thaw cycles. No matrix effect was observed within the linearity range. The assay method was found to be highly reproducible and robust.

The type, frequency, and severity of adverse events (AEs), including treatment-emergent AEs (TEAEs), and the relationship of AEs to the study intervention were recorded. AEs were coded using the Medical Dictionary for Regulatory Activities version 25.1, and TEAEs were assigned to a system organ class and preferred term. The additional safety assessments included vital signs (systolic and diastolic blood pressure, pulse rate, body temperature, and respiration rate), clinical laboratory tests (biochemistry, hematology, and urinalysis), 12-lead electrocardiogram results, and physical examinations.

The sample size was determined from the bisoprolol intra-individual variability observed in a previous Merck bioequivalence study (EMR200006-001) that compared a fixed-dose-combination 5 mg/5 mg bisoprolol–amlodipine tablet with a 5-mg bisoprolol tablet co-administered with a 5-mg amlodipine tablet under fasted and fed conditions (Hu et al., 2018). In that study, the intra-individual coefficients of bisoprolol for AUC_0–tlast, AUC_{0– ∞} , and C_max were, respectively, 10.4%, 10.6%, and 9.4% under fasted conditions and 13.5%, 13.4%, and 15.0% under fed conditions. Thus, considering the bioequivalence criteria for AUC_0–tlast, AUC_{0– ∞} , and C_max (0.80–1.25) and allowing the test/reference ratio to vary between 0.95 and 1.05, a total of 16 participants in the fasted group and 18 in the fed group were considered sufficient to provide 95% power to confirm bioequivalence in each group and 90% power to confirm bioequivalence in both groups. Accounting for an attrition rate of 25%, a total sample size of 46 participants was planned, with 22 participants in the fasted group and 24 participants in the fed group.

The full analysis set (FAS) included all randomized participants. The safety analysis set (SAF) included all participants who had received at least one dose of the study product. The PK parameter analysis set (PKPS) included all participants who had received at least one dose of the study product and had no clinically important protocol deviations or events that would affect the PK analysis. The bioequivalence analysis set (BES) included all participants who had at least one evaluable PK parameter in at least one treatment period and who did not have any relevant protocol deviations, had no factors likely to affect the comparability of PK results, and had adequate study intervention compliance.

Continuous variables were summarized using descriptive statistics [including the number, arithmetic mean, median, standard deviation (SD), and minimum and maximum values], and quantitative variables were summarized by numbers and percentages. Geometric least-squares (LS) mean, coefficient of variation (CV), intra-participant CV percentage, and confidence intervals (CIs) were also calculated. No statistical comparisons were performed on the participant demographics or baseline characteristics.

Analysis of the primary endpoints was conducted in the BES. The data were log-transformed, and a mixed-effects model was applied, including fixed effects for sequence, treatment, and period. Differences between treatments on the log scale were estimated for the parameters, together with their 90% CIs. The geometric LS means and their 95% CIs were also estimated. The bioequivalence of the two tablets was assessed separately for the fasted and fed groups, and bioequivalence was only established if the 90% CIs for the geometric mean ratios for both primary endpoints between the two interventions were within 0.80–1.25 in both the fasted and fed groups. For t_max, Hodges–Lehmann estimates and the corresponding 90% CIs were calculated according to the Tukey method. The 90% CIs of the geometric mean of the test product and reference product were calculated for all other secondary endpoints, but no formal statistical comparisons were performed. The dissolution profiles were characterized by plotting the dissolution rate over time, and the profiles of the two tablets were compared using a model-independent approach (U.S. Food and Drug Administration, 1997).

TEAEs with missing data regarding the association with the study intervention were considered related to the study intervention. No other imputation for missing values was conducted, and outliers were defined as values outside 1.5 × the interquartile range above the upper quartile or below the lower quartile.

PK parameters were derived using Phoenix^® WinNonlin^® version 8.3 (Certara, L.P., Princeton, NJ, United States), and all other statistical analyses were conducted using SAS^® version 9.4 or higher (SAS Institute, Cary, NC, United States).

The participant disposition is shown in Figure 1. A total of 161 adults were screened, and 46 eligible participants were allocated to the fasted group (n = 22) or the fed group (n = 24) and randomized to Sequence A or Sequence B. Two participants in the fasted group experienced AEs prior to receiving the intervention and discontinued the study; these participants were not included in the SAF, PKPS, or BES. The baseline demographics are shown in Table 1. All participants were Chinese, and 69.6% were men. The median age was 32.0 years, the mean (SD) body surface area was 1.73 (0.130) m², the mean (SD) weight was 63.92 (7.219) kg, and the mean (SD) BMI was 22.39 (1.811) kg/m². None of the participants had a history of alcohol consumption or nicotine use.

The PK parameters are summarized in Table 2, and the plasma concentration–time profiles are shown in Figure 2. Following a single dose of the test product and the reference product, the absorption and elimination of bisoprolol were very similar in both the fasted and fed groups. Bisoprolol reached a peak concentration at a median of 1.75 h (test product) and 1.50 h (reference product) in the fasted group and 2.00 h (both test and reference products) in the fed group. In both the fasted and fed groups, both products were eliminated rapidly and plasma levels were below the lower limit of detection after 48 h.

In the fasted group, the geometric mean (geometric CV percentage) C_max was 21.2 (15.0) ng/mL for the test product and 22.1 (17.0) ng/mL for the reference product. In the fed group, the geometric mean (geometric CV percentage) C_max was 22.7 (18.8) ng/mL for the test product and 22.8 (15.2) ng/mL for the reference product. Both AUC_0–tlast and AUC_{0– ∞} of bisoprolol were similar for the test and reference products under both fasted and fed conditions. The mean t_1/2, V_z/F, and CL/F were also similar for both products. Boxplots of C_max, AUC_0–tlast, and AUC_{0– ∞} in both the fasted and fed groups are shown in Figure 3. PK parameters were similar between the test and reference products and between both the fasted and fed groups.

The bioequivalence analysis results are shown in Table 3. Among the fasted group, the geometric LS mean ratio (90% CI) for the test product/reference product was 0.9565 (0.9006–1.0158) ng/mL for C_max, 0.9761 (0.9370–1.0168) h·ng/mL for AUC_0–tlast, and 0.9807 (0.9429–1.0200) h·ng/mL for AUC_{0– ∞} . As all CIs were within the range of 0.80–1.25, this met the pre-defined criteria for bioequivalence. In the fed group, the geometric LS mean ratio (90% CI) for the test product/reference product was 0.9966 (0.9289–1.0691) ng/mL for C_max, 0.9672 (0.9220–1.0145) h·ng/mL for AUC_0–tlast, and 0.9693 (0.9253–1.0155) h·ng/mL for AUC_{0– ∞} . This also met the pre-defined criteria for bioequivalence.

The dissolution assay results of the test and reference products are shown in Figure 4. At all pH conditions (pH 1.2, pH 4.5, and pH 6.8), the mean dissolution rates for both the test and reference products were higher than 85% at 15 min.

The TEAE data are tabulated in Table 4. A higher frequency of TEAEs occurred following the administration of the reference product (10/44 [22.7%]) than after the test product (4/44 [9.1%]). Three participants in the fed group experienced TEAEs considered to be related to the study treatment (two product-related first-degree atrioventricular block events and one product-related diastolic blood pressure decrease). Both first-degree atrioventricular block events were experienced by one participant, once with the test and once with the reference product, and the diastolic blood pressure decrease event occurred with the reference product. All product-related TEAEs were mild in severity and resolved. None of the participants discontinued because of TEAEs, and no serious TEAEs or deaths occurred.