Both studies were conducted in accordance with Good Clinical Practice, the Declaration of Helsinki, and applicable national regulations. The protocols were reviewed and approved by the independent ethics committee of Beijing Shijitan Hospital, Capital Medical University. The non-block trial received ethical approval on November 13, 2019 (Approval No.: 2019 Lun Shen No. 116); the charcoal-block trial obtained ethical approval on April 29, 2020 (Approval No.: 2020 Lun Shen No. 13).

All participants provided written informed consent before any study-specific procedure and could withdraw at any time without prejudice.

The two trials were registered at the Drug Clinical Trial Registration and Information Publicity Platform of the National Medical Products Administration of China under identifiers CTR20200344 (first participant enrolled on December 23, 2019) and CTR20201473 (first participant enrolled on July 20, 2020). The charcoal-block trial was additionally entered in the Chinese Clinical Trial Registry under identifier ChiCTR2000035173 on August 2, 2020, as a supplementary public registration record. Accordingly, although the supplementary ChiCTR entry postdated first participant enrollment in the charcoal-block study, the study had already been registered in the mandatory NMPA drug-trial registry before enrollment. The clinical trial registration identifiers and public registry URLs are reported in the Abstract in accordance with journal requirements. Trial reporting was aligned with the Consolidated Standards of Reporting Trials (CONSORT) principles for randomized clinical trials, and participant disposition is summarized in Figure 1.

Two open-label, randomized, single-dose, four-period crossover PK-BE studies were performed under fasting conditions, one without activated charcoal and one with activated charcoal blockade. A 5-day washout interval separated consecutive study periods.

Sample size was estimated on the basis of pilot data and published variability estimates for inhaled budesonide [12]. In our pilot work, the intra-individual variation (IIV) for Cmax and AUC was approximately 25.0% and 19.0%, respectively. Sample size was determined based on the primary PK endpoints C_max and AUC, assuming α = 0.05, β = 0.20, IIV = 25%, assumed T/R mean ratios of 90%–111%, and BE limits of 80.00%–125.00%. Under these assumptions, 28 evaluable subjects were required; to allow for potential dropout, 36 subjects were enrolled in each study.

The investigational products were two formulations of BIS. The test product was manufactured by Sichuan Purity Pharmaceutical Co., Ltd. (2 mL:1 mg per ampoule; batch No. CP024-201902; manufacture date May 19, 2019; expiry date May 18, 2021).

The reference product was Pulmicort Respules (AstraZeneca Pty Ltd; 2 mL:1 mg per ampoule). Batch No. 325707 was used in the charcoal-block study and batch No. 324560 in the non-block study, with expiry dates in September 2021 and February 2021, respectively.

Drug administration was performed with a PARI LC Jet Plus nebulizer connected to a PARI Master compressor. Before each period, all subjects received standardized nebulization training. During dosing, subjects sat upright, wore nasal clips to prevent nasal breathing, and inhaled slowly and steadily through a mouthpiece while following a standardized audio guide lasting approximately 6 min.

Eligible participants were healthy Chinese adults aged 18–65 years with a body mass index of 19.0–26.0 kg/m². Eligibility was determined by medical history, physical examination, vital signs, lung function testing, 12-lead electrocardiography, and clinical laboratory evaluations. Women of childbearing potential were required to have a negative pregnancy test and to use reliable contraception; male participants were also required to use contraception during the study and for 3 months thereafter. Subjects with asthma, airway hyperresponsiveness, or oral conditions such as ulcers were excluded.

A total of 72 subjects were enrolled, with 36 subjects assigned to each study. Within each study, subjects were randomized in a 1:1 ratio to one of two treatment sequences (T-R-T-R or R-T-R-T), where T denotes the test formulation and R denotes the reference formulation. Randomization was generated using SAS version 9.4.

Subjects were admitted to the study center 2 days before dosing for nebulization training. Only those who completed training successfully were dosed. After an overnight fast of at least 10 h, each subject inhaled a single 1 mg dose of either T or R BIS via the study nebulizer. Nebulization continued until 6 min after sputtering was first heard.

In the charcoal-block study, activated charcoal (3 g suspended in 30 mL water) was administered 2 min before nebulization, immediately after nebulization, and 1 h after nebulization. Participants were instructed to gargle the suspension before swallowing to maximize adsorption of drug remaining in the oral cavity. All dosing procedures were conducted under the supervision of trained study personnel, with protocol-specific instruction before each dosing period. This procedure was designed to minimize gastrointestinal absorption and provide a lung-focused assessment of systemic exposure.

Safety monitoring after each dose included physical examination, electrocardiography, vital signs, clinical laboratory testing, adverse event recording, and documentation of concomitant medication. The washout interval between periods was 5 days. Adverse event severity was graded according to Common Terminology Criteria for Adverse Events, version 5.0.

Blood samples for PK analysis were collected at 1.5 h before administration and at 3, 5, 8, 10, 12, 16, 20, 25, 30, and 45 min and 1.0, 1.5, 2.0, 3.0, 6.0, 10.0, and 15.0 h after administration. The nominal sample volume at each time point was 4 mL.

Plasma samples were separated on a Waters Acquity UPLC BEH C₁₈ column (2.1 mm × 50 mm, 1.7 μm) maintained at 40 °C, with a Phenomenex UHPLC C₁₈ precolumn (2.1 mm internal diameter). The mobile phases were an acetic acid LC-MS solution in ultrapure water and methanol LC-MS grade solvent. The injection volume was 20 μL and the flow rate was 0.4 mL/min.

Quantification of budesonide and budesonide-d₈ was performed by multiple reaction monitoring with electrospray ionization in negative mode. Optimized source parameters were as follows: GAS1 40 psi, GAS2 50 psi, curtain gas 30 psi, collision gas 8 psi, ion spray voltage −4,500 V, and source temperature 450 °C. The monitored ion transitions were m/z 489.3→357.1 for budesonide and m/z 497.3→357.0 for budesonide-d₈. Declustering potential, collision energy, entrance potential, and collision cell exit potential were set according to validated analytical conditions.

Blood samples were centrifuged within 1 h of collection at 1,700 × g and 4 °C for 10 min to obtain plasma. The separated plasma was stored below −60 °C until analysis, within the stability window established during bioanalytical method validation.

Plasma budesonide concentrations were quantified using a validated LC-MS/MS assay in human K2-EDTA plasma. The calibration curve was linear over 10.00–2000 pg/mL, and the lower limit of quantification (LLOQ) was 10.00 pg/mL. During study-sample analysis, quality control (QC) concentrations were 30.00, 250.0, 800.0, and 1500 pg/mL, with a 6000 pg/mL dilution QC, and QC samples were distributed throughout each analytical run. QC precision ranged from 2.72% to 6.03% in the non-block study and from 1.99% to 3.98% in the charcoal-block study; corresponding QC bias ranged from −1.35% to 1.70% and from −1.07% to 7.25%, respectively. Plasma stability met acceptance criteria after four freeze-thaw cycles, 8 h on ice before processing, 72 h in the autosampler after processing, and 78 h of reinjection storage. Long-term plasma stability was demonstrated for 23 days at −30 °C to −10 °C and for 74 days at −50 °C to −30 °C and −80 °C to −60 °C. In the charcoal-block study, whole blood was additionally shown to be stable on ice for 2 h before plasma separation. The storage duration of clinical samples remained within these validated stability windows. Full validation data are provided in Supplementary Material S1.

PK parameters, including C_max, AUC_0-t, AUC_0-inf, terminal elimination half-life, and T_max, were calculated by noncompartmental analysis using Phoenix WinNonlin version 8.2. Actual blood sampling times were used for PK calculations. BE was evaluated in the bioequivalence set (BES). C_max, AUC_0-t, and AUC_0-inf were natural log-transformed and analyzed in SAS 9.4 using a mixed-effects model for the four-period fully replicated crossover design. Sequence, period, and treatment/formulation were included as fixed effects, and subjects were handled as the subject term nested within sequence in the replicated crossover framework. Geometric mean ratios and their 90% CIs were derived under the two one-sided testing approach; BE was concluded when the 90% confidence intervals (CIs) for C_max, AUC_0-t, and AUC_0-inf were entirely within 80.00%–125.00%. In the non-block study, one subject withdrew after completing periods 1 and 2; the evaluable data from these periods were retained in pharmacokinetic parameter set (PKPS)/BES without imputation.

For the exploratory between-cohort comparison, the two replicate administrations of each formulation were averaged within each participant. The resulting subject-level summaries were compared between the non-block and charcoal-block cohorts separately for the test and reference formulations using two-sided Welch unequal-variance tests on the arithmetic scale. T_max was summarized descriptively. No adjustment for multiplicity was performed; therefore, all P values from these comparisons are reported as nominal exploratory P values and were not used for confirmatory inference.

Overall, 210 subjects were screened for the non-block study and 188 for the charcoal-block study (Figure 1). A total of 72 subjects were enrolled across the two studies. In the non-block study, 36 subjects were enrolled and 35 completed all four periods. One subject (K033) withdrew after period 2 because of an adverse event unrelated to study treatment; because dosing and PK sampling for periods 1 and 2 had been completed, this subject contributed evaluable period-1 and period-2 data to pharmacokinetic concentration set (PKCS), PKPS, and BES. In the charcoal-block study, all 36 enrolled subjects completed all four periods.

Demographic characteristics are summarized in Table 1. In the charcoal-block study, the mean age was 33.36 ± 6.89 years, the mean body weight was 61.45 ± 8.16 kg, and the mean body mass index was 22.81 ± 1.78 kg/m². In the non-block study, the corresponding values were 37.17 ± 6.36 years, 63.79 ± 7.43 kg, and 23.71 ± 1.60 kg/m². Male subjects accounted for 44.4% of the charcoal-block cohort and 61.1% of the non-block cohort.

Detailed PK parameters are summarized in Table 2. Budesonide was absorbed rapidly under both study conditions, with a median T_max of approximately 0.20 h. Plasma concentrations then declined promptly, and the terminal half-life was approximately 3 h, consistent with rapid pulmonary absorption followed by systemic elimination (Figures 2, 3). In Figures 2B, 3B, the 0–3 h interval is presented on an expanded linear scale to facilitate visualization of the early absorption phase.

Table 3 presents an exploratory between-study comparison based on subject-level PK summaries derived from replicate administrations of the same formulation within each subject. Because the non-block and charcoal-block conditions were studied in separate cohorts, this analysis should be interpreted as exploratory only.

Median T_max values were similar between conditions for both formulations. For the test formulation, mean C_max was 1142.76 ± 393.10 pg/mL under non-block conditions and 1126.53 ± 406.71 pg/mL under charcoal-block conditions, whereas mean AUC_0-t values were 2,135.20 ± 598.52 and 1927.92 ± 543.73 h·pg/mL and mean AUC_0-inf values were 2,224.24 ± 613.06 and 2000.58 ± 550.29 h·pg/mL, respectively. On the arithmetic scale, the nominal exploratory P values were 0.864 for C_max, 0.129 for AUC_0-t, and 0.108 for AUC_0-inf, with mean differences (charcoal-block minus non-block) of −16.23, −207.28, and −223.67, respectively. For the reference formulation, the corresponding nominal exploratory P values were 0.720 for C_max, 0.456 for AUC_0-t, and 0.432 for AUC_0-inf, with mean differences of 41.41, −113.08, and −122.40, respectively. Thus, activated charcoal was associated with numerically lower AUC values, whereas the between-condition differences in C_max, AUC_0-t, and AUC_0-inf were not statistically significant for either formulation. The shorter t_1/2 observed for the test formulation under charcoal-block conditions (nominal P = 0.016) should be considered descriptive because no multiplicity adjustment was applied.

In the non-block study, all 36 randomized subjects were included in the BE set, although subject K033 contributed data only from the first two periods because of withdrawal after period 2. In the charcoal-block study, all 36 randomized subjects completed the BE evaluation.

As shown in Tables 4, 5, the 90% CIs for the geometric mean ratios of C_max, AUC_0-t, and AUC_0-inf all fell within the accepted BE range of 80.00%–125.00% under both non-block and charcoal-block conditions.

Analysis of log-transformed primary PK parameters (Table 6) showed no significant sequence effect in either study. In the non-block study, no significant formulation or period effects were identified. In the charcoal-block study, period effects were observed for C_max, AUC_0-t, and AUC_0-inf, and formulation effects were observed for AUC_0-t and AUC_0-inf. These analysis-of-variance findings did not alter the BE conclusion because the predefined decision criterion was based on the 90% CIs for the geometric mean ratios, all of which remained within the acceptance interval.

Median T_max values were similar between the test and reference formulations in both studies. Signed-rank testing showed no statistically significant between-formulation difference in T_max.

Safety analyses included all 72 subjects. Treatment-emergent adverse events are summarized in Tables 7, 8.

In the non-block study, 12 subjects (33.33%) reported 20 adverse events, and 10 subjects (27.78%) experienced 14 treatment-emergent adverse events. One subject withdrew because of a grade 2 upper respiratory tract infection; all remaining adverse events were grade 1, and no serious adverse events occurred.

In the charcoal-block study, 11 subjects (30.56%) reported 13 adverse events, and 9 subjects (25.00%) experienced 11 treatment-emergent adverse events. One subject experienced grade 2 chest tightness approximately 2 h after the last dose; this symptom resolved quickly following administration of nasal cannula oxygen therapy; all other adverse events were grade 1. No serious adverse events or adverse-event-related withdrawals occurred, and all events had resolved by the final visit.

The incidence and pattern of adverse events were similar between the test and reference formulations, and the observed safety profile was consistent with the known tolerability of inhaled budesonide [7].