Clinical Therapeutics/Volume 32, Number 3, 2010

Pharmacokinetic Properties and Bioequivalence of Two

Compound Formulations of 1500 mg Ampicillin (1167 mg)/
Probenecid (333 mg): A Randomized-Sequence, Single-Dose,
Open-Label, Two-Period Crossover Study in Healthy Chinese
Male Volunteers
Huizhe Wu, MD1; Mingyan Liu, MD1; Shuang Wang, MD1; Wanyu Feng, MD, PhD2;
Weifan Yao, BS1; Haishan Zhao, BS1; and Minjie Wei, MD, PhD1
1Department of Pharmacology, School of Pharmaceutical Sciences, China Medical University, Shenyang,
People’s Republic of China; and 2Department of Clinical Pharmacology, The First Affiliated Hospital of
China Medical University, Shenyang, People’s Republic of China

ABSTRACT ratios of the log-transformed Cmax and AUC values

Background: Ampicillin/probenecid is an antimi- were within the equivalence range (80%–125%) pre-
crobial formulation indicated for the treatment of re- determined by the State Food and Drug Administration
spiratory, urinary tract, and gastrointestinal infections. (SFDA) of the People’s Republic of China. Tolerability
Ampicillin sodium is the active antimicrobial ingredi- was based on the observation of adverse events (AEs),
ent that can act on the phase of bacterial breeding and monitoring of vital signs (blood pressure, heart rate,
inhibit the biosynthesis of bacterial mucopeptide in temperature, electrocardiography) and laboratory
the cell wall. Probenecid acts synergistically by com- tests (hematology, blood biochemistry, hepatic func-
petitively inhibiting an organic anion transporter in tion, urinalysis), and subject’s interview on AEs.
renal tubules, increasing the plasma concentrations, Results: The study was performed in 20 healthy
and thus extending the plasma elimination t1/2. Chinese male volunteers (mean [SD] age, 21.4 [2.2] years;
Objective: The aim of this study was to assess and weight, 64.1 [5.5] kg; height, 173.7 [5.3] cm; and
compare the pharmacokinetic (PK) properties, bio- body mass index, 21.2 [1.6] kg/m2). The mean (SD)
availability, and bioequivalence of a newly developed Cmax, Tmax, AUC0–24, and AUC0–∞ after administration
dispersible tablet formulation (test) of ampicillin/ of the test and reference formulations, respectively,
probenecid with those of an established branded cap- were as follows: ampicillin, Cmax, 13.45 (3.43) versus
sule formulation (reference) in healthy Chinese male 15.04 (5.68) µg/mL, Tmax, 1.58 (0.49) versus 1.78
volunteers. (0.55) hours, AUC0–24, 50.78 (13.39) versus 57.44
Methods: A randomized-sequence, single-dose, (17.27) µg/mL/h, and AUC0–∞, 51.95 (13.45) versus
open-label, 2-period crossover study was conducted in 58.74 (17.19) µg/mL/h; probenecid, Cmax, 15.56
fasted healthy Chinese male volunteers. Eligible par- (2.94) versus 16.01 (2.88) µg/mL, Tmax, 2.85 (0.78)
ticipants were randomly assigned in a 1:1 ratio to re- versus 3.30 (1.51) hours, AUC0–24, 129.23 (27.59)
ceive 6 dispersible tablets (test) or branded capsules versus 127.29 (26.89) µg/mL/h, and AUC0–∞, 133.85
(reference) (1500 mg total; 250 mg each contain- (28.80) versus 131.21 (28.25) µg/mL/h. On ANOVA,
ing ampicillin 194.5 mg and probenecid 55.5 mg), neither period nor sequence effects were observed for
followed by a 7-day washout period and admin- any of the PK properties. The 90% CIs of ampicillin
istration of the alternate formulation. Plasma samples for the log-transformed ratios of Cmax, AUC0–24, and
were collected over a 24-hour period following admin-
istration and analyzed for ampicillin and probenecid
Accepted for publication February 5, 2010.
content by HPLC. PK parameters such as Cmax, AUC0–t, doi:10.1016/j.clinthera.2010.03.017
and AUC0–∞ were also determined. The formulations 0149-2918/$ - see front matter
were considered bioequivalent if the geometric mean © 2010 Excerpta Medica Inc. All rights reserved.

March 2010 597

 Clinical Therapeutics

AUC0–∞ were 86.5% to 108.0%, 96.7% to 107.8%, and neously in the mouth to be swallowed without the aid
83.3% to 100.7%, respectively, and the corresponding of water and a branded capsule formulation (refer-
values for probenecid were 90.2% to 108.3%, 96.8% to ence) in fasted healthy Chinese male volunteers. The
107.8%, and 97.2% to 108.5%. No AEs were observed study was conducted to satisfy the People’s Republic
or reported up to 1 week after study end. of China State Food and Drug Administration (SFDA)
Conclusions: In this small study in healthy Chinese guidelines requirement for marketing of a generic
male volunteers, a single 1500-mg dose of the dispers- formulation.
ible tablet formulation (test) of ampicillin/probenecid
met the SFDA’s regulatory criteria for bioequivalence SUBJECTS AND METHODS
to the reference capsule formulation based on the rate Formulations and Subject Selection
and extent of absorption. Both formulations were well Ampicillin/probenecid dispersible tablets (250 mg/
tolerated. (Clin Ther. 2010;32:597–606) © 2010 Ex- tablet: ampicillin, 194.5 mg and probenicid, 55.5 mg;
cerpta Medica Inc. lot no. 20050801; expiration date, September 2007),
Key words: ampicillin, probenecid, bioequivalence, manufactured and provided by Shenyang Shidte Phar-
pharmacokinetics, HPLC. maceutical Co. Ltd. (Shenyang, People’s Republic of
China), and ampicillin/probenecid capsules* (250 mg/
capsule ampicillin: 194.5 mg and probenicid, 55.5 mg;
INTRODUCTION lot no. 050307; expiration date, March 2007), purchased
Ampicillin is an active β-lactam broad-spectrum oral from the manufacturers, were used as the test and refer-
antibiotic that is indicated for respiratory, urinary, and ence formulations, respectively, for assessing bioequiva-
gastrointestinal tract infections and meningitis. It has lence. Healthy Chinese male volunteers aged 20 to
been used in the treatment of pneumonia, intestinal 30 years old, weight within 20% of ideal, and with
infections, urinary tract infections, postoperative in- body mass index (BMI) between 20 and 24 kg/m2
fection in the soft tissue, otitis media, Haemophilus were eligible for this study. Additional inclusion crite-
influenzae, and other infections caused by susceptible ria were nonsmoking status and normal findings in
microorganisms.1,2 It inhibits bacterial multiplication the medical history. Volunteers were assessed by 12-lead
by preventing biosynthesis of bacterial mucopeptide ECG, physical examination (oral body temperature,
in the cell wall, which is necessary for structural integ- sitting at rest blood pressure, heart rate, pulse), and
rity and avoiding exposure to the external environ- laboratory tests (hematology, blood biochemistry, he-
ment. Ampicillin is absorbed orally, distributed through- patic function, urinalysis), and were tested for hepati-
out the body, and ~20% to 60% is excreted unchanged tis B surface antigen. Subjects who had a history or
through the kidneys with higher concentration in the evidence of a renal, gastrointestinal, hepatic, or hema-
bile and urine.3 tologic abnormality; any acute or chronic disease; or
In the compound formulation of ampicillin/ an allergy to any chemicals were excluded. All eligible
probenecid, ampicillin plays the principle antibiotic subjects were informed of the details and purpose of
role in the treatment of infections. Probenecid is a the study by the clinical investigators, and each pro-
benzoic acid derivative that acts synergistically by vided written informed consent prior to study partici-
competing with ampicillin for the same organic anion pation. Subjects were instructed to abide by the stan-
transporter in proximal renal tubules4,5; therefore, the dard protocol and to abstain from using any medication
blood concentration of ampicillin can be increased by at least 2 weeks before admission, and not to consume
~30% to 40%, and elimination t1/2 also can be pro- any alcohol or tobacco within 48 hours before or dur-
longed by ~1 to 1.5 hours.6,7 ing the course of the study.
The present study was designed to assess and com-
pare the pharmacokinetic (PK) parameters, bioavail- Study Design
ability, and bioequivalence of two 1500-mg formula- This randomized-sequence, single-dose, open-label,
tions of ampicillin (1167 mg)/probenecid (333 mg): 2-period crossover study was conducted between Sep-
a newly developed generic dispersible tablet formula-
tion (test) designed to disintegrate rapidly in water *Trademark: Aerodyn® (Changchun Gaipu Pharmaceutical
to form a stabilized suspension or disperse instanta- Co. Ltd., Changchun, People’s Republic of China).

598 Volume 32 Number 3

 H. Wu et al.

tember 2005 and March 2006 at the Clinical Practice was equipped with a dual plunger pump (model, LC-
Center of the First Affiliated Hospital of China Medical 10ATVP) and an ultraviolet-visible (UV-Vis) detector
University, Shenyang, People’s Republic of China (State (SPD-10AVP).
Food and Drug Administration [SFDA] registration Chromatographic separation of ampicillin was achieved
number, 2005L02145). The protocol was approved by on a 150 × 4.6 mm, 5-µm column (Scienhome Kroma-
the ethics and research committee of The First Affiliated cil, Scientific Instrument Co. Ltd, Tianjin, People’s Re-
Hospital of China Medical University, and the study was public of China), which was protected with a C18 guard
performed in accordance with the principles of the Dec- cartridge (10 × 4.6 mm inner diameter [ID]) (Zhonghui-
laration of Helsinki and its amendments8 for biomedical da Scientific Instrument Co. Ltd., Dalian, People’s Re-
research involving human subjects and the Guideline for public of China) at a column temperature of 20°C. The
Good Clinical Practice recommended by the SFDA.9 mobile phase consisted of acetonitrile and 0.068 mol/L
Volunteers were hospitalized at 5:00 pm the day monopotassium phosphate (7.5:92.5 vol/vol) at a flow
before the study and fasted 10 hours before each period rate of 1.0 mL/min, and the wavelength of the UV-Vis
of drug administration. Based on a random-number detector was set at 210 nm.
table,10 all subjects were assigned in a 1:1 ratio to re- The chromatography conditions for probenecid were
ceive a single dose of the test or reference formula- also optimized for the analytical column, mobile phase,
tions of 1500 mg (six 250-mg dispersible tablets or and sample detection. The separation method includ-
capsules) with 250 mL of water. Volunteers were not ed an octadecylsilane column (200 × 4.6 mm ID,
allowed to eat until 4 hours after administration, and 5-µm; Diamonsil ODS, Dikma Technology, Beijing,
were provided a standardized lunch and dinner11 (200– People’s Republic of China) fitted with C18 security
400 g cooked rice or other staple food, ~400 g vege- guard cartridges (10 × 4.6 mm ID; Zhonghuida Scien-
tables, ~50 g meat, and 250 mL fruit juice) according tific Instrument Co. Ltd.). The mobile phase was pre-
to a regular time schedule. pared by mixing methanol 0.1% and phosphoric acid
The volunteers were under continuous medical su- at a ratio of 78:22 (vol/vol). The analyses were
pervision at the Clinical Practice Center of The First conducted at room temperature (20°C), a flow rate of
Affiliated Hospital of China Medical University through- 1.0 mL/min, and a detection wavelength of 254 nm.
out the study. Blood samples of ~5 mL were collected
from a suitable forearm vein into vacuum heparinized Sample Preparation
tubes containing 0.5 mL 0.4% heparin sodium (Shen- Blood sample analysis was conducted at the Deter-
yang Biochemistry Company, Shenyang, People’s Re- mination Laboratory of the First Affiliated Hospital
public of China). Samples were obtained before study of China Medical University.
drug administration (0 hour; baseline) and 0.25, 0.5, The procedures described here were applied to all
0.75, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0, 6.0, 8.0, 10.0, subject samples as well as to the samples for calibra-
15.0, and 24.0 hours after study drug administration. tion curve and quality control (QC) processes. The de-
Just before each blood sample was collected, heparin ionized water was purified (Milli-Q water system,
in the heparin-locked catheter was discarded with Millipore, Bedford, Massachusetts). Blank plasma was
1 mL of blood. Plasma was immediately separated by obtained from the blood center in Shenyang.
centrifugation at 1000g for 10 minutes at room tem- An extraction method for ampicillin in plasma was
perature (20°C), then transferred directly into labeled developed and validated. Briefly, 200 µL of plasma or
2-mL polypropylene tubes and stored at –20°C until working standard of ampicillin (prepared in plasma),
quantitative analysis for the determination of plasma 50 µL (80 µg/mL) of hydrochlorothiazide (as an inter-
drug concentrations. After administration of the alter- nal standard [IS]; purity: >99.5%; National Institute
native formulation, blood samples were drawn and for the Control of Pharmaceutical and Biological
analyzed in the same manner. Products, Beijing, People’s Republic of China) and
50 µL of HPLC-grade methanol (Kangkede Technol-
HPLC Method Development ogy Co., Ltd., Tianjin, People’s Republic of China)
In this study, ampicillin and probenecid in plasma were combined in a 2.0-mL microcentrifuge tube. Af-
were quantified by a validated HPLC method.12 The ter mixing, 200 µL of 10% perchloric acid was added,
HPLC system (Shimadzu Corporation, Kyoto, Japan) the mixture was vigorously shaken for 1 minute, and

March 2010 599

 Clinical Therapeutics

then centrifuged at 6000g for 5 minutes. The superna- profen was linear (r2 > 0.999) and in the range of 0.20
tant was filtered through 0.45-µm membrane filters to 20.0 µg/mL.
(HAWG01300 MF-Millipore filters, Millipore [Shang- Intra- and inter-day precision and recoveries for
hai] Trading Co., Ltd., Shanghai, People’s Republic of ampicillin and probenecid determination in human
China) for injection into the HPLC column. plasma are shown in Table I.
Plasma sample deproteinization and extraction pro- The short-term stability of extracted samples over
cedures for probenecid were as follows: 100 µL of plas- 24 hours at room temperature and the freeze–thaw
ma, 50 µL (250 µg/mL) of ibuprofen (IS; purity: >99.7%; cycle stability of frozen samples (frozen at –20°C,
National Institute for the Control of Pharmaceutical thawed at room temperature [20°C], for 3 cycles,
and Biological Products), and 50 µL of methanol were 2 hours per cycle) were determined to ensure that
mixed in a tapered tube, then 100 µL of HPLC-grade the samples were stable during the handling and
acetonitrile (Kangkede Technology) was added and analysis processes. Long-term storage stability of plas-
mixed. This mixture was centrifuged at 6000g for ma samples at –20°C was determined after 30 days to
5 minutes and 20 µL of filtered supernatant was in- ensure the stability of frozen samples at storage tempera-
jected for analysis. ture. The results of these tests are shown in Table II.
Ampicillin and probenecid were stable for 24 hours at
Method Validation and Quantification room temperature in extracted plasma samples, for
The validation scheme involved the analysis of cali- 30 days at –20°C, and for 3 freeze–thaw cycles in
bration curves and QC samples at different concentra- plasma.
tions to determine linearity, inter- and intra-assay The analytical method for ampicillin and probene-
precision and accuracy, limit of quantitation, stability, cid quantification in plasma samples was validated
and recovery. and could be used in a bioequivalence study according
Calibration standards and blanks were freshly pre- to international guidelines.13,14
pared in duplicate for each assay and extracted in
parallel with plasma and QC samples. The extraction Tolerability
recovery of ampicillin or probenecid was determined Tolerability of the drug was determined by clinical
by comparing the peak area obtained from the plasma assessment and vital sign monitoring (body tempera-
sample with the peak area obtained by direct injection ture, blood pressure, heart rate, electrocardiography)
of pure drug standard at 3 QC levels in 6 replicates. at baseline (before dosing) and completion of the
Linearity of ampicillin or probenecid was assessed by study. Subjects were interviewed by the investigators
plotting area ratios versus standard concentrations (2 physicians, ≥2 nurses) during the study concerning
and using a linear regression weighted 1/concentration2 the occurrence of adverse events (AEs). AEs were as-
(w = 1/x2). sessed at the time of each blood draw using direct
Chromatographs are shown in Figure 1. Typical observation, spontaneous reporting, and nonspecific
retention time for ampicillin and IS were 6.18 and questioning. Any undesirable sign, symptom, or medi-
11.06 minutes, respectively, while probenecid and cal conditions occurring after the start of the study
ibuprofen were 5.98 and 10.68 minutes, respectively. were recorded in the case-report forms by the study
Though the retention times were different, the varia- physicians. Laboratory tests (hematology, blood bio-
tion was within a reasonable range. No significant chemistry, hepatic function, urinalysis) and 12-lead
interference or endogenous peaks at the retention ECG were also performed at baseline, 24 hours after
times of ampicillin or probenecid or their ISs were study drug administration, 7 days after completion of
observed in chromatographs of blank plasma. the study, and 6 months self-reported thereafter by
Under these conditions, calibration curves of am- subjects. For hematology (hemoglobin; red blood cell;
picillin were constructed at the concentrations of platelet; white blood cell with lymphocyte, monocyte,
0.25, 0.5, 1.25, 2.5, 5.0, 12.5, and 25.0 µg/mL, and eosinophil, and basophil counts), blood was assayed
the lower limit of quantitation was established as using a hematology analyzer (XE-2100, Sysmex Cor-
0.25 µg/mL. The calibration curve had a correlation poration, Kobe, Japan). Blood chemistry (total protein,
coefficient (r2) of 0.9992. The relationship between glucose, blood urea nitrogen, albumin, total bilirubin,
probenecid concentration and peak area ratio of ibu- total cholesterol, creatinine) was determined with the

600 Volume 32 Number 3

 H. Wu et al.

A D
50 10

25 5
mV

mV
0 0

0 5 10 15 0 5 10
Time (min) Time (min)

B E
50 10
IV
10.680

25 II
5 III
I 5.978
mV

mV

11.059
6.183

0 0

0 5 10 15 0 5 10
Time (min) Time (min)

C F
50 10

25 5
II
mV
mV

I III IV
11.051 6.000
6.179 10.687

0 0

0 5 10 15 0 5 10
Time (min) Time (min)

Figure 1. Chromatography of (A) blank plasma for ampicillin, (B) ampicillin 0.25 µg/mL and hydrochlorothi-
azide internal standard 80 µg/mL, and (C) human plasma sample obtained 4.0 hours after single-
dose administration of 1500-mg ampicillin/probenecid dispersible tablets spiked with hydro-
chlorothiazide 80 µg/mL; (D) blank plasma for probenecid, (E) probenecid 4 µg/mL and ibuprofen
internal standard 250 µg/mL, and (F) human plasma sample obtained 15 hours after single-dose
administration of 1500-mg ampicillin/probenecid capsules spiked with ibuprofen 250 µg/mL. Peak
I = ampicillin; peak II = hydrochlorothiazide; peak III = probenecid; peak IV = ibuprofen.

March 2010 601

Clinical Therapeutics

Table I. Intra- and inter-day precision and recoveries of ampicillin and probenecid in human plasma quality
control samples prepared from healthy Chinese male volunteers. Values are mean (SD) unless other-
wise specified.

Intra-day (n = 6) Inter-day (n = 6)

Theoretical Measured Measured Relative

Concentration, Concentration, Concentration, Recovery,
µg/mL µg/mL RSD, % RE, % µg/mL RSD, % RE, % %
Ampicillin
0.5 0.48 (0.02) 4.93 –5.00 0.49 (0.02) 3.55 –1.33 94.63 (9.55)
2.5 2.52 (0.21) 8.38 0.87 2.43 (0.19) 7.76 –2.67 88.20 (4.46)
12.5 11.94 (0.91) 7.61 –4.42 12.55 (0.96) 7.62 0.37 93.45 (7.78)
Probenecid
0.4 0.42 (0.03) 8.14 5.83 0.39 (0.04) 9.97 –1.25 98.04 (12.55)
2.0 1.92 (0.13) 6.51 –4.00 1.87 (0.14) 7.63 –5.67 94.75 (11.88)
16.0 15.61 (1.34) 8.59 –2.46 15.27 (1.10) 7.18 –4.59 89.97 (5.75)

RSD, % = relative standard deviation, calculated as SD/mean × 100%; RE, % = relative error, calculated as (measured
concentration – theoretical concentration)/theoretical concentration × 100%.

Table II. Stability data for ampicillin and probenecid in human plasma (n = 6).* Values are mean (SD).

Spiked Concentration, µg/mL

Ampicillin Probenecid

Stability 0.5 2.5 12.5 0.4 2.0 16.0

Short-term stability
Baseline 0.49 2.46 12.52 0.38 2.07 15.95
(0.01) (0.01) (0.16) (0.05) (0.13) (0.72)
After 24 hours at room 0.49 2.45 12.56 0.39 2.03 16.20
temperature (0.02) (0.05) (0.16) (0.05) (0.15) (0.75)
Long-term stability
Baseline 0.48 2.48 12.54 0.40 2.08 15.75
(0.04) (0.05) (0.15) (0.05) (0.12) (0.77)
After 30 days at –20°C 0.50 2.45 12.79 0.39 2.06 15.95
(0.02) (0.06) (0.19) (0.05) (0.17) (0.72)
Freeze–thaw stability
Baseline 0.51 2.47 12.52 0.42 2.01 15.80
(0.04) (0.08) (0.19) (0.07) (0.15) (0.69)
After 3 freeze–thaw cycles† 0.48 2.44 12.86 0.39 2.08 16.03
(0.07) (0.05) (0.70) (0.06) (0.12) (0.74)
* No significant differences versus baseline were found.
† Freeze temperature, –20°C; thaw temperature, room temperature, 20°C; 2 hours for each cycle.

602 Volume 32 Number 3

 H. Wu et al.

same system. Urinalysis (pH, specific gravity, ketone, [range, 18.6–24.8 kg/m2]) were included and com-
bilirubin, occult blood, urobilinogen) was conducted pleted the study. The results of physical examination
with a urine analyzer (UF-1000i, Sysmex Corpora- of all subjects during prestudy and poststudy visits
tion). All laboratory tests were performed at the Clini- were all normal. No volunteer was withdrawn from
cal Laboratory of The First Affiliated Hospital of the bioequivalence study.
China Medical University.
Pharmacokinetic Properties
Pharmacokinetic Evaluations and Statistical Analysis The mean (SD) serum concentration–time curves of
A noncompartmental PK method was used to calcu- the 2 formulations of ampicillin/probenecid are shown
late the PK parameters of ampicillin and probenecid.15,16 in Figure 2. The primary PK parameters (Cmax, Tmax,
Individual plasma concentration–time curves were t1/2, ke, AUC0–24, AUC0–∞) for both drugs in both
constructed to obtain Cmax and Tmax. AUC0–24 was formulations are summarized in Table III. The mean
calculated using the trapezoidal rule,17 and AUC0–∞ (SD) Cmax values of the test and reference formula-
was calculated using the following formula: tions were 13.45 (3.43) and 15.04 (5.68) µg/mL for
ampicillin and 15.56 (2.94) and 16.01 (2.88) µg/mL
AUC0–24 + C24/ke, for probenecid, respectively. The mean Tmax values
were 1.58 (0.49) and 1.78 (0.55) hours for ampicillin
where C24 was the concentration at 24 hours and ke was and 2.85 (0.78) and 3.30 (1.51) hours for probenecid.
the elimination rate constant. From the terminal log- The mean values for t1/2 and ke for the test formula-
decay phase, ke was obtained from the slope of the linear tion were 1.92 (0.67) hours and 0.59 (0.20) h–1 for
regression of the log-transformed concentration versus ampicillin and 4.08 (1.08) hours and 0.18 (0.04) h–1
time data in the terminal portion of the curve, and t1/2 for probenecid, respectively. For the reference formu-
was estimated using the following equation: lation, these values were 1.82 (0.64) hours and 0.61
(0.19) h–1 for ampicillin and 3.74 (0.83) hours and
t1/2 = 0.693/ke. 0.19 (0.05) h–1 for probenecid. The AUC0–24 of
ampicillin in the test and reference formulations was
An ANOVA for a 2 × 2 crossover design in natural 50.78 (13.39) and 57.44 (17.27) µg/mL/h, respective-
log-transformed Cmax, AUC0–t, and AUC0–∞ was car- ly, and the corresponding values for probenecid
ried out to determine the bioavailability. The 90% CIs were 129.23 (27.59) and 127.29 (26.89) µg/mL/h.
for the corresponding differences in Cmax, AUC0–t, and There was no statistically significant difference in PK
AUC0–∞ were calculated and ANOVA was performed parameters between the 2 groups. On ANOVA, no
using SPSS version 13.0 (SPSS Inc., Chicago, Illinois). period or sequence effects were observed for any
Probability of exceeding the limits of acceptance PK property.
(80%–125%) was obtained by two 1-sided t tests de-
scribed by Schuirmann18 and Anderson and Hauck.19 Bioequivalence
The 2 formulations were considered bioequivalent if Table IV lists the 90% CIs of the ratios (test:reference)
the geometric mean ratios of the Cmax and AUC were for the log-transformed Cmax (an index of the rate of
within the predetermined range of 80% to 125% and absorption) and for AUC0–t and AUC0–∞ (indexes of
if P for the 90% CIs was ≤0.05. All PK and statistical the extent of absorption), as well as the probability of
analyses were conducted using Program 3P97 soft- exceeding the limits of acceptance for bioavail-
ware (Chinese Pharmacologic Society, Beijing, People’s ability.7,9–11 The 90% CIs for the ratios of Cmax,
Republic of China). AUC0–t, and AUC0–∞ values were as follows: 86.5% to
108.0%, 96.7% to 107.8%, and 83.3% to 100.7%
RESULTS for ampicillin and 90.2% to 108.3%, 96.8% to
Twenty-five healthy Chinese male subjects were as- 107.8%, 97.2% to 108.5% for probenecid, respec-
sessed for inclusion in the study. Twenty subjects (mean tively. The relative bioavailability of ampicillin and
[SD] age, 21.4 [2.2] years [range, 19–23 years]; weight, probenecid in the test formulation compared with
64.1 [5.5] kg [range, 51–69 kg]; height, 173.7 [5.3] cm the reference formulation was 92.2% (21.8%) and
[range, 160–178 cm]; and BMI, 21.2 [1.6] kg/m2 102.3% (13.4%), respectively.

March 2010 603

 Clinical Therapeutics

Ampicillin (test)
A Ampicillin (reference)
20
Plasma Ampicillin Concentration (µg/mL)

18

16

14

12

10

0
0 5 10 15 20 25
Time (h)
Probenecid (test)
B Probenecid (reference)
20
Plasma Probenecid Concentration (µg/mL)

18

16

14

12

10

0
0 5 10 15 20 25
Time (h)
Figure 2. Mean (SD) concentration–time profiles for (A) test (Shenyang Shidte Pharmaceutical Co. Ltd.,
Shenyang, People’s Republic of China) and (B) reference (Aerodyn®, Changchun Gaipu Pharmaceutical
Co. Ltd., Changchun, People’s Republic of China) 1500-mg ampicillin/probenecid formulations after
single-dose administration in healthy Chinese male volunteers (N = 20).

Tolerability vomiting, diarrhea, enteritis) were not reported during

Both ampicillin/probenecid formulations were well the study or within 1 week after the study end.
tolerated by all of the volunteers. No AEs were reported
by the subjects or found on analysis of vital signs or DISCUSSION
laboratory tests. AEs associated with ampicillin and This study examined the PK properties, bioavailabili-
probenecid (eg, mild skin rash, upset stomach, nausea, ty, and bioequivalence of 2 compound formulations of

604 Volume 32 Number 3

 H. Wu et al.

Table III. Pharmacokinetic parameters after administration of 1500 mg of test* (ampicillin/probenecid dis-
persible tablet) and reference† (ampicillin/probenecid capsule) formulations in healthy Chinese male
volunteers (N = 20). Values are mean (SD).

Ampicillin Probenecid

Parameter Test Reference Test Reference

Cmax, µg/mL 13.45 (3.43) 15.04 (5.68) 15.56 (2.94) 16.01 (2.88)
Tmax, h 1.58 (0.49) 1.78 (0.55) 2.85 (0.78) 3.30 (1.51)
AUC0–24, µg/mL/h 50.78 (13.39) 57.44 (17.27) 129.23 (27.59) 127.29 (26.89)
AUC0–∞, µg/mL/h 51.95 (13.45) 58.74 (17.19) 133.85 (28.80) 131.21 (28.25)
t1/2, h 1.92 (0.67) 1.82 (0.64) 4.08 (1.08) 3.74 (0.83)
ke , h–1 0.59 (0.20) 0.61 (0.19) 0.18 (0.04) 0.19 (0.05)

ke = elimination rate constant.

* Manufacturer: Shenyang Shidte Pharmaceutical Co. Ltd., Shenyang, People’s Republic of China.
† Trademark: Aerodyn® (Changchun Gaipu Pharmaceutical Co. Ltd., Changchun, People’s Republic of China).

ampicillin and probenecid—a newly developed dis-

Table IV. Comparison of 90% CIs for the natural persible tablet and a marketed branded capsule—in
log-transformed parameters (Cmax, healthy Chinese male volunteers. The 90% CIs were
AUC0–t, and AUC0–∞) of 2 oral com- completely contained within the predefined SFDA’s
pound formulations of ampicillin and bioequivalence criteria of 80% to 125% for the pri-
probenecid following a single 1500-mg mary end point of AUC and 70% to 143% for Cmax.9,11
dose of administration in healthy Chinese The mean plasma decay curves obtained for the test
male volunteers (N = 20). and reference formulations in the present study were
not significantly different. The statistical comparison
Test*:Reference† of Cmax, AUC0–t, and AUC0–∞ showed no significant dif-
Parameter Ratio, % 90% CI ferences between the test and reference formulations.
Ampicillin The present study had a few limitations that should
be considered. The PK data were obtained only from
Cmax 97.3 86.5–108.0
healthy subjects who were administered a single dose,
AUC0–t 102.3 96.7–107.8 and the PK characteristics of ampicillin and probenecid
AUC0–∞ 92.0 83.3–100.7 might differ in target populations (female and patients
Probenecid with infection in clinical practice). This study was
conducted in fasted subjects; therefore, the results
Cmax 99.2 90.2–108.3
cannot be applied to predict the PK properties of am-
AUC0–t 102.3 96.8–107.8 picillin and probenecid concomitantly administered
AUC0–∞ 102.8 97.2–108.5 with food. The study was conducted with a small sample
size (20 volunteers) according to the related guid-
*Manufacturer: Shenyang Shidte Pharmaceutical Co.
ance9,11 of SFDA in China. Therefore, the results of
Ltd., Shenyang, People’s Republic of China.
†Trademark: Aerodyn® (Changchun Gaipu Pharma- the present study require confirmation from addi-
ceutical Co. Ltd., Changchun, People’s Republic of tional clinical studies with larger and more varied
China). sample sizes to evaluate the tolerability and to repre-
sent the Chinese PK characteristics.

March 2010 605

 Clinical Therapeutics

CONCLUSIONS publications/10policies/b3/index.html. Accessed August

In this small study in healthy Chinese male volunteers, 16, 2006.
a single 1500-mg dose of the dispersible tablet com- 9. State Food and Drug Administration. Good Clinical Prac-
tice Guideline. http://www.sda.gov.cn/WS01/CL0053/
pound formulation (test) of ampicillin and probenecid
24473.html. Accessed August 6, 2003.
met the SFDA’s regulatory criteria for bioequivalence
10. Xu Y, Sun Z. Medical Statistics. 2nd ed. Beijing, China:
to a single 1500-mg dose of the established capsule
Higher Education Press; 2004; Appendix 15:343–344.
compound formulation (reference) based on the rate 11. State Food and Drug Administration. Provisions for Drug.
and extent of absorption. Both formulations were well http://www.sda.gov.cn/WS01/CL0053/24529.html. Ac-
tolerated. cessed July 10, 2007.
12. Yinglan Z, Juying H, Songqing L, et al. Content determina-
ACKNOWLEDGMENTS tion of ampicillin and probenecid in ampicillin/probenecid
The authors would like to thank Shenyang Shidte Phar- capsules by HPLC. China Pharmaceuticals. 2006;15:27–29.
maceutical Co. Ltd. for providing the formulations of 13. US Department of Health and Human Services, Food and
ampicillin and probenecid used in this study. The au- Drug Administration (FDA), Center for Drug Evaluation
thors have indicated that they have no other conflicts and Research (CDER). Guidance for Industry: Bioavail-
ability and bioequivalence studies for orally administered
of interest with regard to the content of this article.
drug products—general considerations. http://www.fda.
gov/downloads/Drugs/GuidanceComplianceRegulatory
REFERENCES Information/Guidances/ucm070124.pdf. Accessed No-
1. Katzung BG. Basic and Clinical Pharmacology. 2nd ed. Los
vember 8, 2008.
Altos, Calif: Lange Medical Publications; 1996.
14. US Food and Drug Administration. In vivo bioequivalence
2. Hardman JG, Limbird LE, Goodman Gilman A. Goodman &
guidance. Pharmacopeial Forum. 1993;19:6501–6508.
Gilman’s The Pharmacological Basis of Therapeutics. 10th ed.
15. US Department of Health and Human Services, Food and
New York, NY: McGraw-Hill Inc; 2001.
Drug Administration (FDA), Center for Drug Evaluation
3. Schumacher I, Margalit R. Liposome-encapsulated ampi-
and Research (CDER). Guidance for Industry: Statistical
cillin: Physiochemical and antibacterial properties. J Pharm
Sci. 1997;86:635–641. approaches to establishing bioequivalence. http://www.
4. Lafforgue G, Arellano C, Vachoux C, et al. Oral absorp- scribd.com/doc/4204696/Statistical-Approaches-to-
tion of ampicillin: Role of paracellular route vs. PepT1 Establishing-Bioequivalence. Accessed March 3, 2010.
transporter. Fundam Clin Pharmacol. 2008;22:189–201. 16. US Food and Drug Administration. Instruction for the In
5. DiPiro JT, Albert RL. Pharmacotherapy: A Pathophysiologic Vivo Bioequivalence Study Protocol Development. http://
Approach. 5th ed. New York, NY: McGraw-Hill Inc; 2002. www.app1.fda.moph.go.th/drug/zone_bioequivalence/
6. Gehring R, van der Merwe D, Pierce AN, et al. Multivariate files/be_inhuman.pdf. Accessed September 24, 2008.
meta-analysis of pharmacokinetic studies of ampicillin 17. Williams RL, Chen ML, Hauck WW. Equivalence approach-
trihydrate in cattle. Am J Vet Res. 2005;66:108–112. es. Clin Pharmacol Ther. 2002;72:229–237.
7. Kampffmeyer HG, Hartmann I, Metz H, et al. Serum concen- 18. Schuirmann DJ. A comparison of the two one-sided tests
trations of ampicillin and probenecid and ampicillin excre- procedure and the power approach for assessing the
tion after repeated oral administration of a pivampicillin- equivalence of average bioavailability. J Pharmacokinet Bio-
probenecid salt (MK-356). Eur J Clin Pharmacol. 1975;9: pharm. 1987;15:657–680.
125–129. 19. Anderson S, Hauck WW. A new procedure for testing
8. World Medical Association (WMA). WMA Declaration equivalence in comparative bioavailability and other clini-
of Helsinki - Ethical Principles for Medical Research In- cal trials. Commun Stat Theory Methods. 1983;12:2663–
volving Human Subjects. http://www.wma.net/en/30 2692.

Address correspondence to: Minjie Wei, MD, PhD, Department of

Pharmacology, School of Pharmaceutical Sciences, China Medical
University, Shenyang 110001, People’s Republic of China. E-mail: mjwei@
mail.cmu.edu.cn

606 Volume 32 Number 3