Am. J. Trop. Med. Hyg., 71(6), 2004, pp.

703–710
Copyright © 2004 by The American Society of Tropical Medicine and Hygiene

COMPARISON OF THE BIOEQUIVALENCE OF THREE ORAL FORMULATIONS

OF DIHYDROARTEMISININ BASED ON EX VIVO BLOOD SCHIZONTOCIDAL
ACTIVITIES AGAINST PLASMODIUM FALCIPARUM
MONTICHA KONGTHAISONG, KESARA NA-BANGCHANG, MATHIRUT MUNGTHIN, NUTTANAN SINCHAIPANID,
AND PEERAPAN TAN-ARIYA
Department of Microbiology, Faculty of Science, and Department of Pharmacognosy, Faculty of Pharmacy, Mahidol University,
Bangkok, Thailand; Department of Medical Technology, Faculty of Allied Health Sciences, Thammasat University, Klongluang,
Pathumthani, Thailand; Department of Parasitology, Phramongkutklao College of Medicine, Bangkok, Thailand

Abstract. Sera collected at various time intervals from healthy Thai male subjects after the administration of the
three oral formulations of dihydroartemisinin (Cotecxin威 manufactured in the People’s Republic of China, a formulation
manufactured by Arenco n.v. Pharmaceutica in Belgium, and a formulation manufactured by the Faculty of Pharmacy
of Mahidol University in Thailand) were investigated for their ex vivo blood schizontocidal activities against the K1
strain of Plasmodium falciparum. Blood schizontocidal activities of sera were evaluated using the maximum inhibitory
dilution as a parameter. Sera obtained following the administration of the three formulations of dihydroartemisinin
showed significantly distinct degree of ex vivo antimalarial activities. The differences may reflect the bioinequivalence
between these three formulations of dihydroartemisinin. The ex vivo blood schizontocidal activity profiles generally
coincided with plasma concentration-time profiles. Thus, the ex vivo model might be the useful tool for evaluating and
comparing the bioequivalence of the interesting drugs especially where high-performance liquid chromatography with
reductive electrochemical detection for drug analysis is not available. The effect of inoculum size of P. falciparum was
shown in the ex vivo model as presented in the in vitro sensitivity test. To determine the effect of the inoculum size on
the drug activity, the ex vivo model might be superior to the in vitro model since the pharmacokinetic profiles can be
considered.

INTRODUCTION from 10 healthy Thai male subjects were investigated after the
Malaria remains a major public health problem worldwide administration of the three oral formulations of dihydroarte-
due to the resistance of Plasmodium falciparum to the cur- misinin (Cotecxin威 from the People’s Republic of China, a
rently used antimalarial drugs. At present, artemisinin and its formulation manufactured in Belgium, and a formulation
derivatives are the most promising antimalarials available for manufactured by the Faculty of Pharmacy of Mahidol Uni-
the treatment of malaria in many areas of the world, espe- versity in Thailand). One of the most important factors in
cially in the Southeast Asian region where multidrug-resistant determining drug sensitivity in vitro is the inoculum effect.14–20
P. falciparum exists.1,2 Dihydroartemisinin, which is the ma- The term of inoculum effect is defined as an increase in drug
jor active metabolite of all artemisinin derivatives, has proved concentrations necessary to inhibit parasite growth with high
to be more active against P. falciparum than the parent com- percentages of parasitemia.15 The study of Duraisingh and
pounds.3–7 Since dihydroartemisinin is the most potent and others underlines the importance of inoculum size on in vitro
least expensive to manufacture, it has the potential to be de- sensitivity testing with artemisinin and dihydroartemisinin.20
veloped and used as a formulated drug. Dihydroartemisinin They demonstrated that the 50% inhibition concentration
was originally produced in China as a tablet formulation and (IC50) of both drugs were increased depending on the increas-
has been shown to be effective in the treatment of multidrug- ing inoculum size. Despite the higher drug concentration
resistant P. falciparum malaria in this country.8 Recently, di- achieved in vivo with treatment doses, the inoculum effect
hydroartemisinin has been developed in Vietnam, Thailand, may occur because the absolute concentrations of parasitized
and the Netherlands, and has become available for clinical erythrocytes are high. The possibility of an inoculum effect on
trials.9 The study of the pharmacokinetics of dihydroartemisi- ex vivo blood schizontocidal activities was also investigated in
nin in healthy Vietnamese subjects was reported by Le and this study.
others.10 It was found that the observed pharmacokinetics of
oral dihydroartemisinin (Cotecxin威 manufactured in the Peo- MATERIALS AND METHODS
ple’s Republic of China) in this study were different from
those previously investigated by Na-Bangchang and others Subjects. Ten healthy Thai male subjects (age range ⳱
using a dihydroartemisinin formulation manufactured in Bel- 21–37 years, weight range ⳱ 43–64 kg) with no history of liver
gium.11 The differences may reflect the bioinequivalence be- or kidney disease were recruited into the study. Written in-
tween these two formulations of dihydroartemisinin. The ob- formed consent for participation was obtained from all sub-
served bioequivalence of drugs in various drug formulations jects before initiation of the study. Individuals who smoked,
has been previously reported.12,13 Results indicated the influ- drank alcoholic beverages, or were taking medications were
ence of pharmaceutical preparations and/or formulations on excluded, and no other drugs were given during the period of
the kinetics of the drugs. Despite the fact that several phar- study. On admission, these subjects were hospitalized over-
maceutical formulations of dihydroartemisinin have been night at the Bangkok Hospital for Tropical Diseases. Each
generated extensively, their relative bioequivalence has yet to was given a physical examination and the following tests were
be examined. preformed: monitoring of heart rate, blood pressure, routine
In the present study, ex vivo blood schizontocidal activities blood examinations (hematology and clinical chemistry), a
and bioequivalence of sera collected at various time intervals 12-lead electrocardiogram, and urinalysis. The study was re-

703
704 KONGTHAISONG AND OTHERS

viewed and approved by the Ethical Committee of Mahidol Evaluation of results. Following serum exposure for 48
University. hours, thin blood smears were made from each well and
Drug administration and blood sampling. Blood samples stained with Giemsa. The number of infected erythrocytes
were collected before (0 hour) and 0.5, 1, 1.5, 2, 3, 4, 6, 8, and and the morphology of parasites were examined under a light
10 hours after administration of a single oral dose (300 mg) of microscope. The assay was considered successful if at least a
the three different formulations of dihydroartemisinin from 4–5-fold increasing number of infected erythrocytes in the
10 healthy Thai male subjects. Formulation I was a single oral control wells was achieved. The number of infected erythro-
dose of 300 mg of dihydroartemisinin (20 mg per tablet, Co- cytes with normal apparent parasites was counted per 10,000
tecxin威; Beijing Cotec New Technology Corp., Beijing Sixth erythrocytes. The effect of sera containing dihydroartemisinin
Pharmaceutical Factory, Bejing, People’s Republic of China). on the parasite growth was assessed microscopically by both
Formulation II was a single oral dose of 300 mg of dihydroar- the decrease in parasite density and the viability of the re-
temisinin (50 mg per tablet; Arenco n.v. Pharmaceutica, Geel, maining parasites. Ex vivo blood schizontocidal activities of
Belgium). Formulation III was a single oral dose of 300 mg sera were evaluated using the maximum inhibitory dilution
dihydroartemisinin (100 mg per tablet; Dr. Nuttanan Sin- (MID) as the parameter indicating antimalarial activity. The
chaipanid, Faculty of Pharmacy, Mahidol University, Bang- MID was the final dilution of sera containing drug at which
kok, Thailand). complete inhibitory effect on parasite growth (at least 95%
The study was a cross-over, randomized design in which inhibition, IC95) was observed.
subjects were randomized to receive one of the three different Determination of inoculum effect on ex vivo blood schiz-
formulations of dihydroartemisinin on three separate occa- ontocidal activities of sera. Due to the limited amount of
serum samples, sera collected from seven healthy Thai male
sions. The wash-out period during each occasion of drug ad-
subjects after the administration of the three different formu-
ministration was seven days. During this study, all subjects
lations of dihydroartemisinin were selectively used to inves-
were given the same diet and allowed to eat two hours after
tigate the inoculum effect. The K1 strain of P. falciparum was
drug administration. Plasma samples for quantification of di-
used in this experiment. A parasite inoculum was prepared as
hydroartemisinin concentrations were obtained by centrifu-
a 3% cell suspension with 1% and 5% ring-stage parasit-
gation (1,200 × g for 10 minutes) within five minutes of ob-
emias. The methodology and determination of the inoculum
taining the blood sample and stored at −80°C until analysis.
effect was performed using the same method used for assess-
Serum samples for assessment of ex vivo antimalarial activi-
ment of ex vivo antimalarial activities, as described earlier in
ties were obtained from the remaining portion of blood by this report.
centrifugation (1,200 × g for 10 minutes) immediately after Statistical analysis. Pharmacokinetic characteristics for
clotting (approximately two hours) and stored at −80°C until rate and extent of dihydroartemisinin absorption were the
analysis. maximum plasma concentrations (Cmax), the time at which
Drug analysis. Dihydroartemisinin concentrations in plas- maximum plasma concentrations occurred (tmax), and the
ma were measured using high-performance liquid chromatog- area under the plasma concentration-time curve (AUC). All
raphy with reductive electrochemical detection (HPLC-ED) pharmacokinetic parameters of dihydroartemisinin were ana-
at Department of Medical Technology, Faculty of Allied lyzed using model-independent methods for non-normally
Health Sciences at Thammasat University according to the distributed data.24 The tmax and Cmax were obtained directly
method of Na-Bangchang and others.21 from plasma concentration-time profiles. The area under the
Assessment of ex vivo blood schizontocidal activities of plasma concentration-time curve from zero time to the last
sera. All serum samples were investigated for their blood observed time (AUC0-t) or certain specific time points were
schizontocidal activities against the chloroquine-resistant, di- calculated using the linear trapezoidal rule for ascending data
hydroartemisinin-sensitive K1 strain of P. falciparum. A 50% points and by the log trapezoidal rule for descending data
(v/v) cell suspension in complete RPMI 1640 medium with points.
1% ring stages of the K1 strain was used as the parasite in- Ex vivo blood schizontocidal activity profiles, i.e., maximal
oculum. This suspension was then adjusted to a 3% cell sus- MIDs, time to maximum MIDs, and area under the MID
pension with complete RPMI 1640 medium and is referred to pattern (AUD) were analyzed using model-independent
as the parasite inoculum. methods for non-normally distributed data.24 Maximum
Assessment of ex vivo blood schizontocidal activities of MIDs was determined by noting the highest MID (peak an-
sera was performed using the in vitro microtechnique accord- timalarial activity) of sera collected from individual samples
ing to the method of Rieckmann and others22 with some following the administration of each oral formulation of di-
modifications.23 Briefly, serum samples collected at each in- hydroartemisinin. The time at which maximum MIDs ob-
terval were serially diluted two-fold and six-fold with normal served and the maximum MIDs were obtained directly from
AB serum to dilutions of 1:2, 1:4, 1:6, 1:8, 1:12, 1:16, 1:24, 1:32, ex vivo antimalarial activity profiles (MID patterns). The area
1:48, 1:64, 1:96, and 1:128, respectively. Throughout the study, under the MID pattern from zero time to the last observed
the same batch of pooled AB serum was used for the dilution. time (AUD0-t) or certain specific time points were calculated
In the control wells, normal AB serum and serum collected using the trapezoidal rule for all data points.
prior to drug administration (hour 0) were used. Eighty mi- Comparison of both pharmacokinetic parameters (i.e.,
croliters of parasite inoculum was added into each well con- Cmax, tmax, and AUC) of dihydroartemisinin and ex vivo an-
taining 20 ␮L of undiluted and/or diluted serum samples. The timalarial activity profiles (i.e., maximum MIDs, time to
final volume in each well of the microtiter plate was 100 ␮L. maximum MIDs, and AUD) of sera obtained following the
After gentle mixing, the microtiter plates was placed in a administration of dihydroartemisinin between formulations
candle jar and incubated at 37°C for 48 hours. was performed using the Mann-Whitney U test for non-
 BIOEQUIVALENCE OF THREE DIHYDROARTEMISININ FORMULATIONS 705

normally distributed data. The level of statistical significance healthy Thai male subjects following the administration of the
was set at P < 0.05. three oral formulations of dihydroartemisinin are shown in
The consideration of bioequivalence of the three oral for- Figure 2. No inhibitory effect was observed in the undiluted
mulations of dihydroartemisinin was based on the two-sided sera collected prior to drug administration, as well as in the
t-test approach that included classic 95% confidence intervals. control wells. In most serum samples, activities of sera con-
taining dihydroartemisinin for all three oral formulations
RESULTS were detectable in the first serum sample (30 minutes after
drug administration). As for the MID pattern of Cotecxin威, a
Morphologic changes of P. falciparum after exposure to complete inhibitory effect was detected starting at 0.5–8 hours
sera. In the wells containing sera collected before drug ad- in most serum samples. The MIDs of individual samples var-
ministration, as well as in the control (normal AB serum) ied throughout the investigation period, with the range of
wells, all parasite stages, i.e., rings, trophozoites, and schizonts, dilution between 1:1 and 1:16. The highest MID was achieved
were found to be normal in appearance (Figure 1a). In con- at three hours with a dilution of 1:16. After the peak of an-
trast, no developmental growth of parasites was observed in timalarial activities, decreases in the MIDs were apparent,
the test wells after exposure to sera containing dihydroarte- and no inhibitory effect was observed at 10 hours in all serum
misinin at inhibitory level. In wells with complete inhibition, samples. Serum samples at a dilution <1:1 (undiluted serum)
all parasites were inhibited at the ring stage. Most of dead had no inhibitory effect. The MID pattern of the formulation
rings had purple-stained pyknotic nuclei with no cytoplasm manufactured by Arenco n.v. Pharmaceutica was generally
(Figure 1b). The wells that showed incomplete inhibition (at similar to that of Cotecxin威. There were no differences in
a higher titer of sera containing dihydroartemisinin) con- either the range of MIDs (1:1–1:16) and the highest MID
tained both viable and dead parasites. In these wells, only a (1:16). However, the formulation manufactured by Arenco
few normal mature schizonts and rings were detected to- n.v. Pharmaceutica reached the peak of antimalarial activities
gether with some dead trophozoites and dead schizonts. As earlier than that of Cotecxin威 (1.5–2 hours versus 3 hours). A
for dead trophozoites, the contour of their nucleus could highly variable MID pattern was observed for the formulation
not be distinguished from their cytoplasm (Figure 1c). Dead manufactured by the Faculty of Pharmacy of Mahidol Uni-
schizonts had fewer nuclei when compared with normal ap- versity. The MIDs of individual samples varied throughout
pearance parasites in the control wells (Figure 1d). The nuclei the investigation period, with the range of dilution between
were fragmented and the contour of cytoplasm was not 1:2 and 1:64. The highest MID was achieved at 1.5 hours with
sharply pronounced. The number of infected erythrocytes a dilution of 1:64. After the peak of antimalarial activities,
with normal-appearing parasites was increased, correspond- decreases in MIDs were observed.
ing to an increase in the dilution of sera containing dihydroar- The correlations of ex vivo blood schizontocidal activities
temisinin. In the wells that showed no inhibitory effect, nor- (MIDs) and plasma concentration-time profiles of dihydroar-
mal parasites were observed, most of them in the ring stage, temisinin in 10 healthy Thai male subjects following the ad-
and the number of infected erythrocytes increased from ministration of dihydroartemisinin for all three oral formula-
the initial number at time zero as that seen with the control tions are shown in Figure 3. Ex vivo blood schizontocidal
wells. activities of sera reflected by the MID patterns generally co-
Ex vivo blood schizontocidal activities of sera. The MID incided with plasma concentrations of dihydroartemisinin.
patterns of sera collected at various time intervals from 10 The MIDs were found to increase and corresponded to an

FIGURE 1. Morphologic analysis of the K1 isolate of Plasmodium falciparum showing (a), normal mixed stages; (b), abnormal or dead parasites
after exposure to sera containing dihydroartemisinin; (c), dead trophozoites; and (d), dead schizonts. (Giemsa stained, magnification × 100.)
706 KONGTHAISONG AND OTHERS

FIGURE 2. Patterns of maximum inhibitory dilution (MID) of sera collected after the administration of the three oral formulations of
dihydroartemisinin against the K1 strain of Plasmodium falciparum. The dots below the dilution 1:1 represent undiluted serum samples that had
no inhibitory effect. h ⳱ hours.

increase in plasma concentrations of the drug. Maximum other two formulations (1.5 hours versus 2 hours and 2 hours)
MIDs was found within the period of time that concentration (P ⳱ 0.018 and 0.022, respectively). A significantly lower
of dihydroartemisinin reached a peak in most subjects. median AUD was observed with sera obtained following the
Comparison of bioequivalence of the three oral formula- administration of the formulation manufactured by Arenco
tions of dihydroartemisinin. Statistical analysis of ex vivo an- n.v. Pharmaceutica than with Cotecxin威 and the formulation
timalarial activity profiles and pharmacokinetic parameters of manufactured by the Faculty of Pharmacy of Mahidol Uni-
all three oral formulations of dihydroartemisinin using the versity (1,597 versus 2,606 and 2,511 ng/mL) (P ⳱ 0.049 and
Mann-Whitney U test are summarized in Table 1. Sera ob- 0.041, respectively). The median AUD of the dihydroarte-
tained following the administration of the formulation manu- misinin formulation manufactured by the Faculty of Phar-
factured by the Faculty of Pharmacy of Mahidol University macy of Mahidol University was not significantly different
showed significantly higher median maximum MIDs than from that of Cotecxin威. Statistical analysis did not show sig-
Cotecxin威 and the formulation manufactured by Arenco n.v. nificant differences in maximum MID, AUD, Cmax, and tmax
Pharmaceutica (1:24 versus 1:7 and 1:6) (P ⳱ 0.003). The between sera obtained following the administration of Cotec-
times to maximum MIDs of sera obtained following the ad- xin威 and sera obtained following the administration of dihy-
ministration of dihydroartemisinin for all three formulations droartemisinin formulation manufactured by Arenco n.v.
were not significantly different. The median AUD value of Pharmaceutica.
sera obtained following the administration of the formulation Inoculum effect of K1 strains of P. falciparum on ex vivo
manufactured by the Faculty of Pharmacy of Mahidol Uni- blood schizontocidal activities of sera. An increase in the
versity was significantly higher than those of the formulation inoculum size from 1% to 5% parasitemia affected ex vivo
manufactured by Arenco n.v. Pharmaceutica (54 versus 18) blood schizontocidal activities of sera. The MIDs of sera with
(P ⳱ 0.033); however, it was not significantly different from a 5% parasitemia were lower than those with a 1% parasit-
that of Cotecxin威. A significantly higher median Cmax was emia at nearly all time intervals, with varying degrees depend-
observed with sera obtained following the administration of ing on the parasites and the drug formulations.
the formulation manufactured by the Faculty of Pharmacy of The MID patterns of sera collected from healthy Thai male
Mahidol University than with Cotecxin威 and the formulation subjects following the administration of the three oral formu-
manufactured by Arenco n.v. Pharmaceutica (894.5 versus lations of dihydroartemisinin against the K1 strain of P. fal-
686.5 and 501 ng/mL) (P ⳱ 0.028 and 0.00043, respectively). ciparum using 1% and 5% parasitemias are shown in Figure
The tmax of sera obtained following the administration of the 4. When a 1% parasitemia was used, the MIDs of sera ob-
formulation manufactured by the Faculty of Pharmacy of Ma- tained following the administration of Cotecxin威 and the for-
hidol University was significantly shorter than those of the mulation manufactured by Arenco n.v. Pharmaceutica were
 BIOEQUIVALENCE OF THREE DIHYDROARTEMISININ FORMULATIONS 707

FIGURE 3. Correlation of maximum inhibitory dilutions (bar) and plasma concentrations (line) of dihydroartemisinin following the admin-
istration of the three oral formulations. Data are median values. h ⳱ hours.

in the range of dilutions between 1:1 and 1:16. The highest highest MID of sera obtained following the administration of
MID of sera obtained following the administration of both Cotecxin威 and the formulation manufactured by the Faculty
formulations was 1:16. The MIDs of sera obtained following of Pharmacy of Mahidol University decreased from a dilution
the administration of the formulation manufactured by the of 1:64 to a dilution of 1:8, an eight-fold decrease. Two serum
Faculty of Pharmacy of Mahidol University were in the range samples obtained following the administration of Cotecxin威
of dilutions between 1:1 and 1:64 and the highest MID was and the formulation manufactured by the Faculty of Phar-
1:64. When inoculum size of the K1 strain was increased from macy, Mahidol University had no inhibitory effect at all time
1% to 5% parasitemia, the MIDs decreased at nearly all time intervals, as shown by dots below the dilution 1:1.
intervals in most serum samples with all three formulations of Statistical analysis of ex vivo antimalarial activity (MIDs)
dihydroartemisinin, especially during the peak of antimalarial profiles of sera against 1% and 5% parasitemias of the K1
activities. The highest MID of sera obtained following the strain (Table 2) showed that the MIDs of sera with a 5%
administration of Cotecxin威 and the formulation manufac- parasitemia were significantly lower than those of sera with a
tured by Arenco n.v. Pharmaceutica decreased from a dilu- 1% parasitemia with all three oral formulations of dihydroar-
tion of 1:16 to a dilution of 1:4, a four-fold decrease. The temisinin (P < 0.05).

TABLE 1
Ex vivo blood schizontocidal activity profiles and pharmacokinetic parameters for consideration of bioequivalence of the three oral formulations
of dihydroartemisinin*

Dihydroartemisinin (300 mg)

Ex vivo antimalarial
activity profiles and Formulation Formulation manufactured by
pharmacokinetic parameters Cotecxin® manufactured by Arenco Faculty of Pharmacy, Mahidol University

Maximum MIDs 1:7 (1:2–1:16) 1:6 (1:2–1:16) 1:24† (1:8–1:64)

Time to maximum MIDs (hours) 1.75 (1–3) 1.5 (1.5–2) 1.5 (1.5–2)
AUD 25.5 (6–54) 18 (7–52) 54‡ (16.5–139)
Cmax (ng/mL) 686.5 (232–703) 501 (211–740) 894.5§ (555–1,020)
tmax (hours) 2 (1.5–2) 2 (1.5–2) 1.5¶ (1.5–2)
AUC (ng/hour/mL) 2,606# (450–2,907) 1,597 (556–2,172) 2,511** (1,186–3,559)
* Values are medians (ranges). MID ⳱ maximum inhibitory dilution; AUD ⳱ area under the MID pattern; AUC ⳱ area under the plasma concentration time curve; CI ⳱ confidence interval.
† Significantly different from Cotecxin威 (P ⳱ 0.003, 95% CI ⳱ 0.002, 0.003) and formulation produced by Arenco (P ⳱ 0.003, 95% CI ⳱ 0.001, 0.003).
‡ Significantly different from formulation produced by Arenco (P ⳱ 0.033, 95% CI ⳱ 0.026, 0.033).
§ Significantly different from Cotecxin威 (P ⳱ 0.028, 95% CI ⳱ 0.023, 0.030) and formulation produced by Arenco (P ⳱ 0.000, 95% CI ⳱ 0.000, 0.000).
¶ Significantly different from Cotecxin威 (P ⳱ 0.018, 95% CI ⳱ 0.000, 0.286) and formulation produced by Arenco (P ⳱ 0.022, 95% CI ⳱ 0.053, 0.062).
# Significantly different from formulation produced by Arenco (P ⳱ 0.049, 95% CI ⳱ 0.047, 0.055).
** Significantly different from formulation produced by Arenco (P ⳱ 0.041, 95% CI ⳱ 0.043, 0.051).
708 KONGTHAISONG AND OTHERS

FIGURE 4. Patterns of maximum inhibitory dilution (MID) of sera collected after the administration of the three oral formulations of
dihydroartemisinin against the K1 strain of Plasmodium falciparum at 1% (left) and 5% (right) parasitemia. The dots below the dilution 1:1
represent undiluted serum samples that had no inhibitory effect. h ⳱ hours.

DISCUSSION TABLE 2
Ex vivo blood schizontocidal activity (MIDs) profiles of sera obtained
Dihydroartemisinin was effective against all asexual blood following the administration of the three oral formulations of di-
stages of P. falciparum. This observation was consistent with hydroartemisinin against the K1 strain of Plasmodium falciparum
those of previous reports.25–28 It was found that all three with 1% and 5% parasitemias*
formulations of dihydroartemisinin caused similar morpho- Maximum MIDs of sera containing dihydroartemisinin
logic changes in P. falciparum. Ring, trophozoite, and schiz-
Formulation Formulation manufactured
ont stages showed abnormalities in both nuclei and cyto- % Parasitemia of manufactured by Faculty of Pharmacy,
P. falciparum Cotecxin威 by Arenco Mahidol University
plasm, as shown in Figures 1b–d. The ex vivo blood schizon-
tocidal activities of sera obtained following the administration 1% 8 (2–16) 8 (4–16) 32 (8–64)
5% 2† (2–4) 2‡ (2–4) 2§ (2–4)
of the three oral formulations of dihydroartemisinin observed
* Data are medians (ranges). MID ⳱ maximum inhibitory dilution; CI ⳱ confidence
in the present study were consistent with those previously interval.
reported, in which the rapid onset and short duration of ac- † Significantly different from 1% parasitemia (P ⳱ 0.010, 95% CI ⳱ 0.011, 0.016).
‡ Significantly different from 1% parasitemia (P ⳱ 0.003, 95% CI ⳱ 0.002, 0.004).
tivity were observed.11,29,30 As shown in Figure 2, the inhibi- § Significantly different from 1% parasitemia (P ⳱ 0.001, 95% CI ⳱ 0.000, 0.002).
 BIOEQUIVALENCE OF THREE DIHYDROARTEMISININ FORMULATIONS 709

tory effect of sera obtained following the administration of of drug absorption of these two formulation was not signifi-
dihydroartemisinin in all three oral formulations were detect- cantly different.
able in the first serum sampling (30 minutes after drug ad- The differences in ex vivo antimalarial activities and phar-
ministration). From 0.5 to 4 hours, the MIDs were highly macokinetic parameters of these three oral formulations of
variable, especially during the peak antimalarial activities. dihydroartemisinin are probably due to the differences in tab-
The lesser variable MIDs were found at 6–10 hours in most let excipients or crystal modification of each formulation. The
serum samples. This indicates the rapid decrease in ex vivo kinetics of dihydroartemisinin release (rate and extent of dis-
blood schizontocidal activities of dihydroartemisinin. Such a solution) from each formulation as it passes through the gas-
high variation in MIDs during the peak antimalarial activities trointestinal tract might reflect the rate and extent of drug
might be affected by the variability among individual subjects absorption. Modification of the pharmaceutical formulation
in the process of drug absorption from the gastrointestinal of dihydroartemisinin to improve drug dissolution, while pro-
tract, resulting in a difference in achieving the peak drug viding high and sustained plasma concentration, e.g., a sus-
concentration in the blood.31 tained release formulation, could be an alternative approach
The ex vivo blood schizontocidal activity profiles of sera to achieve immediate and prolonged therapeutic concentra-
collected from healthy Thai male subjects after the adminis- tion of dihydroartemisinin in blood circulation.33 The ex vivo
tration of the three oral formulations of dihydroartemisinin model used for the assessment of antimalarial activities of
observed in this study generally coincided with the plasma sera observed in this study has been shown to be an useful
concentration-time profiles of dihydroartemisinin measured alternative tool for evaluating and/or comparing the bioequiv-
using HPLC-ED. Dihydroartemisinin concentrations in plas- alence of interesting drugs, especially in laboratories where
ma were detectable in the first sampling (30 minutes after facilities for drug analysis by HPLC-ED are limited. Addi-
drug administration) with all three oral formulations. The tionally, the determination of bioequivalence based on anti-
peak antimalarial activities of sera were observed during the malarial activity using ex vivo experiment is much simpler and
time to reach the peak plasma concentrations of dihydroar- cheaper than investigating pharmacokinetic characteristics of
temisinin with all three oral formulations. the drug using the HPLC-ED method.
The present study has demonstrated the applicability of the The present study clearly demonstrated the influence of
ex vivo model in evaluating the bioequivalence of the three inoculum sizes of P. falciparum on ex vivo blood schizonto-
oral formulations of dihydroartemisinin based on antimalarial cidal activities of sera obtained following the administration
activity. The results, in conjunction with previous findings, of the three oral formulations of dihydroartemisinin. By in-
suggest that the pharmaceutical preparations and/or formu- creasing inoculum sizes from 1% to 5% parasitemia, the
lations have an effect on antimalarial activity and pharmaco- MIDs of sera were decreased at most time intervals with vary-
kinetic profiles of the drug.12,13,32 Evaluation of bioequiva- ing degrees depending on the parasite strains and the drug
lence of various drug formulations in the previous studies formulations. The observation of an inoculum effect in the
present study was consistent with the effect reported in
relied on statistical analysis of their pharmacokinetic param-
in vitro studies.14–20 The results of the inoculum effect on
eters. In the present study, bioequivalence of the three oral
antimalarial activities of sera containing dihydroartemisinin
formulations of dihydroartemisinin was based on statistical
have emphasized the importance of inoculum size on the
analysis of ex vivo blood schizontocidal activity profiles using
ex vivo drug sensitivity. These findings might have a critical
the Mann-Whitney U test. The results observed in this study
role in considering drug regimens for patients who have high
clearly indicate that the formulation manufactured by Arenco
parasitemia levels, if the hypothesis of an inoculum effect is
n.v. Pharmaceutica was not bioequivalent to Cotecxin威 and to
shown to be true. At the present time, it is still uncertain
the formulation manufactured by the Faculty of Pharmacy of
whether drug regimens for patients who have high parasit-
Mahidol University. As shown in Table 1, the formulation emia levels should be adjusted. Many other factors, which
manufactured by Arenco n.v. Pharmaceutica showed signifi- are important in vivo but cannot be investigated ex vivo or
cantly lower ex vivo antimalarial activity profiles (maximum in vitro, may contribute to and/or have an influence on evalu-
MIDs, time to maximum MIDs, and AUD) and pharmacoki- ating the inoculum effect. Such factors, as reported by Gluz-
netic parameters (Cmax, tmax, and AUC) than Cotecxin威 and man and others, include the role of cell-mediated and hu-
the formulation manufactured by the Faculty of Pharmacy moral immunity and the mechanical role of the spleen in
of Mahidol University. These results indicate the lower rate removing parasite from the blood.15 Whether the inoculum
(represented by Cmax and tmax) and extent (represented by effect occurs in vivo has yet to be examined. The ex vivo
AUC) of absorption of the dihydroartemisinin formulation model used for determination of inoculum effect on dihy-
manufactured by Arenco n.v. Pharmaceutica when compared droartemisinin activity observed in the present study might be
with the other two oral formulations. The dihydroartemisinin superior to the in vitro model since an importance factor
formulation manufactured by the Faculty of Pharmacy of (pharmacokinetics) was considered. Nonetheless, the deter-
Mahidol University showed significantly higher maximum mination of ex vivo drug sensitivity of malarial parasites
MID, Cmax, and tmax values than those of Cotecxin威. How- should be conducted using standardized parasitemias to avoid
ever, the median AUD and AUC values observed with the the effect of inoculum size.
dihydroartemisinin formulation manufactured by the Faculty
of Pharmacy of Mahidol University were not significantly dif- Received January 15, 2004. Accepted for publication June 22, 2004.
ferent from those of Cotecxin威. These results indicate that
Authors’ addresses: Monticha Kongthaisong and Peerapan Tan-
dihydroartemisinin formulation manufactured by the Faculty ariya, Department of Microbiology, Faculty of Sciences, Mahidol
of Pharmacy of Mahidol University showed a greater rate of University, Rama VI Road, Bangkok 10400, Thailand. Kesara Na-
drug absorption than that of Cotecxin威; however, the extent Bangchang, Department of Medical Technology, Faculty of Allied
710 KONGTHAISONG AND OTHERS

Health Sciences, Thammasat University, Klongluang, Pathumthani to chloroquine and its experimental testing in vitro. Biochem
12121, Thailand. Mathirut Mungthin, Department of Parasitology, Pharmacol 40: 685–691.
Phramongkutklao College of Medicine, 315 Ratchawithi Road, 17. Bray PG, Hawley SR, Ward SA, 1996. 4-Aminoquinoline resis-
Ratchathewi, Bangkok 10400, Thailand, Telephone and Fax: 66-2- tance of Plasmodium falciparum: insights from the study of
354-7761, E-mail: [email protected]. Nuttanan Sinchaipanid, De- amodiaquine uptake. Mol Pharmacol 50: 1551–1558.
partment of Pharmacognosy, Faculty of Pharmacy, Mahidol Univer- 18. Ritchie GY, Mungthin M, Green JE, Bray PG, Hawley SR, Ward
sity, Sri-Ayuthaya Rd., Bangkok 10400, Thailand. SA, 1996. In vitro selection of halofantrine resistance in Plas-
modium falciparum is not associated with increased expression
of Pgh1. Mol Biochem Parasitol 83: 35–46.
REFERENCES 19. Hawley SR, Bray PG, O’Neill PM, Naisbitt DJ, Park BK, Ward
SA, 1996. Manipulation of the N-alkyl substituent in amodi-
1. McIntosh HM, Olliaro P, 2000. Artemisinin derivatives for treat- aquine to overcome the verapamil-sensitivity chloroquine re-
ing uncomplicated malaria. Cochrane Database of Systemic sistance component. Antimicrob Agents Chemother 1996:
Reviews. 2345–2349.
2. McIntosh HM, Olliaro P, 2000. Artemisinin derivatives for treat- 20. Duraisingh MT, Jones P, Sambou I, von Seildlein L, Pinder M,
ing severe malaria. Cochrane Database of Systemic Reviews. Warhurst DC, 1999. Inoculum effect leads to overestimation of
3. Luo XD, Shen CC, 1987. The chemistry, pharmacology, and clini- in vitro resistance for artemisinin derivatives and standard an-
cal applications of qinghaosu (artemisinin) and its derivatives. timalarials: a Gambian field study. Parasitology 119: 435–440.
Med Res Rev 7: 29–52. 21. Na-Bangchang K, Congpoung K, Hung LN, Molunto P, Karb-
4. Alin MH, Bjorkman A, Ashton M, 1990. In vitro activity of ar- wang J, 1998. Simple high-performance liquid chromatograph-
temisinin, its derivatives, and pyronaridine against different ic method with electrochemical detection for the simultaneous
strains of Plasmodium falciparum. Trans R Soc Trop Med Hyg determination of artesunate and dihydroartemisinin in biologi-
84: 635–637. cal fluids. J Chromatogr B Biomed Sci Appl 708: 201–207.
5. Basco LK, Le Bras J, 1993. In vitro activity of artemisinin deriva- 22. Rieckmann KH, Campbell GH, Sax LJ, Mrema JE, 1978. Drug
tives against African isolates and clones of Plasmodium falci- sensitivity of Plasmodium falciparum and in vitro microtech-
parum. Am J Trop Med Hyg 49: 301–307. nique. Lancet 1: 22–23.
6. Le Bras J, 1998. In vitro susceptibility of African Plasmodium 23. Tan-ariya P, Brockelman CR, Menabadhu C, 1987. Optimal con-
falciparum isolates to dihydroartemisinin and the risk factors centration of p-aminobenzoic acid and folic acid in the in vitro
for resistance to qinghaosu. Med Trop (Mars) 58 (Suppl): 18– assay of antifolates against Plasmodium falciparum. Am J Trop
21. Med Hyg 37: 42–48.
7. Klayman DL, 1985. Qinghaosu (artemisinin): an antimalarial 24. Gibaldi M, 1991. Biopharmaceutics and Clinical Pharmacokinet-
drug from China. Science 228: 1049–1055. ics. Fourth edition. London: Lea and Febiger, 14–23.
8. Li GQ, Guo XB, Fu LC, Jian HX, Wang HX, 1994. Clinical trials 25. Geary TG, Divo AA, Jensen JB, 1989. Stage specific actions of
of artemisinin and its derivatives in the treatment of malaria in antimalarial drugs on Plasmodium falciparum in culture. Am
China. Trans R Soc Trop Med Hyg 88 (Suppl 1): S5–S6. J Trop Med Hyg 40: 240–244.
9. Looareesuwan S, Wilairatana P, Vanijanonta S, Pitisuttithum P, 26. ter Kuile F, White NJ, Holloway P, Pasvol G, Krishna S, 1993.
Viravan C, Kraisintu K, 1996. Treatment of acute, uncompli- Plasmodium falciparum: in vitro studies of the pharmacody-
cated, falciparum malaria with oral dihydroartemisinin. Ann namic properties of drugs used for the treatment of severe
Trop Med Parasitol 90: 21–28. malaria. Exp Parasitol 76: 85–95.
10. Le NH, Na-Bangchang K, Le TD, Thrinh KA, Karbwang J, 1999. 27. Alin MH, Bjorkman A, 1994. Concentration and time depen-
Pharmacokinetics of a single oral dose of dihydroartemisinin in dency of artemisinin efficacy against Plasmodium falciparum
Vietnamese healthy volunteers. Southeast Asian J Trop Med in vitro. Am J Trop Med Hyg 50: 771–776.
Public Health 30: 11–16. 28. Maeno Y, Toyoshima T, Fujioka H, Ito Y, Meshnick SR, Benakis
11. Na-Bangchang K, Congpoung K, Ubalee R, Thanavibul A, Tan- A, Milhous WK, Aikawa M, 1993. Morphologic effects of ar-
ariya P, Karbwang J, 1997. Pharmacokinetics and ex vivo anti- temisinin in Plasmodium falciparum. Am J Trop Med Hyg 49:
malarial activity of sera following a single oral dose of dihy- 485–491.
droartemisinin in healthy Thai males. Southeast Asian J Trop 29. Na-Bangchang K, Tippawangkosol P, Thanavibul A, Ubalee R,
Med Public Health 28: 731–735. Karbwang J, 1999. Pharmacokinetic and pharmacodynamic in-
12. Na-Bangchang K, Karbwang J, Congpoung K, Thanavibul A, teractions of mefloquine and dihydroartemisinin. Int J Clin
Ubalee R, 1998. Pharmacokinetic and bioequivalence evalua- Pharmacol Res 19: 9–17.
tion of two generic formulations of oral artesunate. Eur J Clin 30. Ubalee R, Songthammawat D, Na-Bangchang K, Tan-ariya P,
Pharmacol 53: 375–376. Karbwang J, 1999. Ex vivo blood schizontocidal activities of
13. Na-Bangchang K, Karbwang J, Palacios PA, Ubalee R, Saeng- artemisinin derivatives against Plasmodium falciparum. South-
tertsilapachai S, Wernsdorfer WH, 2000. Pharmacokinetics east Asian J Trop Med Public Health 30: 225–231.
and bioequivalence evaluation of three commercial tablet for- 31. Karbwang J, Na-Bangchang K, 1992. Principles and planning of
mulations of mefloquine when given in combination with di- clinical pharmacokinetic studies. Karbwang J, Back DJ, Breck-
hydroartemisinin in patients with acute uncomplicated falci- enridge D, eds. Clinical Pharmacology in Parasitic Diseases.
parum malaria. Eur J Clin Pharmacol 55: 743–748. Bangkok: Thai Wattana Panich, 65–71.
14. Ponnudurai T, Leewenburg ADEM, Meuwissen HET, 1982. 32. Karbwang J, Na-Bangchang K, Congpuong K, Molunto P, Tha-
Chloroquine sensitivity of isolates of Plasmodium falciparum navibul A, 1997. Pharmacokinetics and bioavailability of oral
adapted in in vitro culture. Trop Geogr Med 33: 50–54. and intramuscular artemether. Eur J Clin Pharmacol 52: 307–
15. Gluzman IY, Schlesinger PH, Krogstad DJ, 1987. Inoculum effect 310.
with chloroquine and Plasmodium falciparum. Antimicrob 33. Food and Drug Administration, 1992. FDA Guidance: Statistical
Agents Chemother 31: 32–36. Procedures for Bioequivalence Studies Using a Standard Two-
16. Geary TG, Divo AD, Jensen JB, Zangwill M, Ginsburg H, 1990. Treatment Cross-Over Design. Washington, DC: Food and
Kinetic modelling of the response of Plasmodium falciparum Drug Administration.