THE JOURNAL OF INFECTIOUS DISEASES • VOL. 150, NO.

6 • DECEMBER 1984
© 1984 by The University of Chicago. All rights reserved. 0022-1899/84/5006-0016$01.00

Clindamycin Activity Against Chloroquine-Resistant Plasmodium falciparum

Lavanya S. Seaberg, Arlene R. Parquette, From the Divisions of Infectious Diseases and
I1ya Y. Gluzman, Grady W. Phillips, r-, Laboratory Medicine, Departments of Medicine and
Pathology, and The Center for Basic Cancer Research,
Thomas F. Brodasky, and Donald J. Krogstad
Washington University School of Medicine,
St. Louis, Missouri; and The Upjohn Company,
Kalamazoo, Michigan

The clindamycin dose-response curves observed with both chloroquine-resistant and

chloroquine-susceptible strains of Plasmodium falciparum in vitro demonstrated a
plateau region that extended from 10-2 to 101 jJg/ml of drug (22 nM to 22 ~. Similar
dose-response curves were also observed with the three major metabolites of clindamycin

Downloaded from http://jid.oxfordjournals.org/ at New York University on September 10, 2015

(clindamycin sulfoxide, de-N-methyl clindarnycin, and de-N-methyl clindamycin sulfox-
ide). The position of this plateau was time dependent and rose from 20070-25% to
90%-95% inhibition of parasite growth between 24 and 72 hr of exposure to the drug. Clin-
damycin treatment reduced plasmodial protein and nucleic acid synthesis (as measured
by the incorporation of [3H]isoleucine and [JH]hypoxanthine, respectively) but 'did not
interfere with knob formation. The combination of quinine plus a fixed concentration of
clindamycin (0.1 jJg/ml) inhibited growth of the quinine-resistant Indochina I strain,
although most of the antiplasmodial activity observed at quinine concentrations <50
ng/rnl (154 nM) could be attributed to clindamycin alone.

Malaria continues to have an impact on world
health that is difficult to overestimate, with more
than 100 million cases and one million deaths each
year [1]. Of the four plasmodia that infect humans,
Plasmodium falciparum poses the greatest risk of
death because of its ability to produce overwhelm-
ing parasitemia (~1 0 6 parasitized red cells/iii) [2]
and because many P. falciparum strains are resis-
tant to chloroquine [3].
The increasing prevalence of chloroquine resis-
tance (in Southeast Asia, South America, and Afri-
ca) has led to greater use of the pyrimethamine plus
sulfadoxine combination (Fansidar"; Roche,
Nutley, NJ) and of quinine for the treatment of pa-
Received for publication April 23, 1984, and in revised form
July 21, 1984.
This work was presented in part at the 32nd annual meeting of
the American Society of Tropical Medicine and Hygiene in San
Antonio, Texas, December 1983.
This work was supported in part by a grant from The Upjohn
Company.
We thank Richard L. Westerman, William L. Lummis, and
William L. Stevens for their suggestions and assistance; Phuc
Nguyen-Dinh, C. C. Campbell, Ayoade M. F. Oduola, and
Robert E. Desjardins for providing the test strains used in these
studies; and Robert E. Desjardins for his thoughtful review of
the manuscript.
Please address requests for reprints to Dr. Donald J.
Krogstad, Division of Laboratory Medicine, Washington
University School of Medicine, St. Louis, Missouri 63110.
tients with chloroquine-resistant P. falciparum in-
fections. However, resistance to pyrimethamine
plus sulfadoxine has recently been reported from
Africa, as well as from South America and South-
east Asia [4] and now limits the use of that com-
bination. In addition, the use of quinine is limited
by its toxicity [5], its cost, and - to some ex-
tent - by the recent development of quinine resis-
tance in Indochina [6]. The quinine plus tetracy-
cline combination is now used commonly in South-
east Asia, and mefloquine is effective against
many, but not all, of the chloroquine-resistant
strains in that region [7]. Thus, there remains an
urgent and unfilled need for new effective anti-
malarials for the treatment of patients with multi-
ply resistant P. falciparum infection [8].
Clindamycin was first shown to have antimala-
rial activity in 1967 by Lewis [9], who studied the
murine malaria parasite Plasmodium berghei.
Subsequently, Powers [10] and Schmidt et al. [11]
demonstrated that clindamycin was also active
against Plasmodium cynomolgi and chloroquine-
resistant P. falciparum infections [12] in primates.
More recently, Geary and Jensen [13] reported that
clindamycin was active against the FCR-3 strain of
P. falciparum in vitro (on the basis of quantitative
parasite counts), and its efficacy in the treatment
of human P. falciparum infection was tested in
areas with known chloroquine resistance at the

904
Antiplasmodial Activity ofClindamycin
R-II and R-III level, such as the Philippines
[14-16] and Indochina [17]. In those studies, the
efficacy of clindamycin as an antimalarial was
more closely related to the duration of treatment
(95070-100070 of patients recovered after five to
seven days of therapy vs. 50%-70% after three
days of therapy [14-17]) than it was to the dose
(oral regimens containing 300-2,400 mg/day pro-
duced similar results when given for five to seven
days [14, 15]). Therefore, we undertook these in
vitro studies to determine the following: (1) the
dose-response relation between clindamycin con-
centration and its antiplasmodial activity, (2)
whether the metabolites of clindamycin were also
active against the parasite, (3) whether clindamycin
resistance emerged readily in vitro, (4) to examine
potential mechanisms of clindamycin action, and
(5) to define the in vitro activity of antimalarial
combinations containing clindamycin plus quinine.
Materials and Methods
Parasite strains and culture conditions. The
chloroquine-resistant Indochina I1CDC and chlo-
roquine-susceptible Honduras I/CnC strains of
R jalciparum used for these studies were provided by
Drs. P. Nguyen-Dinh and C. C. Campbell of the
Centers for Disease Control (Atlanta). By the
PH]hypoxanthine assay developed by Desjardins
et al. [18], the ED so (effective dose) of chloroquine
for these strains was 100 and 3 ng/ml, respectively.
The Indochina I strain was also quinine resistant in
vitro with a quinine ED so of 135 ng/ml. The W-2
clone (derived from the Indochina III strain) [19]
was provided by Drs. A. M. J. Oduola and R. E.
Desjardins of the Department of Parasitology,
School of Public Health, University of North
Carolina at Chapel Hill.
Parasites were grown in the culture system
developed by Trager and Jensen [20] in an atmos-
phere with reduced oxygen in modular incubation
chambers (Billups-Rothenberg, Del Mar, Calif).
The oxygen tension for susceptibility testing was
achieved by gassing the chambers for 5 min with
gas of known composition (3% O 2 , 3070 CO lJ and
balance nitrogen [Union Carbide Linde Division,
New York]) [21].
Susceptibility testing. Susceptibility testing
was performed as previously described [18, 21]. A
2% red-cell suspension with 0.2070 parasitemia was
placed in 24-well tissue culture plates (CoStar,
905
Cambridge, Mass) and incubated overnight with-
out the addition of drug. The next two days (at
zero time and at + 24 hr) the culture medium was
changed and replaced with medium containing the
desired concentrations of clindamycin. Clin-
damycin and its metabolites were purified [22] and
characterized [23] by thin-layer chromatography
before their use in these studies. On the third day
(at + 48 hr), the fresh medium added to the wells
of the tissue culture plates contained 1.33 times the
desired final clindamycin concentrations. Aliquots
(150 /-ll) taken from the wells of the tissue culture
Downloaded from http://jid.oxfordjournals.org/ at New York University on September 10, 2015
plates were then added to the wells of microtiter
plates (Linbro-Titertek; Flow Laboratories, Me-
Lean, Va) containing 50 /-ll of culture medium with
[3H]hypoxanthine (specific activity of 1 Ci/rnmol;
Amersham, Arlington Heights, Ill). This protocol
produced the desired range of clindamycin concen-
trations with 0.5 /-lCi of PH]hypoxanthine in each
microtiter well. After overnight incubation the
contents of the microtiter wells were then harvested
(at + 66 hr) onto 11 mm glass-fiber filter-paper
disks (Reeve Angel 934; Whatman, Clifton, NJ)
with a Titertek Cell Harvester (Flow Laboratories).
After they were thoroughly washed with glass-
distilled water and dried, the glass- fiber filter-
paper disks were placed in glass scintillation vials
(Wheaton Scientific, Millville, NJ) containing 8 ml
of Ready Solv Mp® (Beckman, Fullerton, Calif)
and counted in an LS-8000 liquid scintillation
counter (Beckman). The anti plasmodial activity of
the clindamycin concentrations tested was defined
by their inhibition of PH]hypoxanthine uptake
relative to drug-free controls. On the basis of the
observations of Sherman (who showed that para-
sitized red cells incorporated isoleucine during pro-
tein synthesis) [24], the uptake of PH]isoleucine
was used to estimate plasmodial protein synthesis.
Thus, the effects of different clindamycin concen-
trations on plasmodial protein synthesis were
estimated by their effect on pHisoleucine uptake
with 0.5 /-lCi of ['Hltsolcuclne per microtiter well in
50 /-ll of culture medium (4, 5- 3H-isoleucine, 110
Ci/mmol; New England Nuclear, Boston).
Morphological assessment of drug activity was
based on quantitative parasite counts and was per-
formed with a similar protocol [21]. However, at
+ 48 hr, the medium added to the microtiter wells
contained no PH]hypoxanthine. In the morpho-
logical assays, antiplasmodial activity was defined
by the ability of the clindamycin concentrations
906
tested to reduce the parasite counts observed after
66 hr of exposure to drug (compared with drug-
free controls).
The anti plasmodial activity of the clindamycin
plus quinine combination was defined by its ability
to inhibit [3H]hypoxanthine uptake by the parasite
by use of the protocol described above for the
testing of clindamycin and its metabolites.
Electron microscopy. For determination of the
effect of clindamycin on parasite maturation (espe-
cially on the formation of knobs in the trophozoite
stage), the gelatin technique of Jensen [25] was
used to prepare red-cell suspensions enriched with
trophozoites and schizonts. After overnight in-
cubation with fresh red cells in drug-free medium,
these cultures (in contrast to the [3H]hypoxanthine
and [3H]isoleucine studies that were performed
with asynchronous cultures) contained primarily
ring stages of the parasite and were placed in
culture medium containing clindamycin for an ad-
ditional30 hr of incubation. At that time, the con-
trol (clindamycin-free) cultures contained pre-
dominantly trophozoites, and both the treated and
control cultures were harvested for examination by
electron microscopy. These parasitized red-cell
suspensions were initially fixed in 1.25070 glu-
taraldehyde (pH 7.3) with 0.1 M sodium caco-
dylate and 0.116 M sucrose and, subsequently, in
0.2 M sodium cacodylate (pH 7.2) with 1070 os-
mium tetroxide [26, 27]. After staining the suspen-
sions with 0.5070 uranyl magnesium acetate, thin
sections were obtained with an LKB IV microtome
and examined with a Philips 201 electron micro-
scope (Einhoven, The Netherlands).
Statistical methods. Standard methods were
used to calculate the mean and standard deviation
(table 1) [28]. Results presented in the figures are
from individual experiments (based on the mean
percent inhibition of ['Hlhypoxanthlne or [3H]iso-
leucine uptake observed with three or more deter-
minations for each drug concentration tested),
although each experiment was performed two or
more times under identical conditions.
Results
Antiplasmodial activity of clindamycin. Quan-
titative parasite counts after 66 hr of exposure to
clindamycin suggested a plateau in the dose-
response curve from ~0.1 ug/rnl to 30 ug/rnl (0.22
J..lM to 66 J..lM; table 1). Concentrations>100 J,.lg/ ml
Seaberget al.
Table 1. Antiplasmodial activity of clindamycin against
the Indochina I strain of P. falciparum.
Clindamycin Percent
concentration Percentage of parasitized reduction in
(ug/rnl) red blood cells (mean ± SO) parasite count
o 9.58 ± 0.17
0.1 6.48 ± 0.41 32.4
0.3 5.02 ± 0.45 47.6
I 4.17 ± 0.22 56.2
3 3.62 ± 0.33 62.2
10 3.09 ± 0.25 67.7
30 2.49 ± 0.15 74.0
100 0.98 89.8
Downloaded from http://jid.oxfordjournals.org/ at New York University on September 10, 2015
300 No recognizable parasites, 100
obvious red blood cell
damage
1,000 No recognizable parasites, 100
gross hemolysis
NOTE. The percent parasitemia was determined by examin-
ing at least 1,000 red blood cells on ~3 Giemsa-stained slides
from each of three microtiter wells at each drug concentration.
All recognizable parasites were counted. Quantitative parasite
counts were performed after 66 hr of exposure to clindamycin.
produced hemolysis and obviously damaged the re-
maining red blood cells. Although examination of
synchronous cultures at the trophozoite stage
revealed evidence of vacuolation and decreased
pigment formation, neither light nor electron
microscopy demonstrated stage-specific inhibition
of parasite maturation by clindamycin.
The radiometric assay (based on the inhibition
of [3H]hypoxanthine uptake) permitted the testing
of a larger number of drug concentrations. It also
defined a plateau in the dose-response curve with
the chloroquine-resistant Indochina I strain that
extended from 10-2 to 101 J,.lg ofclindamycin/ml (22
nM to 22 J..lM; figure 1). Similar results were also
observed with the W-2 clone obtained by Oduola et
al. [19] from the Indochina III strain and with the
chloroquine-susceptible Honduras I strain (data
not shown).
Because clinical trials had suggested that the
antimalarial activity of clindamycin in vivo was
more dependent on the duration of treatment than
on the dose [14-17], we examined the effect of an
increased duration of exposure to clindamycin on
its antiplasmodial activity in vitro with the In-
dochina I strain. The results of those studies in-
dicated that the activity of the drug increased
markedly with time. The percent inhibition of
[3H]hypoxanthine uptake observed in the plateau
 Antiplasmodial Activity ofClindamycin 907

LU
100 region increased from 20%-25010 to 90070-95% be-
u; ~
tween 42 and 90 hr of incubation (i.e., after 24-72
0« I- hr of exposure to drug before the addition of
za.... 75 [3H]hypoxanthine), although the MIC did not
o~
i=LU change; it remained at 10-2 ug/rnl (22 nM; figure 2).
diZ Antiplasmodial activity of c1indamycin metab-
II
ZI-
-z 50 olites. To determine whether the in vivo anti-
1-« malarial activity of clindamycin resulted from the
Zx parent compound or one of its several metabolites,
LUo
ua....
01:::>- 25 we examined the antiplasmodial activity of clinda-
~I mycin sulfoxide, de-N-methyl clindamycin, and
I de-N-methyl clindarnycin sulfoxide in vitro with

Downloaded from http://jid.oxfordjournals.org/ at New York University on September 10, 2015

M
I
I
the radiometric assay. These studies revealed
-7 -5 -3 -1 + 1 +3 similar dose-response curves (with plateau regions
LOGlo CLiNDAMYCIN extending from 10-1 to 101 ug/rnl, 22 nM to 221lM>
CONCENTRATION (lJg/ml) for each of these three metabolites (figure 3).
Figure 1. Clindamycin concentrations of 10-2 to 101 Development of c1indamycin resistance in vitro.
ug/rnl (22 nM-221JM) produced a 45070-50070 inhibition To test whether preexisting clindamycin resistance
of parasite growth (as measured by [3H]hypoxanthine was present or would emerge readily in vitro, we
uptake) with the chloroquine-resistant Indochina I strain cultured the Indochina I strain in the highest con-
after 48 hr of exposure to drug before the addition of centration of clindamycin that permitted parasite
[3H]hypoxanthine. Similar results were also observed with
the W-2 clone derived from the Indochina III strain by growth (10- 3 /lg/ml, 2.2 nJ\.1), as determined by
Oduola et al. [18] and with the chloroquine-susceptible quantitative parasite counts of cultures continu-
Honduras I strain (data not shown). ously exposed to clindamycin for more than seven
days. Parasites were then examined for the emer-
gence of clindamycin resistance by two methods: (1)
subculture into media containing greater concen-
traions of clindamycin (2-5 x 10-3 ug/ml, 4.5-10
W
100 nJ\.1), and (2) repeated radiometric susceptibility
LL.~
O<{
testing with clindamycin and PH] hypoxanthine.
..... After continuous in vitro exposure to clindamy-
6~ 75 cin (10- 3 ug/rnl 2.2 nJ\.1) for as long as 10 weeks
i=w (two subcultures each week), the parasites were still
ceZ unable to replicate in media containing> 10-3 /lg of

Zz
II

..... <{
50
clindamycin/ml (>2.2 nJ\.1). Similarly, radiometric
susceptibility testing revealed no significant in-
crease in the resistance of the parasite to clindamy-
Z>< cin; the position of the plateau was unchanged and
wO
~~ 25 the steep portions of the dose-response curve re-
wI
0..
mained between 10-3 to 10-2 and 101 to 102 ug/rnl
I
CO')
(2.2-22 nM and 22-220 1lM>, respectively (data not

o ,t , shown).
-2 -1 0 +1 +2 +3 Potential mechanisms of c1indamycin action.
Because several observers have reported that some
LOGlO CLiNDAMYCIN patients treated with clindamycin actually have
CONCENTRATION (~g/ml)
greater parasitemia on the second or third day of
Figure 2. Increasing the duration of exposure to clinda- treatment [14, 15], we also tested the possibility
mycin (before the addition of [3H]hypoxanthine) from that clindamycin might inhibit knob formation
24 hr to 72 hr progressively increased the percent inhibi-
tion of [3H]hypoxanthine uptake observed in the plateau [29]. This inhibition could increase the parasite
region (10-2 to 101 /-lg/ml) from 20070-25010 to 90%-95% count in the peripheral blood by interfering with
with the Indochina I strain. the adherence of red blood cells containing tropho-
908 Seaberget al.

w 10~ 100
~
«
~ 75
w INDOCHINA I
c, ~
~
w z«
z 50
I
at
t=::J
75
~
z 25 caW
~«
ZX -z
I-
wO
u~ Clindornycin Clindornvcin Sulohoxide zU
-::J 50
ffi:::C
~ I
100
~w
I
M
u,
0 75 w a
Z""'"
z U~
0:::'
0
f=
50 wI 25
iii
a.. M
25 LL.

Downloaded from http://jid.oxfordjournals.org/ at New York University on September 10, 2015

I
~
a
-7 -5 -3 -1 +1 +3 -7 -5 -3 -1 +1 +3
De-N-Methyl C1indamycin De-N-Methyl Clmdamycln Sulphoxide -7 -5 -3 -1 +1 +3
LOG 10 ANTIMICROBIAL CONCENTRATION (lJg/ml)
LOGlO CLiNDAMYCIN
Figure 3. By the radiometric assay, the inhibition of CONCENTRATION (~g/ml)
['Hlhypoxanrhine uptake produced by the three major
metabolites of clindamycin (clindamycin sulfoxide, de- Figure 4. The effects of clindamycin on protein syn-
N-methyl clindamycin, and de-N-methyl clindamycin sul- thesis (as measured by PH]isoleucine incorporation by
foxide) with the Indochina I strain was virtually identical the Indochina I strain) were similar to its effects on nucleic
to that observed with clindamycin. Cultures were exposed acid synthesis (figure 1). Cultures were exposed to clinda-
to drugs for 48 hr before the addition of PH]hypo- mycin for 48 hr before the addition of PH]isoleucine.
xanthine.
Antiplasmodial activity of the clindamycin plus
quinine combination. Because treatment with
zoites or schizonts to the peripheral microvascula- clindamycin alone produces a relatively slow
ture. However, comparison of clindamycin-treated clearance of parasites from the blood in vivo
cultures with untreated controls indicated that the [14-16], several investigators have proposed using
drug did not interfere with knob formation. Trans-
mission electron microscopy revealed knobs on the
surfaces of trophozoite- and schizont-containing 100
red blood cells after treatment with clindamycin
that were indistinguishable from those observed in
untreated controls (data not shown). Likewise,
Quinine
similar trophozoite- and schizont-rich fractions +
Clindamycin
were obtained from both control (untreated) and
50 (O.1IJg / ml) .......
clindamycin-treated cultures of the Indochina I
strain by the gelatin technique [25], which is effec-
tive only with knob-positive strains (data not
shown). Thus, at present, there is no clear explana-
tion for the increased parasitemia observed in some
patients two to three days after beginning treat-
ment with clindamycin.
Because clindamycin is known to inhibit bac-
terial protein synthesis [30], we also examined its
effect on plasmodial protein synthesis on the basis
of [3H]isoleucine incorporation by the parasite
[24]. These studies revealed that the effects of clin- Figure S. The addition of quinine did not significantly
damycin on protein synthesis were parallel to its ef- increase the antiplasmodial activity of clindamycin (0.1
ug/ml) against the chloroquine- and quinine-resistant
fects on nucleic acid synthesis, as measured by the Indochina I strain (at quinine concentrations of ~30
inhibition of PH]isoleucine and ['Hlhypoxanthine ng/ml). Cultures were exposed to drug for 48 hr before
incorporation, respectively (figures 4 and 1). the addition of [3H]hypoxanthine.
A ntiplasmodial A ctivity ofClindamycin
clindamycin plus quinine for the treatment of resis-
tant strains. Therefore, we examined this combina-
tion in vitro with use of a quinine-resistant strain
(Indochina I). Because the clindamycin dose-
response curve is flat in the range of clinical in-
terest (from 10-2 to 101 ug/ml, 22 nMto 22 lAM), we
tested the effects of clindamycin on the quinine
dose-response curve with a fixed clindamycin con-
centration of 0.1 IAg/ml (220 nM). Quinine alone
had no effect at concentrations ~50 ng/ml (~154
nM) against this strain, which has a quinine ED so
of 135 ng/ml. In addition, most of the antiplas-
modial activity observed with the combination
could be accounted for by the 0.1 ug/rnl of clin-
damycin alone, which reduced [3H]hypoxanthine
uptake by 30070 in the absence of quinine (figure 5).
Discussion
The plateau-shaped dose-response curves observed
with clindamycin (figures 1 and 2) are consistent
with the results of the clinical trials in the Philip-
pines and Indochina (i.e., the outcomes observed
in those trials were more dependent on the dura-
tion of therapy than on the dose) [14-17]. They are
also consistent with the greater sensitivity of the
[3H]hypoxanthine assay system vs. morphology
(figure 1 vs. table 1) and with the earlier observa-
tions of Powers et al. [l0, 12], who reported that
the rate of parasite clearance was not proportional
to the dose in primates treated with clindamycin
for P. cynomolgi or chloroquine-resistant P. fal-
ciparum infections. However, the shape of these
curves is distinctly unusual. To our knowledge, the
known antimalarials and other compounds with
antiplasmodial activity all have single-step dose-
response curves (i.e., their antiplasmodial activity
is proportional to the log., of their concentration
over the critical range of the dose-response curve)
[18,21,31]. Potential explanations for the unusual
(plateau) shape of the clindamycin dose-response
curve include the following: (1) a mixture of strains
with different susceptibilities to the drug, (2) more
than one mechanism of drug action, which would
also be consistent with the hemolysis observed at
clindamycin concentrations ~100 IAg/ml (table 1),
and (3) action on a pathway or process whose in-
hibition has a slowly increasing (cumulative) effect
on parasite viability.
Two lines of evidence suggest that the plateau-
shaped dose-response curve observed with clin-
damycin did not result from a mixture of strains. It
909
was, in fact, observed with two different isolates
(Indochina I and Honduras I), both of which have
demonstrated single-step dose-response curves
with the other antimalarials that have been tested
[21], and with a clone derived from the Indochina
III strain by Oduola et al. [19]. Retesting after pro-
longed exposure to clindamycin in vitro always
yielded the same (plateau-shaped) dose-response
curve rather than a single-step rise at the upper
range of these concentrations that would be
observed with a more resistant clone.
On the basis of the hemolysis observed at clinda-
Downloaded from http://jid.oxfordjournals.org/ at New York University on September 10, 2015
rnycin concentrations> 100 IAg/ml (>220 lAM> with
both infected and control (uninfected) red blood
cells, there must be at least two mechanisms of clin-
damycin action against the parasite: one that
causes red blood cell damage at the higher (non-
physiological) concentrations, and one or more ad-
ditional mechanisms in the clinically relevant range
(from 10-2 to 101 IAg/ml, 22 nM to 22 IAM>.
Although clindamycin did not inhibit knob for-
mation (as judged by transmission electron
microscopy and by the efficacy of the gelatin
technique), clindamycin-treated parasites did have
reduced protein synthesis (as measured by the in-
corporation of ['Hlisoleucine). Possible mecha-
nisms of clindamycin action consistent with these
observations include clindamycin inhibition of
plasmodial protein synthesis and other, as yet
undefined, effects. Although the ribosomes of
P. falciparum have not yet been characterized in
detail, previous studies with other eukaryotes sug-
gest that mitochondrial ribosomes are usually 70S
and susceptible to clindamycin, whereas cytoplas-
mic ribosomes are usually 80S and are generally
unaffected by clindamycin [30]. Examination of
clindamycin-treated parasites by light microscopy
revealed vacuolation of trophozoites and de-
creased pigment formation similar to observations
of Plasmodium knowlesi previously reported by
Powers et al. [31].
The similar in vitro antiplasmodial activities
observed with clindamycin and its metabolites
(figure 3) are different from the pattern that has
been observed with bacteria, against which de-N-
methyl clindamycin is distinctly more active than
are the other metabolites of clindamycin [22]. They
also suggest that metabolic transformation by the
liver, kidney, or other organs is not necessary for
the antimalarial activity of clindamycin in vivo.
Similarly, the lack of alterations in the clindamycin
dose-response curve after up to 10 weeks of ex-
910
posure to the drug in vitro suggests that the pla-
teau-shaped dose-response curve reflects the in-
trinsic response of the parasite to the drug.
The results observed with the clindamycin plus
quinine combination (figure 5) indicate that
quinine concentrations active in vitro against
chloroquine-susceptible strains (10-30 ng/ml) [18]
are inactive against the Indochina I (chloroquine-
and quinine-resistant) strain, unless combined
with clindamycin (0.1 JAg/mI). Although higher
clindamycin levels are normally achieved in vivo
(1-8 ug/rnl) [32], even 0.1 ug/ml of clindamycin
has greater activity against this strain than does
50-100 ng/ml of quinine. This suggests that the ad-
dition of quinine to clindamycin may not substan-
tially reduce the time required for parasite clear-
ance with clindamycin treatment of chloroquine-
and quinine-resistant P. falciparum infection.
These findings also suggest that the clindamycin
plus quinine combination is unlikely to be antag-
onistic in vivo. The clinical experience (of which
we are aware) with clindamycin plus quinine com-
binations [33, 34] is not sufficient to determine
whether those combinations are more effective
than clindamycin or quinine alone for the treat-
ment of multiply-resistant P. falciparum infec-
tions, although Clyde et al. [34] observed that the
parasite clearance times with clindamycin alone
were longer than those observed with the quinine
plus clindamycin combination.
Taken together, these results suggest the follow-
ing: (1) that the time dependence of the antimalar-
ial action of clindamycin results from the cumu-
lative effects of the drug on the parasite, (2) the an-
timalarial effects of clindamycin do not require
metabolic transformation of the drug in vivo, (3)
clindamycin-treated parasites have reduced protein
and nucleic acid synthesis, (4) resistance to the
antiplasmodial effects of clindamycin does not
emerge readily in vitro, and (5) the combination of
clindamycin plus quinine may not be more effec-
tive against quinine-resistant strains than is
clindamycin alone. However, additional clinical
studies will be necessary to define the role of clin-
damycin in the treatment of multiply-resistant
R falciparum infection in vivo.
References
l. Wyler DJ. Malaria-resurgence and research. N Engl J
Med 1983;308:875-8, 934-40
Seaberg et al.
2. Field JW. Blood examination and prognosis in acute fal-
ciparum malaria. Trans R Soc Trop Med Hyg 1949;
43:33-56.
3. Dixon KE, Williams RG, Pongsupat T, Pitaktong U,
Phintuyothin P. A comparative trial of mefloquine and
Fansidar in the treatment of falciparum malaria: failure
of Fansidar. Trans R Soc Trop Med Hyg 1982;
76:664-7.
4. Hess U, Timmermans PM, Jones M. Combined chloro-
quine/Fansidar-resistant falciparum malaria appears in
East Africa. Am J Trop Med Hyg 1983;32:217-20
5. White NJ, Warrell DA, Chanthavanich P, Looareesuwan
S, Warrell MJ, Path MRC, Krishna S, Williamson DH,
Turner RC. Severe hypoglycemia and hyperinsulinemia
Downloaded from http://jid.oxfordjournals.org/ at New York University on September 10, 2015
in falciparum malaria. N Engl J Med 1983;309:61-6
6. Chongsuphajaisiddhi T, Sabcharoen A, Attanath P. In
vivo and in vitro sensitivity of fa1ciparum malaria to
quinine in Thai children. Ann Trop Pediatr 1981;1:21-6
7. Boudreau EF, Webster HK, Pavanand K, Thosinga L.
Type II mefloquine resistance in Thailand [letter].
Lancet 1982;2:1335
8. Rieckmann KH. Fa1ciparum malaria: the urgent need for
safe and effective drugs. Annu Rev Med 1983;34:
321-35
9. Lewis C. Antiplasmodial activity of halogenated lincomy-
cin analogues in Plasmodium berghei-infected mice. An-
timicrob Agents Chemother 1967;1968:537-42
10. Powers KG. Activity of chlorinated lincomycin ana-
logues against Plasmodium cynomolgi in rhesus
monkeys. Am J Trop Med Hyg 1969;18:485-90
II. Schmidt LH, Harrison J, Ellison R, Worcester P. The
activities of chlorinated lincomycin derivates against in-
fections with Plasmodium cynomolgi in Macaca mulat-
tao Am J Trop Med Hyg 1970;19:1-11
12. Powers KG, Jacobs RL. Activity of two chlorinated lin-
comycin analogues against chloroquine-resistant falci-
parum malaria in owl monkeys. Antimicrob Agents
Chemother 1972;1:49-53
13. Geary TG, Jensen J8. Effect of antibiotics on Plasmo-
dium falciparum in vitro. Am J Trop Med Hyg 1983;
32:221-5
14. Cabrera BD, Rivera DG, Lara NT. Study on clindamy-
cin in the treatment of fa1ciparum malaria. Rev Inst Med
Trop Sao Paulo 1982;24(Suppl 6):62-9
15. Rivera DG, Cabrera BD, Lara NT. Treatment of falcipa-
rum malaria with clindamycin. Rev Inst Med Trop Sao
Paulo 1982;24(Suppl 6):70-5
16. Alecrim MdG, Dourado H, Alecrim W. Albuquerque BC,
Wanssa E, Wanssa MdC. Tratamento da malaria (P. fal-
ciparum) con clindamicina. Rev Inst Med Trop Sao
Paulo 1981;23:86-91
17. Hall AP, Doberstyn EB, Nanakorn A, Sonkom P. Fal-
ciparum malaria semi-resistant to clindamycin. Br Med J
1975;2:12-4
18. Desjardins RE, Canfield CJ, Haynes JD, Chulay JD.
Quantitative assessment of antimalarial activity in vitro
by a semiautomated microdilution technique. Anti-
microb Agents Chemother 1979;16:710-8
19. Oduola AMJ, Weatherly NF, Bowdre JH, Desjardins
RE. Cloned strains of Plasmodium falciparum prepared
by single cell micromanipulation of infected erythrocytes
A ntiplasmodial Activity ofClindamycin
in vitro [abstract no. 58]. In: Program abstracts of the
joint meeting of the American Society of Tropical Medi-
cine and Hygiene and the American Society of Parasito-
logists in San Antonio, Texas 1983
20. Trager W, Jensen JB. Human malaria parasites in con-
tinuous culture. Science 1976;193:673-5
21. Pfaller MA, Krogstad OJ. Oxygen enhances the antimal-
arial activity of the imidazoles. Am J Trop Med Hyg
1983;32:660-5
22. Magerlein BJ, Kagan F. Lincomycin. VIII. 4'-aIkyl-l'-
demethyl-4'-depropylclindamycins, potent anti bacterial
and antimalarial agents. J Med Chern 1969;12:780-4
23. Brodasky TF, Lewis C, Eble TE. Bioautographic thin-
layer chromatographic analysis of antibiotics and their
metabolites in the whole animal: I. Clindamycin in the
rat. J Chromatogr 1976;123:33-44
24. Sherman IW. Biochemistry of Plasmodium (malarial para-
sites). Microbiol Rev 1979;43:453-95
25. Jensen JB. Concentration from continuous culture of ery-
throcytes infected with trophozoites and schizonts of
Plasmodium jalciparum. Am J Trop Med Hyg 1978;
27:1274-6
26. Aikawa M, Cochrane AH, Nussenzweig RS, Rabbage J.
Freeze-fracture study of malaria sporozoites: antibody-
induced changes of the pellicular membrane. J Protozool
1979;26:273-9
27. Fitch CD, Chevli R, Banyal HS, Phillips G, pfaller MA,
Krogstad OJ. Lysis of Plasmodium falciparum by ferri-
911
protoporphyrin IX and a chloroquine-ferriprotopor-
phyrin IX complex. Antimicrob Agents Chemother
1982;21:819-22
28. Colton T. Statistics in medicine. Boston: Little, Brown,
1974
29. Udeinya 11, Schmidt JA, Aikawa M, Miller LH, Green I.
Falciparum malaria-infected erythocytes specifically
bind to cultured human endothelial cells. Science 1981;
213:555-7
30. Pestka S. Insights into protein biosynthesis and ribosome
function through inhibitors. Prog Nucleic Acid Res Mol
Bioi 1976;17:217-45
31. Powers KG, Aikawa M, Nugent KM. Plasmodium knowl-
esi: morphology and course of infection in rhesus
Downloaded from http://jid.oxfordjournals.org/ at New York University on September 10, 2015
monkeys treated with clindamycin and its N-demethyl-
4'-pentyl analog. Exp ParasitoI1976;40:13-24
32. Leigh DA. Antibacterial activity and pharmacokinetics
of clindamycin. J Antimicrob Chemother 1981;7(Suppl
A):3-9
33. Miller LH, Glew RH, Wyler OJ, Howard WA, Collins WE,
Contacos PG, Neva FA. Evaluation of clindamycin in
combination with quinine against multidrug-resistant
strains of Plasmodium falciparum, Am J Trop Med Hyg
1974;23:565-9
34. Clyde OF, Gilman RH, McCarthy vc. Antimalarial ef-
fects of clindamycin in man. Am J Trop Med Hyg 1975;
24:369-70