Efficacy of Atovaquone/Proguanil for Malaria Prophylaxis in Children and Its Effect on the Immunogenicity of Live Oral Typhoid and Cholera Vaccines

Transcription

1 MAJOR ARTICLE Efficacy of Atovaquone/Proguanil for Malaria Prophylaxis in Children and Its Effect on the Immunogenicity of Live Oral Typhoid and Cholera Vaccines Jean-François Faucher, 1 Ronnie Binder, 1,2 Michel A. Missinou, 1,2 Pierre-Blaise Matsiegui, 1 Holger Gruss, 1 Rajko Neubauer, 1 Bertrand Lell, 1,2 John U. Que, 3 Gerri B. Miller, 4 and Peter G. Kremsner 1,2 1 Medical Research Unit, Albert Schweitzer Hospital, Lambaréné, Gabon; 2 Department of Parasitology, Institute for Tropical Medicine, University of Tübingen, Tübingen, Germany; 3 Berna Biotech, Berne, Switzerland; and 4 GlaxoSmithKline, Research Triangle Park, North Carolina A double-blind, placebo-controlled study was conducted to measure the impact of malaria prophylaxis with atovaquone/proguanil (A-P) on the immunogenicity of vaccines against typhoid fever and cholera, Salmonella serotype Typhi Ty21a and Vibrio cholerae CVD103-HgR, respectively. A total of 330 Gabonese schoolchildren were assigned to receive either A-P or placebo for 12 weeks. Vaccination occurred 3 weeks after the start of prophylaxis, and immunogenicity was assessed 4 weeks after vaccination. The protective efficacy of A-P against Plasmodium falciparum malaria was of 97% (95% confidence interval, 79% 100%). The 2 treatment groups did not differ significantly with regard to changes in antibody titers after vaccination ( P p.96 for anti S. Typhi IgG antibodies, P p.07 for anti S. Typhi IgA antibodies, and P p.64 for vibriocidal antibodies). The A-P combination was highly effective for malaria prophylaxis, without interfering with the in vivo immunogenicity of CVD103-HgR and Ty21a vaccines, and it could therefore be simultaneously administered with these vaccines. One-third of travelers who seek medical advice regarding prophylaxis are vaccinated for typhoid fever [1]. Live, oral, attenuated vaccines against typhoid fever and cholera (Salmonella serotype Typhi Ty21a and Vibrio cholerae CVD103-HgR, respectively) are licensed for human use and may be indicated for people living in or traveling to areas where malaria is endemic [2, 3]. Because some vaccine strain replication is needed to induce a protective immune response, administration of antibiotics may interfere with this process [4]. Me- Received 9 April 2002; accepted 25 June 2002; electronically published 25 October Financial support: GlaxoSmithKline. Reprints or correspondence: Dr. Peter G. Kremsner, Dept. of Parasitology, Institute for Tropical Medicine, University of Tübingen, Tübingen, Germany Clinical Infectious Diseases 2002; 35: by the Infectious Diseases Society of America. All rights reserved /2002/ $15.00 floquine and chloroquine exhibit in vitro antibacterial activity and are able to inhibit the Ty21a vaccine [5 8]. Chloroquine also has immunosuppressive properties and may interfere with the response to nonbacterial vaccines as well [9]. Proguanil, a dihydrofolate reductase inhibiting antimalarial drug, has some antibacterial activity. It can theoretically suppress multiplication of vaccine strains and can diminish the immunogenicity of the live, oral, attenuated vaccines. In fact, a previous study has shown that proguanil decreases the immunogenicity of the S. Typhi Ty21a vaccine but not the immunogenicity of the V. cholerae CVD103-HgR vaccine [8]. Several clinical studies have shown a high protection level against malaria with the atovaquone/proguanil hydrochloride (A-P) combination in semi-immune and nonimmune populations [10 14]. In many countries, A-P was initially approved for treatment of Plasmodium falciparum malaria [15] and is now approved for use as Atovaquone/Proguanil for Malaria CID 2002:35 (15 November) 1147

2 malaria prophylaxis. The impact of the A-P combination on plasma antibody responses to CVD103-HgR and Ty21a vaccines is unknown. Therefore, we conducted a double-blind, placebocontrolled study to evaluate the effectiveness of the A-P combination for the prevention of P. falciparum parasitemia and to assess its effect on the immune response elicited by CVD103- HgR and Ty21a vaccines in Gabonese children. METHODS The study took place at the Lalala Public school in Lambaréné, Gabon, from January through June Lambaréné is in a tropical rain forest area where P. falciparum malaria is hyperendemic [16]; the parasites in this region are resistant to chloroquine [17]. Healthy schoolchildren aged 4 16 years were enrolled in the trial. Exclusion criteria were as follows: pregnancy, lactation, unwillingness to avoid pregnancy during the trial, intake of antimalarial drugs 7 days before receipt of a radical cure regimen, evidence of AIDS or other immunodeficiency, and clinically significant abnormal hematological or clinical chemistry indexes. Written informed consent was obtained from the parents of the children. The study was approved by the ethics committee of the International Foundation of the Albert Schweitzer Hospital in Lambaréné and conducted in accordance with good clinical practice and all applicable regulations, including the Declaration of Helsinki (October 1996 version). The target sample size was based on a previous study performed at the same site [12]. We decided to enroll 110 subjects into each of 3 body-weight strata (group 1, kg; group 2, kg; group 3, kg), for a total of 330 subjects. In each group, all subjects who successfully completed the radical cure phase were randomly allocated to blinded placebo or A- P groups for chemosuppression at a 1:1 ratio. Because a high rate of parasitemia was expected in these children at baseline, all of the children received an open-label curative treatment course of artesunate (Plasmotrim Lactab; Mepha) once daily for 3 days; the daily dose of artesunate was determined by the children s weight: group 1, 50 mg; group 2, 100 mg; group 3, 150 mg. This short treatment regimen was chosen because artemisinin derivatives have a short half-life, are now marketed in Gabon, and achieve efficacy rates similar to the standard treatment duration (P.G.K., unpublished data). A thick blood smear was performed at the start and 7 days after the start of the initial curative treatment course. The smears were stained with Giemsa stain, and parasite counts were performed, as described elsewhere, with a detection limit corresponding to 5 parasites/ml [18]. During the chemosuppression phase, volunteers received a daily dose of placebo or A-P (Malarone; GlaxoSmithKline) according to their weight, as follows: group 1, 62.5 mg of atovaquone and 25 mg of proguanil hydrochloride; group 2, 125 mg of atovaquone and 50 mg of proguanil hydrochloride; group 3, mg of atovaquone and 75 mg of proguanil hydrochloride. The tablets were given each day under the supervision of field workers. Thick blood smears were performed weekly or in the event of chemoprophylaxis failure, and adverse events were assessed daily. A child was withdrawn from the study if he or she had not taken the study drug for 12 continuous days or if he or she had taken an additional drug with known antimalarial properties. The chemosuppression phase began 7 days after the start of the initial treatment and continued for 12 weeks. During 4 weeks of follow-up after the chemosuppressive phase, thick blood smears were performed at weekly visits, as was the reporting of adverse events. Blood samples were obtained to measure plasma drug concentrations of atovaquone, proguanil, and its active metabolite, cycloguanil. Bioanalysis was performed with use of plasma samples obtained at the trough level before administration of the study drug during weeks 3 and 5 of the chemoprophylaxis phase or at any time a subject developed parasitemia. Bioanalysis of plasma samples was performed as described elsewhere [19, 20]. CVD103-HgR and Ty21a vaccines were manufactured by the Swiss Serum and Vaccine Institute, as reported elsewhere [21]. The vaccines were presented in a doubled-chambered aluminium foil sachet. One chamber contained either cfu of CVD103-HgR or cfu of Ty21a. The other chamber contained a buffer. In vitro MICs of proguanil for the CVD103- HgR and Ty21a strains were determined by use of a standard tube dilution method. The MIC was defined as the lowest concentration of drug that resulted in no visible growth of the organism after incubation at 37 C for 24 h. All children remaining in the study at week 3 while taking the double-blind chemoprophylaxis were vaccinated with a single oral dose of combined CVD103-HgR and Ty21a. Additional doses of monovalent Ty21a were administered 3 and 5 days after the first dose. Venous blood samples were obtained during the screening and 28 days after immunization. Plasma was obtained, placed into labeled, dated tubes, and stored at 20 C. Vibriocidal antibody titers were determined by use of a microtiter plate assay, as described elsewhere, with the exception that the vaccine strain V. cholerae CVD 103-HgR rather than 569B (classical, Inaba) was used as the test strain [22]. A significant response was defined as a 4-fold increase in titer from the baseline level. Anti S. Typhi IgG and IgA lipopolysaccharide (LPS) antibodies were quantified by use of an ELISA described elsewhere [23], with some modification. In brief, the modifications included serially diluting the test sera, control sera, and reference sera. Anti-LPS antibody concentration (in endotoxin units per milliliter [EU/mL]) was calculated by extrapolation using the 4-parameter method, with the reference sera used as standards. A significant response was defined as a 1148 CID 2002:35 (15 November) Faucher et al.

3 Figure 1. 2-fold increase in titer from the baseline level, with a postvaccination antibody titer of 11 EU/mL. The parasitemia rate was calculated for P. falciparum parasitemia only; cases due to other Plasmodium species are discussed in the text. Efficacy data are reported here by perprotocol analysis. The protective efficacy was defined as 100 [1 (parasitemia in A-P recipients/parasitemia in placebo recipients)]. Groups were compared by means of the Mantel-Haenszel test (stratified by weight). The incidence density was calculated for adverse events and P. falciparum parasitemia as the number of cases divided by the total personyears (exposure time) for each group in the appropriate population. Serum antibody titers were log-transformed and then summarized and compared by use of Student s t test for the difference before and after vaccination. All subjects who had received 1 dose of prophylactic drug and for whom preand postvaccination serum samples were available were included in the analysis. Trial profile of schoolchildren who received atovaquone/proguanil for malaria prophylaxis RESULTS Of the 1040 schoolchildren attending Lalala school, 376 volunteered actively, and 355 of these 376 were screened (figure 1). We obtained venous blood samples from 355 children. Twenty-five children were not eligible: 3 had elevated alanine aminotransferase activity, 1 had anemia, 1 withdrew consent, 3 were lost to follow-up, and 17 withdrew for clinical reasons. Thus, 330 children were enrolled and randomly assigned to a treatment group. There was no significant difference between the groups with regard to demographic characteristics, baseline hematological findings, clinical chemistry, vital signs, or parasitemia indexes (table 1). All enrolled children received curative treatment; 112 (34%) of 330 had a positive blood smear result at the start of the initial cure regimen: 106 tested positive for P. falciparum, 4 tested positive for Plasmodium malariae, and 2 tested positive for Plasmodium ovale. All 330 completed the curative treatment Atovaquone/Proguanil for Malaria CID 2002:35 (15 November) 1149

4 Table 1. Clinical and laboratory data at baseline screening for the safety population of 330 children who received either atovaquone/proguanil (A-P) or placebo for malaria prophylaxis. Characteristic (n p 165) (n p 165) Sex, no. male/no. female 76/89 76/89 Age, years Weight, kg Parasitemia, a no. (%) 54 (32.7) 58 (35.1) Infection with Plasmodium falciparum only, no. Vital signs Systolic blood pressure, mm Hg Diastolic blood pressure, mm Hg Heart rate, beats/min Temperature, C Laboratory data Hemoglobin level, g/dl Leukocyte count, 10 9 cells/l Platelet count, 10 6 platelets/l Creatinine level, mm Total bilirubin level, mm Alanine aminotransferase level, IU/L NOTE. Data are mean SD, unless otherwise indicated. Safety population includes all subjects who took 1 dose of study drug (including radical cure treatment). a All causes. regimen. On day 0 of chemoprophylaxis, all schoolchildren had cleared their parasitemia except for 1 child, who had P. falciparum parasitemia when the curative treatment regimen began and was subsequently found to be positive for P. malariae on day 0. From day 0 to day 21, 6 schoolchildren (5 of whom had a positive blood smear at the start of the initial radical cure regimen) who were allocated to receive placebo developed parasitemia (P. falciparum). Thus, a 3-day treatment course with artesunate led to a parasitological cure until day 22 in 53 (91%) of 58 parasitemic schoolchildren receiving placebo. There were no parasitological failures in schoolchildren allocated to receive A-P. A total of 36 (11%) of 330 subjects reported having 1 adverse event during the radical cure treatment phase; none were serious or severe in intensity. Throughout the chemosuppressive phase, treatment-emergent adverse events (regardless of relationship to study drug) were experienced by 42% of the placebo recipients and 44% of the A-P recipients. Most were mild or moderate in intensity; the most common adverse events are shown in table 2. There was no chemoprophylaxis interruption related to an adverse event. Thus, A-P chemoprophylaxis was well tolerated. Of the 330 schoolchildren allocated to receive the chemosuppressive therapy, 165 were assigned to receive A-P and 165 were assigned to receive placebo. Eighteen children assigned to the placebo group and 19 children assigned to the discontinued treatment during the course of the study (prophylaxis phase and follow-up phase) for reasons other than prophylaxis failure. For the per-protocol population, during the total observation time of 30.7 person-years (prophylaxis phase) in the placebo group, P. falciparum positive blood smears were noted for 31 children, which is an incidence density of 1.01 cases per person per year. Children in the were observed for 36.9 person-years. One child in this group had a positive blood smear result during the chemosuppression phase, which is an incidence density of 0.03 cases per person per year; the protective efficacy was 97% (95% CI, 79% 100%; table 3). One subject in the had a blood smear that was positive for malaria during the chemoprophylaxis phase (week 9) of the study. This subject was in the lower end of weight group 3 (33 kg) and received 3 tablets of the A-P combination daily. At the time of the positive blood smear, the child had concentrations of atovaquone, proguanil, and cycloguanil of 3.8 mg/ml, 22.3 ng/ml, and less than the quantifiable limit, respectively. These values were similar to those observed at weeks 3 and 5 and were within the range of the values observed in the other subjects in weight group 3 (at least one-third of 1150 CID 2002:35 (15 November) Faucher et al.

5 Table 2. Adverse events that occurred during the initial curative treatment and chemosuppression for the safety population of 330 children who received either atovaquone/proguanil (A-P) or placebo for malaria prophylaxis. Chemosuppression phase Adverse event Initial cure phase, no. (%) (n p 330) No. (%) (n p 165) Incidence density (n p 165) No. (%) Incidence density Any 36 (11) 73 (44) (42) 2.13 Headache 12 (4) 22 (13) (13) 0.67 Abdominal pain 11 (3) 22 (13) (8) 0.40 Cough 4 (1) 16 (10) (10) 0.52 Fever 5 (2) 9 (5) (12) 0.58 Vomiting 5 (2) 9 (5) (3) 0.15 Diarrhea 1 (!1) 2 (1) (5) 0.24 NOTE. Only treatment-emergent adverse events that occurred in 5% of the subjects who received A-P or placebo during the chemoprophylaxis phase are presented. The safety population includes all subjects who took 1 dose of the study drug (including radical cure treatment). the children in this weight group had low or undetectable plasma cycloguanil levels). During the 4 weeks of follow-up, during which the children did not receive study medication, 4 P. falciparum positive blood smears were recorded, 3 of which were in the and 1 of which was in the placebo group (table 3). One of the children whose malaria failed to respond to A-P prophylaxis was coinfected with P. malariae. The illnesses of 3 patients in the that failed to respond to therapy were found during the last 2 weeks of follow-up, when the drug concentrations may no longer have been chemosuppressive, and may have been attributable to infections acquired after the end of prophylaxis. Of the 41 total positive blood smears recorded throughout both the chemosuppressive and the follow-up phases, most were due to P. falciparum (36 were due to P. falciparum, 4 were due to P. malariae, 1 was due to P. ovale). The MICs of proguanil were 150 mg/ml against CVD103- HgR and 1200 mg/ml against Ty21a. The effect of concomitant administration of A-P or placebo on the vibriocidal and Anti S. Typhi IgG and IgA LPS antibody responses engendered by CVD103-HgR in combination with Ty21a was determined. Preand postvaccination plasma samples were available for 152 schoolchildren allocated to receive A-P and 123 schoolchildren allocated to receive placebo. One child did not complete vaccination because of vomiting that occurred soon after vaccine intake on vaccination day 1. The results are listed in tables 4 and 5. The geometric mean baseline vibriocidal antibody titers were high for both groups (144 and 139 in the placebo and A- P groups, respectively) and did not differ significantly. The vibriocidal antibody response rates were low (22% and 27% in the placebo and s, respectively), and there was no between-groups statistical difference in the change in vibriocidal antibody titer after vaccination ( P p.64). The relationship between baseline vibriocidal antibody titer and the propensity to seroconvert was analyzed (table 6). The rate of seroconversion decreased with increasing baseline titers (P!.01; determined by the runs test for trend, post hoc analysis). Immunization with the bivalent oral Ty21a and CVD103- Table 3. Percentage protective efficacy of malaria treatment (per-protocol population) in children who received prophylaxis with atovaquone/proguanil (A-P) or placebo. Value (n p 150) Chemosuppression phase (n p 144) P (n p 143) Follow-up (n p 112) No. (%) 1 (!1) 31 (22)! (2) 1 (!1) Incidence density Efficacy (95% CI) 96.9% (78.9% 100%) NOTE. Efficacy was calculated from the following formula: 100 [1 (parasitemia proportion in A-P recipients/ parasitemia proportion in placebo recipients)]. Only Plasmodium falciparum parasitemia was assessed. Groups were compared by use of the exact Mantel-Haenszel test (stratified by weight). The per-protocol population excluded those subjects who withdrew during the prophylaxis phase for reasons other than P. falciparum parasitemia or a treatmentrelated adverse event. Atovaquone/Proguanil for Malaria CID 2002:35 (15 November) 1151

6 Table 4. Mean serum antibody titers in the intent-to-treat population for Salmonella serotype Typhi and Vibrio cholerae in children who received atovaquone/proguanil (A-P) or placebo as malaria prophylaxis who were then assessed for the immunogenicity of oral attenuated vaccines against typhoid fever and cholera. Antibody No. (%) GM of log titers (mean SD) No. (%) GM of log titers (mean SD) S. Typhi IgG.958 Before ( ) ( ) After ( ) ( ) Difference ( ) ( ) S. Typhi IgA.072 Before ( ) ( ) After ( ) ( ) Difference ( ) ( ) Vibriocidal.637 Before ( ) ( ) After ( ) ( ) Difference ( ) ( ) NOTE. The intent-to-treat population included all subjects randomized to a treatment group who took 1 dose of the assigned chemoprophylactic drug. GM, geometric mean. a Comparison between treatment groups (atovaquone-proguanil vs. placebo) of the within-group change in antibody titers (i.e., pre- to postvaccination antibody titer change of the treatment groups) was compared by Student s t test via log-transformed data. HgR increased the geometric mean anti S. Typhi LPS IgG and IgA antibody titers in both groups. However, the betweengroups difference in mean antibody increase for both IgG and IgA was not statistically significant (table 4). In the placebo and s, the anti S. Typhi IgG LPS antibody response rates were 67% and 74%, whereas the anti S. Typhi IgA LPS antibody response rates were 33% and 45%, respectively. There was no between-groups statistical difference for the change in anti S. Typhi IgG and IgA LPS antibody titers after vaccination ( P p.96 and P p.07, respectively). DISCUSSION The high effectiveness of the A-P combination for malaria prophylaxis was already established by 3 placebo-controlled trials in areas where chloroquine-resistant malaria was endemic; 677 patients (479 adults and 265 children) were enrolled in these trials [15]. The primary end point of our clinical trial was to confirm that pediatric tablets formulated with proguanil hydrochloride from a different manufacturer would result in prophylaxis efficacy and safety comparable to the efficacy and safety noted in the first pediatric trial [12] and in studies of adults [10, 11]. The present study demonstrates that the A-P combination was well tolerated and that the level of protection was 97%, which is comparable to that observed in previous studies in which the protective efficacy range was 95% 100% [15]. Many of the positive smear results observed within 28 days after the initiation of the clearance treatment course may be radical cure treatment failures. The level of protection is similar when excluding all schoolchildren (all allocated to receive placebo), with positive smears observed within 28 days after the initiation of clearance treatment course (data not shown). Table 5. Antibody response in the intent-to-treat population for Salmonella serotype Typhi and Vibrio cholerae in children who received atovaquone/proguanil (A-P) or placebo as malaria prophylaxis and who were then assessed for the immunogenicity of oral attenuated vaccines against typhoid fever and cholera. Antibody (n p 152) P a No. (%) (n p 123) Positive a Negative Positive a Negative S. Typhi IgG 112 (74) 40 (26) 83 (67) 40 (33) IgA 69 (45) 83 (55) 40 (33) 83 (67) V. cholerae 41 (27) 111 (73) 27 (22) 96 (78) NOTE. The intent-to-treat population included all subjects randomized to a treatment group and who took 1 dose of the assigned chemoprophylactic drug. a Positive response was considered a 2-fold increase in S. Typhi antibody titer and a 4-fold increase in V. cholerae antibody titer CID 2002:35 (15 November) Faucher et al.

7 Table 6. Relationship between baseline reciprocal vibriocidal titer and rate of vibriocidal seroconversion in children who received atovaquone/proguanil prophylaxis for malaria and who were then studied for immunogenicity of oral attenuated vaccines against typhoid fever and cholera. Baseline vibriocidal titer No. who had seroconversion/ no. vaccinated (%)!20 6/6 (100) /1 (100) /21 (62) /118 (24) /129 (16) NOTE. The relationship was significant for trend ( P!.01; determined by means of the runs test for trend, post hoc analysis). To our knowledge, only 1 other study has been published on interactions between antimalarials and oral attenuated vaccines for typhoid fever and cholera. This open trial showed that coadministered chloroquine and proguanil (at 200 mg daily) have a negative impact on the immunogenicity of the V. cholerae CVD103-HgR and S. Typhi Ty21a vaccines and concluded that chloroquine and proguanil should not be simultaneously administered with the V. cholerae CVD103-HgR and S. Typhi Ty21a strains [8]. It is currently advised to delay proguanil treatment for 3 days after receipt of the last dose of Ty21a [24]. The present double-blind study was also designed to address the effect of the A-P combination on the immune response to live attenuated vaccines. Response rates to the Ty21a vaccine were similar in both groups, and the between-groups difference was not statistically significant. Thus, A-P chemoprophylaxis did not appear to inhibit the humoral immune response to concurrently administered typhoid and cholera vaccinations. Although Ty21a vaccine should be administered before malaria chemoprophylaxis in most cases, extreme situations can lead travelers or people in areas of endemicity to receive concomitant chemoprophylaxis and vaccination against typhoid fever. In such cases, when use of the A-P combination is advised for malaria chemoprophylaxis, the data presented here suggest that the Ty21a vaccine is suitable for vaccination against typhoid fever. It is not clear why our study did not show that the A-P combination had any impact on Ty21a vaccine immunogenicity, even though use of proguanil alone may reduce Ty21a immunogenicity [8]. It is possible that the combined effect of atovaquone and proguanil reduces the antibacterial activity of proguanil on the Ty21a vaccine or that the low dose of proguanil used in this study did not have an effect. In addition, antibiotics may well be modulating the vaccine-induced humoral immune response in a way independent of their antibacterial properties [25], because it has been shown that the in vitro activity of various antimalarial drugs is not predictive of their in vivo impact on vaccine immunogenicity [8]. This raises questions about the in vivo interference of antimalarials with the immune response to live attenuated bacterial vaccines through their antibacterial properties. Although response rates to CVD103-HgR vaccine were not affected by A-P prophylaxis, they are surprisingly low, given the high rates of seroconversion that have been achieved in previous studies performed in less developed countries that used the same vaccine dose as the present study [26 29]. Because the rate of seroconversion decreased with increasing baseline titers, the low vibriocidal antibody response rate to CVD103-HgR vaccine could be due to the high mean baseline vibriocidal antibody titer in both groups. Previous studies of the CVD103-HgR vaccine from industrialized countries as well as less developed countries have found significantly lower baseline vibriocidal antibody titers among participants who responded to the vaccine [26 30]. The mean baseline vibriocidal antibody titer found in our study is higher than that reported in previous studies of CVD103-HgR vaccine immunogenicity from areas of endemicity [26 29]. Cholera is not highly endemic in Gabon, and no cholera epidemic has been reported in the study area within the past decade. A study on the etiology of acute infantile gastroenteritis in Gabon found no reports of cholera cases [31]. The meaning of the high mean baseline vibriocidal antibody titer found in this area is therefore unclear. However, elevated vibriocidal antibody titers may in part be caused by other species, because vibriocidal antibodies may be induced by bacteria other than V. cholerae [32]. In conclusion, the A-P combination confers a high level of protection against malaria in this region, which harbors chloroquine-resistant strains of P. falciparum, and the A-P combination does not appear to decrease the immunogenicity of CVD103-HgR and Ty21a vaccines. It appears that the A-P combination could therefore be simultaneously administered with the oral attenuated vaccines for typhoid fever and cholera. Low response rates to CVD103-HgR are related to high baseline vibriocidal antibody titers. Acknowledgments We thank Dr. Jeff Chulay, for assistance with study design, and Nadine Jacquemet, Catherine Granier, and Neil Roskell, for statistical support. We thank the director and staff of the Public School of Lalala (Lambaréné, Gabon) for support. Marcel Nkeyi, Anselme Nzengué, and Jean Mbembé provided excellent technical support, and 18 field assistants helped with Atovaquone/Proguanil for Malaria CID 2002:35 (15 November) 1153

8 the field work. We especially thank the study children and their parents. References 1. Behrens RH, Roberts JA. Is travel prophylaxis worthwhile? Economic appraisal of prophylactic measures against malaria, hepatitis A and typhoid in travellers. BMJ 1994; 309: Levine MM, Ferreccio E, Abrego P, San Martin O, Ortiz E, Cryz S. Duration of efficacy of Ty21a, attenuated Salmonella typhi live oral vaccine. Vaccine 1999; 17:S Ryan ET, Calderwood SB. Cholera vaccines. Clin Infect Dis 2000; 31: Wolfe MS. Precautions with oral live typhoid (Ty21A) vaccine. Lancet 1990; 336: Horowitz H, Carbonaro CA. Inhibition of the Salmonella typhi oral vaccine strain, Ty21a, by mefloquine and chloroquine. J Infect Dis 1992; 166: Ambrosch F, Hirschl A, Kremsner P, et al. Orale Typhus-Lebendimpfung. Munch Med Wochenschr 1985; 127: Ambrosch F, Hirschl A, Kollaritsch H, Kremsner P, Rappold E, Wiedermann G. Untersuchungen zur Kompatibilität des neuen Malariamittels Mefloquin mit der oralen Typhus-Lebendimpfung. Mitt Osterr Ges Tropenmed Parasitol 1987; 9: Kollaritsch H, Que JU, Kunz C, Wiedermann G, Herzog C, Cryz SJ. Safety and immunogenicity of live oral cholera and typhoid vaccines administered alone or in combination with antimalarial drugs, oral polio vaccine, or yellow fever vaccine. J Infect Dis 1997; 175: Pappaiouanou M, Fishbein DB, Dreesen DW, et al. Antibody response to preexposure human diploid-cell rabies vaccine given concurrently with chloroquine. N Engl J Med 1986; 314: Shanks GD, Gordon DM, Klotz FW, et al. Efficacy and safety of atovaquone/proguanil for suppressive prophylaxis against Plasmodium falciparum malaria. Clin Infect Dis 1998; 27: Sukwa TY, Mulenga M, Chisaka N, Roskell NS, Scott TR. A randomized, double-blind, placebo-controlled field trial to determine the efficacy and safety of Malarone (atovaquone/proguanil) for the prophylaxis of malaria in Zambia. Am J Trop Med Hyg 1999; 60: Lell B, Luckner D, Ndjave M, Scott T, Kremsner PG. Randomized placebo-controlled study of atovaquone plus proguanil for malaria prophylaxis in children. Lancet 1998; 351: Høgh B, Clarke PD, Camus D, et al., and the Malarone International Study Team. Atovaquone-proguanil versus chloroquine-proguanil for malaria prophylaxis in non-immune travellers: a randomised, doubleblind study. Lancet 2000; 356: Overbosch D, Schilthuis H, Bienzle U, et al. Atovaquone-proguanil versus mefloquine for malaria prophylaxis in nonimmune travelers: results from a randomized, double-blind study. Clin Infect Dis 2001; 33: Shanks GD, Kremsner PG, Sukwa TY, et al., for the Malarone Clinical Trials Study Group. Atovaquone and proguanil hydrochloride for prophylaxis of malaria. J Travel Med 1999; 6(Suppl 1):S Wildling E, Winkler S, Kremsner PG, et al. Malaria epidemiology in the Province of Moyen Ogooué, Gabon. Trop Med Parasitol 1995; 46: Kremsner PG, Winkler S, Brandts C, Neifer S, Bienzle U, Graninger W. Clindamycin in combination with chloroquine or quinine is an effective therapy for uncomplicated Plasmodium falciparum malaria in children from Gabon. J Infect Dis 1994; 169: Kremsner PG, Zotter GM, Feldmeier H, Graninger W, Rocha RM. A comparative trial of three regimens for treating uncomplicated falciparum malaria in Acre, Brazil. J Infect Dis 1988; 158: Hussein Z, Eaves J, Hutchinson DB, Canfield CJ. Population pharmacokinetics of atovaquone in patients with acute malaria caused by Plasmodium falciparum. Clin Pharmacol Ther 1997; 61: Hussein Z, Eaves J, Hutchinson DB, Canfield CJ. Population pharmacokinetics of proguanil in patients with acute P. falciparum malaria after combined therapy with atovaquone. Br J Clin Pharmacol 1996; 42: Cryz SJ Jr, Que JU, Levine MM, Wiedermann G, Kollaritsch H. Safety and immunogenicity of a live oral bivalent typhoid fever (Salmonella typhi Ty21a) cholera (Vibrio cholerae CVD103-HgR) vaccine in healthy adults. Infect Immun 1995; 63: Clements ML, Levine MM, Young CR, Black RE, Lim YL, Robins- Browne RM. Magnitude, kinetics and duration of vibriocidal antibody response in North Americans after ingestion of Vibrio cholerae. J Infect Dis 1982; 145: Levine MM, Herrington D, Murphy JR, et al. Safety, infectivity, immunogenicity and in vivo stability of two attenuated auxotrophic mutant strains of Salmonella typhi 541 Ty and 543 Ty, as live oral vaccines in man. J Clin Invest 1987; 79: Foster RH, Noble S. Bivalent cholera and typhoid vaccine. Drugs 1999; 58: Woo PCY, Tsoy HW, Wong LP, Leung HCH, Yuen KY. Antibiotics modulate vaccine-induced humoral immune response. Clin Diagn Lab Immunol 1999; 6: Su-Arehawatana P, Singaraj P, Taylor D, et al. Safety and immunogenicity of different immunization regimens of CVD103-HgR live oral cholera vaccine in soldiers and civilians in Thailand. J Infect Dis 1992; 165: Suharyono, Simanjuntak C, Witham N, et al. Safety and immunogenicity of single-dose live oral cholera vaccine CVD103-HgR in 5 9-yearold Indonesian children. Lancet 1992; 340: Gotuzzo E, Butron B, Seas C, et al. Safety, immunogenicity and excretion pattern of single dose live oral cholera vaccine CVD103-HgR in Peruvian adults of high and low socioeconomic levels. Infect Immun 1993; 61: Perry RT, Plowe CV, Koumaré B, et al. A single dose of live oral cholera vaccine CVD 103-HgR is safe and immunogenic in HIV-infected and HIV-noninfected adults in Mali. Bull World Health Organ 1998; 76: Cryz S, Levine MM, Losonsky J, Kaper JB, Althaus B. Safety and immunogenicity of a booster dose of Vibrio cholerae CVD103-HgR live oral cholera vaccine in Swiss adults. Infect Immun 1992; 60: Gendrel D, Sitbon M, Richard-Lenoble D, et al. Etiology of acute infantile gastroenteritis in Gabon. Bull Soc Pathol Exot Filiales 1985; 78: Agarwal SC, Ganguly NK. Vibriocidal antibodies induced by Yersinia enterocolitica serotype IX. J Hyg 1972; 70: CID 2002:35 (15 November) Faucher et al.