| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
L.5 However, there has been concern that the widespread use of cotrimoxazole for prophylaxis might accelerate the spread of resistance to sulfadoxine-pyrimethamine (SP), another antifolate combination that is one of the most widely available and affordable antimalarials in Africa.6,7 Sulfadoxine-pyrimethamine used alone or in combination is the first-line treatment of uncomplicated malaria in several African countries and is currently the only recommended drug for the prevention of malaria in pregnancy.8 Resistance to SP among Plasmodium falciparum is mediated by point mutations that accumulate at several sites in the dihydrofolate reductase (dhfr) and dihydropteroate synthase (dhps) genes, resulting in increasing levels of drug resistance in vitro.9 In Africa, clinical failure after treatment with SP either alone or in combination has been associated with the dhfr triple mutant (Asn-108 + Ile-51 + Arg-59) and the dhps double mutant (Gly-437 + Glu-540).10 Laboratory studies have shown that P. falciparum isolates with the dhfr triple mutant have high-level resistance to both trimethoprim and pyrimethamine.11 However, there are few clinical studies evaluating whether cotrimoxazole leads to the selection of parasites resistant to SP. It has been argued that the clinical benefits of cotrimoxazole prophylaxis among HIV-infected patients outweigh potential risks related to the selection of SP resistant parasites.12 However, concerns that cotrimoxazole use by HIV-infected patients might increase the spread of SP-resistant parasites in the community have delayed implementation of prophylaxis guidelines.
To evaluate the potential association between taking cotrimoxazole prophylaxis and the spread of SP-resistant malaria, we analyzed data collected during a study of daily cotrimoxazole prophylaxis among a cohort of HIV-infected persons living in rural Uganda.3 During the study, all HIV-infected and HIV-uninfected household members were followed-up with weekly home visits to detect clinical malaria, and blood specimens were collected on filter paper from each episode of possible malaria for resistance testing. We evaluated the hypothesis that HIV-uninfected persons residing in households where HIV-infected household members were taking cotrimoxazole prophylaxis would have a higher risk of developing malaria caused by parasites containing molecular markers of SP resistance than those in households where HIV-infected household members were not taking cotrimoxazole.
|
METHODS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The cohort and study methods have been described previously, and are abbreviated below.3 During two time periods, between April and May 2001 and between January and April 2002, we enrolled HIV-infected clients of The AIDS Support Organization (TASO) and their household members in Tororo District, eastern Uganda, in a prospective cohort study. After written informed consent was provided, study staff visited clients’ homes to conduct a census, obtain informed consent from household members, and administer an individual, standardized questionnaire. A household was defined as persons who shared their meals and slept in the same house or cluster of houses for at least five days of the week for 3 months before the baseline survey. Households enrolled in 2001 were followed-up for a period of approximately five months before HIV-infected participants were given treatment with daily cotrimoxazole (adults, 160 mg of trimetho-prim/800 mg of sulfamethoxazole; children between 5–12 years of age, 80 mg of trimethoprim/400 mg of sulfamethoxazole; and children less than five years of age, syrup adjusted to weight). To increase sample size, a second group of households were enrolled in 2002 using identical inclusion criteria and consent procedures. Neither consent for HIV testing nor receiving test results were requirements for enrollment in the study, but only persons for whom HIV test results were known were included in analyses. In this study, we used the time period when HIV-infected participants from the first enrollment period were taking cotrimoxazole and HIV-infected participants from the second enrollment period were not yet taking cotrimoxazole (Figure 1
). This overlap period allowed us to compare two parallel populations controlling for any temporal trends.
|
Given that 113 malaria cases in the exposed group and 163 malaria cases in the unexposed group were available for resistance testing, we estimated that with 80% power, an alpha error of 0.05 and the prevalence of either dhps double mutant (Gly-437 + Glu-540) or dhfr triple mutant (Asn-108 + Ile-51 + Arg-59) of 0.4, we would be able to detect a difference in prevalence 8.3% for the dhps double mutant and for the dhfr triple mutant outcomes.
Laboratory methods. Films for malaria parasites were treated with Leishman’s stain and parasite counts were measured per 200 white blood cells. Thin smears were used to identify Plasmodium species. We randomly selected 50% of filter paper specimens from all malaria episodes diagnosed to test for the presence molecular markers of SP resistance. We assessed for the presence of three mutations in the dhfr gene (dhfr Asn-108, Ile-51, and Arg-59) and two mutations in the dhps gene (dhps Gly-437 and Glu-540) commonly found in east Africa and one mutation in the dhfr gene (Leu-164) rarely found in Africa, but associated with high-level SP resistance.13 Parasite DNA was isolated from filter paper using the Chelex extraction method,14 and genotypes were determined using nested polymerase chain reaction amplification followed by restriction enzyme digestion and visualization using gel electrophoresis as previously described.15,16 Specimens were classified as wild-type, pure mutant, or mixed (both mutant and wild-type alleles detected in the same specimen).
Statistical methods. Data were entered using EpiInfo (Centers for Disease Control and Prevention [CDC], Atlanta, GA)17 and analyzed using STATA (Stata Corporation, College Station, TX). We compared the prevalence of mutations in those exposed and unexposed to cotrimoxazole using the chi-square test. A Poisson distribution using a log-link function was used to compare the incidence of malaria episodes controlling for age, sex, and potential temporal trends associated with calendar period. We used rainfall data collected in the geographic area during 2001 and 2002 to derive calendar periods that might have higher mosquito populations. Generalized estimating equation methods with an exchangeable correlation structure model were used for analysis of repeated outcome measures among the same individuals. To compare the incidence of malaria due to parasites containing the dhfr/ dhps quintuple mutant, we first applied the proportions derived from the specimens tested for mutation to the whole study population. We then used the hotdeck imputation method18 to correct for missing mutation results. Variables with a significance level of 0.05 were included in the final model.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The HIV-uninfected household members exposed and unexposed to household cotrimoxazole use were similar in terms of median age (11 versus 10 years) and the proportion that were female (53% versus 49%). In both groups, the median number of HIV-infected and uninfected persons in a household was one and three, respectively.
Prevalence of molecular markers of SP resistance.
Of the 224 episodes of malaria diagnosed among HIV-uninfected patients exposed to household cotrimoxazole use, 113 (50%) filter paper specimens were randomly selected for molecular analysis and 105 (93%) were successfully assayed for all mutations tested. Of the 323 episodes of malaria diagnosed among HIV-uninfected persons not exposed to household cotrimoxazole use, 163 (50%) filter paper specimens were randomly selected for molecular analysis and 146 (90%) were successfully assayed for all mutations tested. None of the specimens contained the dhfr Leu-164 mutation. The dhfr Asn-108 and Ile-51 mutations were virtually ubiquitous in both groups, with 100% and 94% of samples containing pure mutants, respectively. The prevalence of the dhfr Arg-59 pure mutant was less common (41% overall), consistent with the stepwise progression of dhfr mutations, with Arg-59 occurring after Asn-108 and Ile-51. Considering only samples with pure mutants, the prevalence of the dhfr triple mutant (Asn-108 + Ile-51 + Arg-59) was similar in samples from HIV-uninfected patients with malaria living in households either unexposed or exposed to cotrimoxazole prophylaxis (45% versus 37%, P = 0.18) (Figure 2). The dhps Gly-437 and Glu-540 mutations were highly concordant, with 99% of samples having either both mutations or neither. The prevalence of the dhps double mutant (Gly-437 + Glu-540) was similar in samples from HIV-uninfected patients with malaria living in households either unexposed or exposed to cotrimoxazole prophylaxis (62% versus 64%, P = 0.81) (Figure 2). Considering all five mutations, the prevalence of the quintuple mutant was similar in samples from HIV-uninfected patients with malaria living in households unexposed and exposed to cotrimoxazole prophylaxis (32% versus 30%, P = 0.74) (Figure 2). Similar results were obtained when samples that were either mixed or pure mutants, rather than only pure mutants, were considered.
|
5 years of age (IRR =11.8, 95% CI = 9.3–15.0). However, the decreased incidence of malaria among HIV-uninfected household members exposed to cotrimoxazole was consistent across both age groups (Table 1).
|
). To account for missing data on molecular markers of SP resistance, we repeated the analysis using imputation to reconstruct a complete dataset based on a covariate structure that considered gender, age and calendar period. The HIV-uninfected household members in a household using cotrimoxazole prophylaxis had a decreased incidence of malaria due to parasites containing the dhps double mutant (IRR = 0.67, 95% CI = 0.49–0.92, P = 0.014), the dhf triple mutant (IRR = 0.49, 95% CI = 0.34–0.69, P < 0.001), and the dhfr/dhps quintuple mutant (IRR = 0.61, 95% CI = 0.41–0.91, P = 0.014). Age and season were independent predictors for clinical malaria. |
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The efficacy of cotrimoxazole prophylaxis in preventing morbidity and mortality among Africans with HIV infection has been well documented.1,2,19 There is concern, however, that malaria parasites will develop cross-resistance between cotrimoxazole and SP due to the similarities in the modes of action of these two drugs. Cotrimoxazole and SP are both antifolates, with components that inhibit dhfr (trimethoprim and pyrimethamine, respectively) and dhps (sulfamethoxazole and sulfadoxine, respectively). An in vitro study demonstrated significant cross-resistance between trimethoprim and pyrimethamine in laboratory isolates containing the dhfr triple mutation commonly found in Africa.11 However, it is not clear whether significant cross resistance between cotrimoxazole and SP occurs in vivo. In a clinical study from Uganda, children with uncomplicated malaria who failed treatment with cotrimoxazole paradoxically selected for the dhfr 108 wild-type allele.20 In another study of children with malaria, the presence of SP-resistance conferring mutations was not linked to cotrimoxazole resistance in Escherichia coli stool isolates, although measures of resistance were high for both drugs.21
In our study, there was no evidence that cotrimoxazole prophylaxis lead to the spread of SP-resistant malaria parasites among household members not taking the drug. In the context of documented in vitro cross-resistance between cotrimoxazole and SP, why is there no in vivo selection of cross-resistant parasites by cotrimoxazole? It may be that cotrimoxazole does select for SP-resistant parasites, but that our period of observation was not long enough to capture this effect on the surrounding community. Alternatively, the spread of resistant parasites through the community may have occurred so rapidly that we were unable to detect a difference between households exposed and unexposed to cotrimoxazole. However, this explanation seems unlikely given the wide geographic dispersion of households in our rural setting. In addition, the possibility that SP-resistant malaria parasites arose in both the exposed and the unexposed groups due to contamination by mosquitoes crossing to unexposed households is very unlikely given that malaria incidence in general was less in the exposed group.
Despite the lack of evidence that cotrimoxazole prophylaxis contributes to the spread of malaria parasites resistant to SP, the prevalence of molecular markers of SP resistance in our study population was very high. A recent study reported that SP-resistant parasites have spread rapidly across southern and eastern Africa from a few common ancestral clones.22 Our study was conducted during a period when Uganda switched from chloroquine to a combination of chloroquine plus SP as the recommended first-line treatment of uncomplicated malaria. It is unknown whether increasing use of SP in our community contributed to the high prevalence of SP-resistant parasites. From our results, one may infer that in a region with existing high-level resistance, such as Uganda, factors other than cotrimoxazole use in the community are primarily responsible for the selection and spread of SP-resistant parasites.
In addition to our observation that cotrimoxazole did not lead to resistance selection, we found that it offered the benefit of a reduced incidence of malaria among HIV-uninfected persons living in households exposed to cotrimoxazole. Considering high levels of SP resistance in this population, why might prophylaxis effectively prevent malaria? This effect may be seen because there is limited clinically relevant cross-resistance between the two medications, or because the antimicrobial effect needed to prevent an infection is less than that required to treat a clinical episode of malaria.23
Cotrimoxazole prophylaxis by HIV-infected individuals was not associated with an increased prevalence of SP-resistant parasites among HIV-uninfected household members. Since cotrimoxazole prophylaxis reduces malaria and improves survival for HIV-infected persons in sub-Saharan Africa, the implementation of prophylaxis in this population should not be delayed because of theoretical concerns regarding the selection of SP-resistant malaria parasites.
Received February 13, 2006. Accepted for publication March 28, 2006.
Acknowledgement: We thank the staff and clients of TASO, the Uganda Virus Research Institute-CDC laboratory, and the staff of CDC-Uganda for participating in the study.
Financial support: This study was supported by the Fogarty AIDS International Training and Research Program/University of California, Berkeley (1-D43-TW00003), the Fogarty International Center/ National Institutes of Health (TW00007), and the Centers for Disease Control and Prevention.
* Address correspondence to Samuel S. Malamba, Division of Epidemiology, School of Public Health, University of California, 1918 University Avenue, Fourth Floor, Berkeley, CA 94720 and Centers for Disease Control and Prevention–Uganda, Global AIDS Program, National Center for HIV, STD and TB Prevention, c/o Uganda Virus Research Institute, PO Box 49, Entebbe, Uganda. E-mails: malambas{at}berkeley.edu and zcq2{at}ug.cdc.gov 
Authors’ addresses: Samuel S. Malamba, Division of Epidemiology, School of Public Health, University of California, 1918 University Avenue, Fourth Floor, Berkeley, CA 94720 and CDC-Uganda, Global AIDS Program, National Center for HIV, STD and TB Prevention, c/o Uganda Virus Research Institute, PO Box 49, Entebbe, Uganda, Telephone: 256-41-320776, 256-752-790145, or 510-643-4922, Fax: 256-41-321457 or 510-643-4927, E-mails: zcq2{at}ug.cdc.gov and malambas{at}berkeley.edu. Jonathan Mermin, John R. Lule, Robert Downing, Ray Ransom, and Aminah Kigozi, Centers for Disease Control and Prevention–Uganda, Global AIDS Program, National Center for HIV, STD and TB Prevention, PO Box 49, Entebbe, Uganda. Arthur Reingold and Alan Hubbard, Division of Epidemiology, School of Public Health, University of California, Berkeley, CA 94720. Ben M. Hunt, Philip J. Rosenthal, and Grant Dorsey, Department of Medicine, San Francisco General Hospital, University of California, San Francisco, CA 94110.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. J Nunn, P. Mwaba, C. Chintu, A. Mwinga, J. H Darbyshire, A. Zumla, and for the UNZA-UCLMS Project LUCOT Collaboration Role of co-trimoxazole prophylaxis in reducing mortality in HIV infected adults being treated for tuberculosis: randomised clinical trial BMJ, July 10, 2008; 337(jul10_1): a257 - a257. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. K. LAUFER and C. V. PLOWE Cotrimoxazole prophylaxis and malaria in Africa: have the important questions been answered? Am J Trop Med Hyg, September 1, 2006; 75(3): 373 - 374. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
