• View in gallery

    Sample selection. Decision tree that led to the selection of 534 samples for genotyping analysis. The primary outcome of interest was risk of recrudescence for all episodes of uncomplicated falciparum malaria treated with study drugs (sulfadoxine/pyrimethamine [SP] plus placebo, SP plus amodiaquine, and SP plus artesunate).

  • View in gallery

    Parasite genetic diversity detected by msp-1, msp-2, and glurp. Polymerase chain reaction (PCR) products were categorized into molecular weight groups differing by 10 bp for msp-1 and msp-2 and by 20 bp for glurp. (A) msp-1. White bars represent K1; black bars, MAD-20; and checked bars, RO33 allelic families. The 123 K1 alleles ranged in size from 80 to 380 bp; the 35 MAD20 alleles, 80 to 350 bp; and the 55 RO33 alleles, 110 to 160 bp. (B) msp-2. White bars represent FC27 and black bars represent IC3D7 allelic families. The 81 FC27 alleles ranged in size from 280 to 590 bp, and the 130 IC3D7 alleles ranged in size from 170 to 510 bp. (C) glurp. The glurp alleles ranged in size from 340 to 1,040 bp.

  • View in gallery

    Comparison of reinfection survival functions for concordant and discordant genotyping outcomes. Survival analysis techniques were used to compare the accuracy of results based on msp-1 and msp-2 (A) and msp-2 and all 3 genes combined (B). For each comparison pair, the risk of reinfection over 100 days of follow-up based on discordant outcomes (dotted lines) was compared with the same standard based on concordant results (solid lines). P < 0.05 represents a significant difference in the 2 reinfection survival functions compared. It is assumed that reinfection rates should be the same for concordant and discordant outcomes. Compared with msp-1, msp-2 analysis seems to offer a more accurate representation of outcomes, and there was no significant difference between analyses based on msp-2 and all 3 genes.

  • 1

    White N, 1999. Antimalarial drug resistance and mortality in falciparum malaria. Trop Med Int Health 4 :469–470.

  • 2

    World Health Organization, 1996. Assessment of therapeutic efficacy of antimalarial drugs for uncomplicated falciparum malaria in areas with intense transmission. Geneva: World Health Organization.

  • 3

    Snounou G, Beck H, 1998. The ose of PCR genotyping in the assessment of recrudescence or reinfection after antimalarial drug treatment. Parasitol Today 14 :462–467.

    • Search Google Scholar
    • Export Citation
  • 4

    Giao PT, Binh TQ, Kager PA, Long HP, Van Thang N, Van Nam N, de Vries PJ, 2001. Artemisinin for treatment of uncomplicated falciparum malaria: is there a place for monotherapy? Am J Trop Med Hyg 65 :690–695.

    • Search Google Scholar
    • Export Citation
  • 5

    Adjuik M, Agnamey P, Babiker A, Borrmann S, Brasseur P, Cisse M, Cobelens F, Diallo S, Faucher JF, Garner P, Gikunda S, Kremsner PG, Krishna S, Lell B, Loolpapit M, Matsiegui PB, Missinou MA, Mwanza J, Ntoumi F, Olliaro P, Osimbo P, Rezbach P, Some E, Taylor WR, 2002. Amodiaquine-artesunate versus amodiaquine for uncomplicated Plasmodium falciparum malaria in African children: a randomised, multicentre trial. Lancet 359 :1365–1372.

    • Search Google Scholar
    • Export Citation
  • 6

    Dorsey G, Njama D, Kamya MR, Cattamanchi A, Kyabayinze D, Staedke S, Gasasira A, Rosenthal PJ, 2002. Short and long-term efficacies for uncomplicated malaria of sulfadoxine/ pyrimethamine alone or in combination with amodiaquine or artesunate: a longitudinal randomized trial in Ugandan children. Lancet (in press).

  • 7

    Viriyakosol S, Siripoon N, Petcharapirat C, Petcharapirat P, Jarra W, Thaithong S, Brown KN, Snounou G, 1995. Genotyping of Plasmodium falciparum isolates by the polymerase chain reaction and potential uses in epidemiological studies. Bull World Health Organ 73 :85–95.

    • Search Google Scholar
    • Export Citation
  • 8

    Plowe CV, Djimde A, Bouare M, Doumbo O, Wellems TE, 1995. Pyrimethamine and proguanil resistance-conferring mutations in Plasmodium falciparum dihydrofolate reductase: polymerase chain reaction methods for surveillance in Africa. Am J Trop Med Hyg 52 :565–568.

    • Search Google Scholar
    • Export Citation
  • 9

    Zwetyenga J, Rogier C, Tall A, Fontenille D, Snounou G, Trape JF, Mercereau-Puijalon O, 1998. No influence of age on infection complexity and allelic distribution in Plasmodium falciparum infections in Ndiop, a Senegalese village with seasonal, mesoendemic malaria. Am J Trop Med Hyg 59 :726–735.

    • Search Google Scholar
    • Export Citation
  • 10

    Kimura E, Mattei D, di Santi SM, Scherf A, 1990. Genetic diversity in the major merozoite surface antigen of Plasmodium falciparum: high prevalence of a third polymorphic form detected in strains derived from malaria patients. Gene 91 :57–62.

    • Search Google Scholar
    • Export Citation
  • 11

    Fenton B, Clark JT, Khan CM, Robinson JV, Walliker D, Ridley R, Scaife JG, McBride JS, 1991. Structural and antigenic polymorphism of the 35- to 48-kilodalton merozoite surface antigen (MSA-2) of the malaria parasite Plasmodium falciparum.Mol Cell Biol 11 :963–974.

    • Search Google Scholar
    • Export Citation
  • 12

    Borre MB, Dziegiel M, Hogh B, Petersen E, Rieneck K, Riley E, Meis JF, Aikawa M, Nakamura K, Harada M, Wind A, Jakobsen PH, Cowland J, Jepsen S, Axelsen NH, Vuust J, 1991. Primary structure and localization of a conserved immunogenic Plasmodium falciparum glutamate rich protein (GLURP) expressed in both the preerythrocytic and erythrocytic stages of the vertebrate life cycle. Mol Biochem Parasitol 49 :119–131.

    • Search Google Scholar
    • Export Citation
  • 13

    Magesa SM, Mdira KY, Farnert A, Simonsen PE, Bygbjerg IC, Jakobsen PH, 2001. Distinguishing Plasmodium falciparum treatment failures from re-infections by using polymerase chain reaction genotyping in a holoendemic area in northeastern Tanzania. Am J Trop Med Hyg 65 :477–483.

    • Search Google Scholar
    • Export Citation
  • 14

    Winstanley P, 2001. Chlorproguanil-dapsone (LAPDAP) for uncomplicated falciparum malaria. Trop Med Int Health 6 :952–954.

  • 15

    Basco LK, Ringwald P, 2000. Molecular epidemiology of malaria in Yaounde, Cameroon. VII. Analysis of recrudescence and reinfection in patients with uncomplicated falciparum malaria. Am J Trop Med Hyg 63 :215–221.

    • Search Google Scholar
    • Export Citation
  • 16

    Amukoye E, Winstanley PA, Watkins WM, Snow RW, Hatcher J, Mosobo M, Ngumbao E, Lowe B, Ton M, Minyiri G, Marsh K, 1997. Chlorproguanil-dapsone: effective treatment for uncomplicated falciparum malaria. Antimicrob Agents Chemother 41 :2261–2264.

    • Search Google Scholar
    • Export Citation
  • 17

    von Seidlein L, Bojang K, Jones P, Jaffar S, Pinder M, Obaro S, Doherty T, Haywood M, Snounou G, Gemperli B, Gathmann I, Royce C, McAdam K, Greenwood B, 1998. A randomized controlled trial of artemether/benflumetol, a new antimalarial and pyrimethamine/sulfadoxine in the treatment of uncomplicated falciparum malaria in African children. Am J Trop Med Hyg 58 :638–644.

    • Search Google Scholar
    • Export Citation
  • 18

    Watkins WM, Mosobo M, 1993. Treatment of Plasmodium falciparum malaria with pyrimethamine-sulfadoxine: selective pressure for resistance is a function of long elimination half-life. Trans R Soc Trop Med Hyg 87 :75–78.

    • Search Google Scholar
    • Export Citation
  • 19

    Beck HP, Felger I, Huber W, Steiger S, Smith T, Weiss N, Alonso P, Tanner M, 1997. Analysis of multiple Plasmodium falciparum infections in Tanzanian children during the phase III trial of the malaria vaccine SPf66. J Infect Dis 175 :921–926.

    • Search Google Scholar
    • Export Citation
  • 20

    Brockman A, Paul RE, Anderson TJ, Hackford I, Phaiphun L, Looareesuwan S, Nosten F, Day KP, 1999. Application of genetic markers to the identification of recrudescent Plasmodium falciparum infections on the northwestern border of Thailand. Am J Trop Med Hyg 60 :14–21.

    • Search Google Scholar
    • Export Citation
  • 21

    van Vugt M, Brockman A, Gemperli B, Luxemburger C, Gathmann I, Royce C, Slight T, Looareesuwan S, White NJ, Nosten F, 1998. Randomized comparison of artemether-benflumetol and artesunate-mefloquine in treatment of multidrug-resistant falciparum malaria. Antimicrob Agents Chemother 42 :135–139.

    • Search Google Scholar
    • Export Citation
  • 22

    Irion A, Felger I, Abdulla S, Smith T, Mull R, Tanner M, Hatz C, Beck HP, 1998. Distinction of recrudescences from new infections by PCR-RFLP analysis in a comparative trial of CGP 56 697 and chloroquine in Tanzanian children. Trop Med Int Health 3 :490–497.

    • Search Google Scholar
    • Export Citation
  • 23

    Ranford-Cartwright LC, Taylor J, Umasunthar T, Taylor LH, Babiker HA, Lell B, Schmidt-Ott JR, Lehman LG, Walliker D, Kremsner PG, 1997. Molecular analysis of recrudescent parasites in a Plasmodium falciparum drug efficacy trial in Gabon. Trans R Soc Trop Med Hyg 91 :719–724.

    • Search Google Scholar
    • Export Citation
  • 24

    Farnert A, Arez AP, Babiker HA, Beck HP, Benito A, Bjorkman A, Bruce MC, Conway DJ, Day KP, Henning L, Mercereau-Puijalon O, Ranford-Cartwright LC, Rubio JM, Snounou G, Walliker D, Zwetyenga J, do Rosario VE, 2001. Genotyping of Plasmodium falciparum infections by PCR: a comparative multicentre study. Trans R Soc Trop Med Hyg 95 :225–232.

    • Search Google Scholar
    • Export Citation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DISTINGUISHING RECRUDESCENCE FROM REINFECTION IN A LONGITUDINAL ANTIMALARIAL DRUG EFFICACY STUDY: COMPARISON OF RESULTS BASED ON GENOTYPING OF MSP-1, MSP-2, AND GLURP

View More View Less
  • 1 Department of Medicine, San Francisco General Hospital, University of California, San Francisco, San Francisco, California; Department of Biochemistry, Makerere University Medical School, Kampala, Uganda; Department of Biostatistics, School of Public Health, University of California, Berkeley, Berkeley, California

Genotyping frequently is used to distinguish recrudescent from new infections in antimalarial drug efficacy trials, but methodology and interpretation of results have not been standardized. We compared the utility of polymorphisms within 3 Plasmodium falciparum genes during a longitudinal trial in Kampala, Uganda. Merozoite surface protein-1 (msp-1) and merozoite surface protein-2 (msp-2) revealed greater diversity than glutamate-rich protein. Genotypes based on msp-1, msp-2, and all 3 genes combined were compared for 394 initial and subsequent isolates. Classification of most episodes as due to recrudescence or reinfection was straightforward. In 24% (msp-1), 16% (msp-2), and 62% (3 genes combined) of samples, subsequent episodes contained identical and new alleles, however. Our analysis suggested that such episodes should be classified as reinfections and not recrudescence. Comparing the 3 studied genes, msp-2 results were most accurate, and analysis of this single gene effectively distinguished recrudescence from reinfection in our study population.

INTRODUCTION

Resistance to antimalarial therapy is a major concern in Africa.1 Decisions regarding drug policy rely largely on the results of in vivo studies that assess clinical and parasitologic outcomes after therapy. Such studies are limited by their inability to determine whether disease that recurs after therapy is due to recrudescence of resistant parasites or reinfection with new parasite strains. Many studies (and World Health Organization [WHO] guidelines) attempt to minimize this limitation by restricting follow-up to 14 days because during this time reinfection is considered unlikely.2 This approach may underestimate significantly, however, the true risk of treatment failure.3 This underestimation may be especially important when studying drugs with short half-lives, such as artesunate,4 and effective combination regimens,5,6 where a majority of treatment failures may occur after 14 days.

Molecular genotyping is used increasingly to help distinguish recrudescence from reinfection in antimalarial drug efficacy studies.3 Infecting strains of malaria parasites can be “fingerprinted” through polymerase chain reaction (PCR) amplification of polymorphic genes.7 The “fingerprint” patterns of isolates causing successive episodes of malaria can then be compared to distinguish recrudescent from newly infecting parasites. Although the principle behind molecular genotyping is straightforward, the methods used to interpret results have varied widely. In particular, the definitions used to classify treatment failures as recrudescence or reinfection have not been consistent. This underestimation may be especially important when studying drugs with short half-lives, such as artesunate, and effective combination regimens, where a majority of treatment failures may occur after 14 days.

In the present study, we analyzed polymorphisms within the 3 genes—merozoite surface protein-1 (msp-1), merozoite surface protein-2 (msp-2), and glutamate-rich protein (glurp)—most commonly evaluated to distinguish recrudescence from reinfection in antimalarial drug efficacy studies. Our database was from a longitudinal comparison of 3 antimalarial regimens for the treatment of uncomplicated malaria in Kampala, Uganda. Statistical techniques were used to compare different definitions of recrudescence and reinfection and to evaluate the utility of results obtained with each gene and all 3 genes combined. We discuss the impact of different approaches and offer a simple genotyping system using msp-2 alone to type adequately infecting parasites from our study population.

MATERIALS AND METHODS

Clinical trial.

The clinical study took place between July 2000 and August 2001 at the outpatient department of Mulago Hospital, Kampala, Uganda. The full details of the clinical trial were published previously.6 Briefly, 316 healthy children between the ages of 6 months and 5 years were enrolled from the community and randomly assigned to 1 of 3 regimens, which then was given for all episodes of uncomplicated malaria diagnosed during 12 months of follow-up. These regimens were sulfadoxine/pyrimethamine (SP) plus placebo, SP plus amodiaquine, and SP plus artesunate. For each episode of uncomplicated malaria diagnosed, patients were treated with their same preassigned regimen, and outcomes were assessed using a standard WHO 14-day follow-up protocol.2 Clinical treatment failures occurring within 14 days were treated with standard doses of quinine. Any case of malaria diagnosed >14 days after a previous episode was classified as a new episode for treatment purposes.

Sample selection and polymerase chain reaction genotyping.

Blood was collected on filter paper (Whatman No. 3, Whatman Inc., Clifton, NJ) each time a new episode of malaria was diagnosed and when treatment failure occurred. Molecular genotyping was performed for all successive episodes of uncomplicated Plasmodium falciparum malaria treated with study drugs except that failures within 6 days of treatment were classified as recrudescence, episodes caused by P. falciparum followed by nonfalciparum malaria were classified as reinfections, and episodes after quinine therapy (for severe malaria or clinical failure within 14 days) were considered new infections.

Parasite DNA was extracted with chelex 100 Resin (Bio-Rad Laboratories, Hercules, CA) as previously described.8 The polymorphic regions of msp-1 (block 2), msp-2 (block 3), and glurp (R2 repeat region) were amplified by nested PCR. First-round PCR primers corresponded to conserved sequences flanking these regions.9 Second-round PCR primers were used to amplify the K1, MAD20, and RO33 allelic families of msp-110; the IC3D7 and FC27 allelic families of msp-211; and the repeat region of glurp.12 For controls, genomic DNA from HB3 and 3D7 laboratory strains was isolated by standard techniques.

Nested PCR products were analyzed by electrophoresis using 2% agarose for msp-1 and msp-2 and 1.5% agarose for glurp. Samples from an individual patient were run in adjacent lanes. If there was no amplification for any allelic family, PCR was repeated with 5 times the quantity of template DNA. If no amplification was detected after this second reaction, amplification was classified as unsuccessful. Gel images were digitized and molecular weights were assigned to bands using GelCompar II software (Applied Maths, Sint-Martens-Latem, Belgium). Densitometric curves were calculated for each gel lane, and bands with >2% of the density of the dominant band in each lane were assigned molecular weights.

Data interpretation.

Genotyping patterns were compared for successive episodes of uncomplicated P. falciparum malaria. Based on analysis of the HB3 strain, alleles were considered the same if molecular weights were within 10 bp for msp-1 and msp-2 and within 20 bp for glurp. It was assumed that after a patient initially was treated for malaria, a subsequent episode was caused by either parasite strains present before treatment (e.g., recrudescence) or parasite strains acquired after treatment (e.g., reinfection). An outcome was defined as recrudescence if a subsequent sample contained identical alleles or a subset of the alleles present in the first sample. An outcome was defined as reinfection if a subsequent sample contained only new alleles. If a subsequent sample contained alleles present in the first sample and new alleles, the outcome initially was considered indeterminate.

Statistical analysis.

Different criteria for classifying indeterminate outcomes were compared using a statistical modeling approach. Treatment regimen and pretreatment parasite density were identified as strong independent predictors of treatment outcome using a Cox proportional hazards model. Measures of association (hazard ratios) for these predictor variables first were estimated only for episodes in which the outcomes were defined clearly as recrudescence or reinfection. These parameter estimates were considered the reference standards. The analysis was repeated after including episodes with indeterminate outcomes, using varied classification criteria. The best method for classifying indeterminate outcomes was considered that in which the new parameter estimates were closest to our reference standards. In our comparisons of different genes and the application of genotyping to clinical results, the risks of recrudescence and reinfection were estimated using Kaplan-Meier survival analysis techniques. Survival functions were compared using the log-rank test.

RESULTS

Performance of polymerase chain reaction genotyping.

Our clinical study included 661 treatments for malaria. PCR genotyping was performed on 534 samples for which the results could affect the classification of treatment outcome (Figure 1). Genotyping was not performed if an episode was not preceded or followed by a study drug treatment (n = 59), subsequent treatment was <7 days after initial treatment (n = 23), the episode was followed by nonfalciparum malaria (n = 2), or the sample was missing (n = 4). PCR amplification was successful in 99%, 96%, and 98% of samples for msp-1, msp-2, and glurp and in 96% of samples when all 3 genes were combined. Of a possible 392 paired primary and subsequent samples, PCR results were available for 386, 371, 379, and 367 pairs for msp-1, msp-2, glurp, and when all 3 genes were combined.

Population diversity.

To estimate the genetic diversity of msp-1, msp-2, and glurp in the parasite population, the frequency distribution of alleles was determined for the first episode of malaria of each study subject (n = 115) (Figure 2). msp-1 (45 alleles) and msp-2 (57 alleles) revealed considerably greater parasite diversity than glurp (28 alleles). Based on a simple multinomial probability equation, the likelihood of 2 random infections having identical alleles was 2% for msp-2, 5% for msp-1, and 7% for glurp. The mean number of alleles per sample (complexity of infection) also was greater with msp-1 (1.7) and msp-2 (1.7) than with glurp (1.4). Because of the lesser degree of polymorphism detected by glurp, it was not considered further as a single genetic marker.

Classification of indeterminate outcomes.

The percentage of indeterminate outcomes (in which a subsequent episode contained alleles present in the previous episode and new alleles) was 24% (93 of 386) for results based on msp-1 alone, 16% (58 of 371) for results based on msp-2 alone, and 62% (226 of 367) for results based on all 3 genes. To classify best outcomes initially considered indeterminate, parameter estimates (hazard ratios) for samples with only definitive outcomes (our reference standards) were compared with estimates obtained when indeterminate outcomes were included (Table 1). Three classification schemes for indeterminate outcomes were considered: (1) all indeterminate outcomes considered recrudescence, (2) indeterminate outcomes considered reinfection if ≥50% of the alleles in the subsequent sample were new (in this scheme, 90%, 91%, and 71% of episodes were classified as reinfections using results based on msp-1, msp-2, and all 3 genes), and (3) all indeterminate outcomes considered reinfections. Compared with our reference standards, parameter estimates changed considerably when indeterminate outcomes all were considered recrudescence (scheme 1). In contrast, parameter estimates were similar when all indeterminate outcomes were classified as reinfections (scheme 3) or most were classified as reinfections (scheme 2). Based on our prediction model, episodes with indeterminate outcomes were much more consistent with reinfection than recrudescence.

Comparison of analyses using msp-1, msp-2, or all 3 genes.

Reinfection survival functions were used to compare genotyping results based on either msp-1 or msp-2 alone, with indeterminate outcomes classified by scheme 2, as detailed earlier. Outcomes were concordant (classified as either reinfection or recrudescence by msp-1 and msp-2 analyses) in 79% (294 of 370) and discordant in 21% (76 of 370) of comparisons. Reinfection rates were not associated with treatment group or pretreatment parasite densities (data not shown). We assumed that reinfection survival functions should be similar whether results were concordant or discordant. The survival functions based on msp-2 results fit this assumption (Figure 3). In contrast, survival functions based on msp-1 results showed significant differences between concordant and discordant outcomes, suggesting a less accurate estimate of true outcomes. For the same comparison between msp-2 and all 3 genes, outcomes were concordant in 89% (327 of 367) and discordant in only 11% (40 of 367) of comparisons, and survival functions for concordant and discordant data were not significantly different (Figure 3).

Application of genotyping to clinical results.

The risk of recrudescence is the primary outcome of interest for antimalarial drug efficacy studies. Using survival analysis techniques with the follow-up period extended to 42 days, we evaluated how using different genes and different methods of classifying indeterminate outcomes affected the calculated risk of recrudescence in our clinical study (Table 2). When comparing different classification schemes for indeterminate outcomes, scheme 2 provided the most robust results across the different genes (msp-1 alone, msp-2 alone, and all 3 genes combined). Compared with results based on msp-1 or msp-2 alone, the use of 3 genes combined showed the greatest variation in results across different indeterminate outcome classification schemes because of the larger number of indeterminate outcomes.

DISCUSSION

Molecular genotyping is a simple and practical tool to distinguish between recrudescence and reinfection in antimalarial drug efficacy studies. Genotyping is particularly relevant in high transmission areas13 and when follow-up is >14 days.5,6,14–17 Although genotyping analyses have been reported in many trials, the methodology and interpretation of results have not been standardized. In particular, the choice and number of genetic markers and the definitions used to classify outcomes as recrudescence or reinfection have not been consistent.

We have used results from a longitudinal study of antimalarial drug efficacy to compare methodologies critically and offer a standardized genotyping approach. In this study, we had a large sample size, extended follow-up covering multiple episodes of disease, and 3 different treatment groups covering a wide range of risk for recrudescence. We found that polymorphisms in either msp-1 or msp-2 were sufficiently diverse for us to distinguish adequately recrudescence from reinfection in our study population. In a direct comparison of analyses based on msp-1 and msp-2 polymorphisms, msp-2 provided a more accurate measure of treatment outcomes. Analyses considering polymorphisms in glurp and all 3 genes combined provided no additional utility. We propose a standard approach in which the analysis of msp-2 polymorphisms is used to distinguish recrudescence from reinfection.

Our genotyping results agreed with expected outcomes. Most recrudescences occurred within 28 days after treatment. The risk of recrudescence was highest in patients treated with SP monotherapy, the regimen that also was the least effective based on clinical and parasitologic criteria. Most reinfections occurred >28 days after treatment. The rate of reinfection was relatively constant after 42 days, when levels of all study drugs were expected to be low.18

Other investigators have suggested that in areas of high transmission, evaluation of recrudescence is achieved best by detailed analysis of a single, highly polymorphic locus.19 It also has been argued that the use of multiple markers may enhance the detection of diversity and decrease the chance of misclassifying new infections as recrudescence.3 In this study, we detected considerable parasite diversity with analyses of either msp-1 or msp-2, resulting in an acceptably low probability of outcome misclassification. When results from all 3 genes combined were used, the level of diversity increased but at the cost of a considerably larger proportion of outcomes being classified as indeterminate.

It was important to determine how best to classify outcomes initially classified as indeterminate. In these cases, patients developed recurrent malaria, and infecting parasites contained alleles present in the prior episode and new alleles. In our sample, the proportion of outcomes fitting this category was 24% using msp-1, 16% using msp-2, and 62% using all 3 genes. Approaches considered in previous studies have included omitting indeterminate outcomes from the analysis,20,21 considering them all as either recrudescences or reinfections,17,22,23 or other more complex methods.13,15 In many studies, it is unclear what method was used. In this study, we compared 3 different classification schemes, using as the reference standard the subset of outcomes that were definitive (either recrudescence, when subsequent isolates contained only alleles present in the prior episode, or reinfection, when a subsequent isolate contained only new alleles). Our analysis showed that the episodes initially classified as indeterminate were much more likely to be caused by reinfection than by recrudescence. Based on this result, we hypothesize that if residual and new parasite strains cocirculate, clinical symptoms are more likely to be caused by the new strain. This hypothesis is consistent with data from Gabon, in which the time to reappearance of parasites was similar in distribution when subsequent episodes contained only new alleles or when subsequent episodes contained new and old alleles.23

PCR-based genotyping is potentially limited by the sensitivity of the technique for detecting all allelic variants present in an infection.3 This can be especially problematic with low-level parasitemia or when one strain is a minor component of a multistrain infection.24 The problem of sensitivity for multiple alleles is relatively unimportant in drug efficacy studies, however, in which strains responsible for clinical illness are likely to be dominant and detected by PCR. Consistent with this conclusion, in our study 92% of our samples had a parasite density ≥2,000 asexual parasites/μl (96% ≥500 asexual parasites/μl), and geometric mean parasite density was not associated with the number of strains detected (P = 0.23).

The presence of gametocytes in blood samples can confound genotyping analyses.3 Outcomes can be misclassified as recrudescence if asexual parasites are cleared but gametocytes originating from the primary episode are detected at the time that a subsequent episode is diagnosed. Although the presence of gametocytes in subsequent samples was associated with recrudescence (P = 0.001), gametocytes occurred in only 16 of 371 paired samples (11 of these in the group treated with SP alone), suggesting that confounding as a result of gametocytemia was not a significant problem in our study.

Genotyping is an important tool for antimalarial drug efficacy studies, but its value has been limited by inconsistent methodology. We have shown that the results of our clinical trial varied considerably depending on the choice of genes for analysis and the interpretation of genotyping results. We have shown that in this high-transmission area, a single, highly polymorphic genetic marker adequately distinguished between recrudescence and reinfection. We support the analysis of msp-2 polymorphisms because this genetic marker provided more accurate results compared with msp-1 and equivalent results compared with all 3 genes combined. More studies are needed, however, to determine if msp-2 or another marker is the appropriate choice in other high-transmission areas. In low-endemicity regions, multiple genetic markers may be needed to detect sufficient parasite diversity. In addition, we recommend that when recurrent isolates contain old and new alleles, outcomes should be classified as reinfection. Our hope is that additional studies will allow the development of a simple, standardized approach to genotyping analyses in endemic areas of Africa. This approach should allow for more accurate estimation of treatment failure risk and improve comparability among studies.

Table 1

Comparison of indeterminate outcome classification schemes

Hazard ratios for independent predictors
GeneModel*Treatment With SPTreatment with SP/ASLog parasite density
SP, sulfadoxine/pyrimethamine; AS, artesunate.
* Reference standard includes only outcomes clearly considered recrudescence or reinfection. Other models include indeterminate outcomes based on different classification schemes. Scheme 1: All indeterminate outcomes considered recrudescence. Scheme 2: Indeterminate outcomes considered reinfection if half or more of the alleles in the subsequent sample were new. Scheme 3: All indeterminate outcomes considered reinfection.
All P < 0.03.
Compared with baseline of treatment with SP/AQ (amodiaquine).
msp-1Reference standard7.03.31.23
Scheme 13.82.01.17
Scheme 27.03.11.23
Scheme 36.73.31.24
msp-2Reference standard6.72.91.31
Scheme 13.91.81.18
Scheme 26.82.81.31
Scheme 37.03.01.33
All 3 genesReference standard10.54.11.32
Scheme 12.41.41.16
Scheme 26.72.61.24
Scheme 312.04.21.33
Table 2

Risk of recrudescence after 42 days of follow-up among treatment groups

msp-1msp-2All 3 genes
Treatment GroupScheme 1Scheme 2Scheme 3Scheme 1Scheme 2Scheme 3Scheme 1Scheme 2Scheme 3
Note. Schemes as defined in Table 1.
SP, sulfadoxine/pyrimethamine; AQ, amodiaquine; AS, artesunate.
SP47%40%38%47%39%38%52%42%29%
SP/AQ12%5%5%10%6%5%20%6%3%
SP/AS30%22%22%25%21%20%37%20%14%
Figure 1.
Figure 1.

Sample selection. Decision tree that led to the selection of 534 samples for genotyping analysis. The primary outcome of interest was risk of recrudescence for all episodes of uncomplicated falciparum malaria treated with study drugs (sulfadoxine/pyrimethamine [SP] plus placebo, SP plus amodiaquine, and SP plus artesunate).

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 68, 2; 10.4269/ajtmh.2003.68.133

Figure 2.
Figure 2.

Parasite genetic diversity detected by msp-1, msp-2, and glurp. Polymerase chain reaction (PCR) products were categorized into molecular weight groups differing by 10 bp for msp-1 and msp-2 and by 20 bp for glurp. (A) msp-1. White bars represent K1; black bars, MAD-20; and checked bars, RO33 allelic families. The 123 K1 alleles ranged in size from 80 to 380 bp; the 35 MAD20 alleles, 80 to 350 bp; and the 55 RO33 alleles, 110 to 160 bp. (B) msp-2. White bars represent FC27 and black bars represent IC3D7 allelic families. The 81 FC27 alleles ranged in size from 280 to 590 bp, and the 130 IC3D7 alleles ranged in size from 170 to 510 bp. (C) glurp. The glurp alleles ranged in size from 340 to 1,040 bp.

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 68, 2; 10.4269/ajtmh.2003.68.133

Figure 3.
Figure 3.

Comparison of reinfection survival functions for concordant and discordant genotyping outcomes. Survival analysis techniques were used to compare the accuracy of results based on msp-1 and msp-2 (A) and msp-2 and all 3 genes combined (B). For each comparison pair, the risk of reinfection over 100 days of follow-up based on discordant outcomes (dotted lines) was compared with the same standard based on concordant results (solid lines). P < 0.05 represents a significant difference in the 2 reinfection survival functions compared. It is assumed that reinfection rates should be the same for concordant and discordant outcomes. Compared with msp-1, msp-2 analysis seems to offer a more accurate representation of outcomes, and there was no significant difference between analyses based on msp-2 and all 3 genes.

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 68, 2; 10.4269/ajtmh.2003.68.133

Authors’ addresses: Adithya Cattamanchi, Philip J. Rosenthal, and Grant Dorsey, Department of Medicine, San Francisco General Hospital, University of California, San Francisco, P.O. Box 0811, San Francisco, CA 94143, Telephone: 415-206-8687, Fax: 415-648-8425. Daniel Kyabayinze, Department of Biochemistry, Makerere University Medical School, P.O. Box 7072, Kampala, Uganda. Alan Hubbard, Department of Biostatistics, School of Public Health, University of California, Berkeley, 735 University Hall, Berkeley, CA 94720, Telephone: 510-642-8365, Fax: 510-642-5815.

Acknowledgments: We thank the clinical study team of B.M. Karakire, Marx Dongo, Sam Nsobya, Moses Kiggundu, Christopher Bongole, Regina Nakafero, Bridget K. Nzarubara, Pauline Byakika, and Sarah Kibirango; Sarah Staedke, for reviewing the manuscript; community leaders from the Kawempe Division of Kampala; and the study participants and their parents or guardians.

Financial support: This work was supported by the Fogarty International Center/National Institutes of Health (TW00007, TW01506, and AI43301), the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR), and the Doris Duke Charitable Foundation Clinical Research Fellowship.

REFERENCES

  • 1

    White N, 1999. Antimalarial drug resistance and mortality in falciparum malaria. Trop Med Int Health 4 :469–470.

  • 2

    World Health Organization, 1996. Assessment of therapeutic efficacy of antimalarial drugs for uncomplicated falciparum malaria in areas with intense transmission. Geneva: World Health Organization.

  • 3

    Snounou G, Beck H, 1998. The ose of PCR genotyping in the assessment of recrudescence or reinfection after antimalarial drug treatment. Parasitol Today 14 :462–467.

    • Search Google Scholar
    • Export Citation
  • 4

    Giao PT, Binh TQ, Kager PA, Long HP, Van Thang N, Van Nam N, de Vries PJ, 2001. Artemisinin for treatment of uncomplicated falciparum malaria: is there a place for monotherapy? Am J Trop Med Hyg 65 :690–695.

    • Search Google Scholar
    • Export Citation
  • 5

    Adjuik M, Agnamey P, Babiker A, Borrmann S, Brasseur P, Cisse M, Cobelens F, Diallo S, Faucher JF, Garner P, Gikunda S, Kremsner PG, Krishna S, Lell B, Loolpapit M, Matsiegui PB, Missinou MA, Mwanza J, Ntoumi F, Olliaro P, Osimbo P, Rezbach P, Some E, Taylor WR, 2002. Amodiaquine-artesunate versus amodiaquine for uncomplicated Plasmodium falciparum malaria in African children: a randomised, multicentre trial. Lancet 359 :1365–1372.

    • Search Google Scholar
    • Export Citation
  • 6

    Dorsey G, Njama D, Kamya MR, Cattamanchi A, Kyabayinze D, Staedke S, Gasasira A, Rosenthal PJ, 2002. Short and long-term efficacies for uncomplicated malaria of sulfadoxine/ pyrimethamine alone or in combination with amodiaquine or artesunate: a longitudinal randomized trial in Ugandan children. Lancet (in press).

  • 7

    Viriyakosol S, Siripoon N, Petcharapirat C, Petcharapirat P, Jarra W, Thaithong S, Brown KN, Snounou G, 1995. Genotyping of Plasmodium falciparum isolates by the polymerase chain reaction and potential uses in epidemiological studies. Bull World Health Organ 73 :85–95.

    • Search Google Scholar
    • Export Citation
  • 8

    Plowe CV, Djimde A, Bouare M, Doumbo O, Wellems TE, 1995. Pyrimethamine and proguanil resistance-conferring mutations in Plasmodium falciparum dihydrofolate reductase: polymerase chain reaction methods for surveillance in Africa. Am J Trop Med Hyg 52 :565–568.

    • Search Google Scholar
    • Export Citation
  • 9

    Zwetyenga J, Rogier C, Tall A, Fontenille D, Snounou G, Trape JF, Mercereau-Puijalon O, 1998. No influence of age on infection complexity and allelic distribution in Plasmodium falciparum infections in Ndiop, a Senegalese village with seasonal, mesoendemic malaria. Am J Trop Med Hyg 59 :726–735.

    • Search Google Scholar
    • Export Citation
  • 10

    Kimura E, Mattei D, di Santi SM, Scherf A, 1990. Genetic diversity in the major merozoite surface antigen of Plasmodium falciparum: high prevalence of a third polymorphic form detected in strains derived from malaria patients. Gene 91 :57–62.

    • Search Google Scholar
    • Export Citation
  • 11

    Fenton B, Clark JT, Khan CM, Robinson JV, Walliker D, Ridley R, Scaife JG, McBride JS, 1991. Structural and antigenic polymorphism of the 35- to 48-kilodalton merozoite surface antigen (MSA-2) of the malaria parasite Plasmodium falciparum.Mol Cell Biol 11 :963–974.

    • Search Google Scholar
    • Export Citation
  • 12

    Borre MB, Dziegiel M, Hogh B, Petersen E, Rieneck K, Riley E, Meis JF, Aikawa M, Nakamura K, Harada M, Wind A, Jakobsen PH, Cowland J, Jepsen S, Axelsen NH, Vuust J, 1991. Primary structure and localization of a conserved immunogenic Plasmodium falciparum glutamate rich protein (GLURP) expressed in both the preerythrocytic and erythrocytic stages of the vertebrate life cycle. Mol Biochem Parasitol 49 :119–131.

    • Search Google Scholar
    • Export Citation
  • 13

    Magesa SM, Mdira KY, Farnert A, Simonsen PE, Bygbjerg IC, Jakobsen PH, 2001. Distinguishing Plasmodium falciparum treatment failures from re-infections by using polymerase chain reaction genotyping in a holoendemic area in northeastern Tanzania. Am J Trop Med Hyg 65 :477–483.

    • Search Google Scholar
    • Export Citation
  • 14

    Winstanley P, 2001. Chlorproguanil-dapsone (LAPDAP) for uncomplicated falciparum malaria. Trop Med Int Health 6 :952–954.

  • 15

    Basco LK, Ringwald P, 2000. Molecular epidemiology of malaria in Yaounde, Cameroon. VII. Analysis of recrudescence and reinfection in patients with uncomplicated falciparum malaria. Am J Trop Med Hyg 63 :215–221.

    • Search Google Scholar
    • Export Citation
  • 16

    Amukoye E, Winstanley PA, Watkins WM, Snow RW, Hatcher J, Mosobo M, Ngumbao E, Lowe B, Ton M, Minyiri G, Marsh K, 1997. Chlorproguanil-dapsone: effective treatment for uncomplicated falciparum malaria. Antimicrob Agents Chemother 41 :2261–2264.

    • Search Google Scholar
    • Export Citation
  • 17

    von Seidlein L, Bojang K, Jones P, Jaffar S, Pinder M, Obaro S, Doherty T, Haywood M, Snounou G, Gemperli B, Gathmann I, Royce C, McAdam K, Greenwood B, 1998. A randomized controlled trial of artemether/benflumetol, a new antimalarial and pyrimethamine/sulfadoxine in the treatment of uncomplicated falciparum malaria in African children. Am J Trop Med Hyg 58 :638–644.

    • Search Google Scholar
    • Export Citation
  • 18

    Watkins WM, Mosobo M, 1993. Treatment of Plasmodium falciparum malaria with pyrimethamine-sulfadoxine: selective pressure for resistance is a function of long elimination half-life. Trans R Soc Trop Med Hyg 87 :75–78.

    • Search Google Scholar
    • Export Citation
  • 19

    Beck HP, Felger I, Huber W, Steiger S, Smith T, Weiss N, Alonso P, Tanner M, 1997. Analysis of multiple Plasmodium falciparum infections in Tanzanian children during the phase III trial of the malaria vaccine SPf66. J Infect Dis 175 :921–926.

    • Search Google Scholar
    • Export Citation
  • 20

    Brockman A, Paul RE, Anderson TJ, Hackford I, Phaiphun L, Looareesuwan S, Nosten F, Day KP, 1999. Application of genetic markers to the identification of recrudescent Plasmodium falciparum infections on the northwestern border of Thailand. Am J Trop Med Hyg 60 :14–21.

    • Search Google Scholar
    • Export Citation
  • 21

    van Vugt M, Brockman A, Gemperli B, Luxemburger C, Gathmann I, Royce C, Slight T, Looareesuwan S, White NJ, Nosten F, 1998. Randomized comparison of artemether-benflumetol and artesunate-mefloquine in treatment of multidrug-resistant falciparum malaria. Antimicrob Agents Chemother 42 :135–139.

    • Search Google Scholar
    • Export Citation
  • 22

    Irion A, Felger I, Abdulla S, Smith T, Mull R, Tanner M, Hatz C, Beck HP, 1998. Distinction of recrudescences from new infections by PCR-RFLP analysis in a comparative trial of CGP 56 697 and chloroquine in Tanzanian children. Trop Med Int Health 3 :490–497.

    • Search Google Scholar
    • Export Citation
  • 23

    Ranford-Cartwright LC, Taylor J, Umasunthar T, Taylor LH, Babiker HA, Lell B, Schmidt-Ott JR, Lehman LG, Walliker D, Kremsner PG, 1997. Molecular analysis of recrudescent parasites in a Plasmodium falciparum drug efficacy trial in Gabon. Trans R Soc Trop Med Hyg 91 :719–724.

    • Search Google Scholar
    • Export Citation
  • 24

    Farnert A, Arez AP, Babiker HA, Beck HP, Benito A, Bjorkman A, Bruce MC, Conway DJ, Day KP, Henning L, Mercereau-Puijalon O, Ranford-Cartwright LC, Rubio JM, Snounou G, Walliker D, Zwetyenga J, do Rosario VE, 2001. Genotyping of Plasmodium falciparum infections by PCR: a comparative multicentre study. Trans R Soc Trop Med Hyg 95 :225–232.

    • Search Google Scholar
    • Export Citation

Author Notes

Reprint requests: Grant Dorsey, Department of Medicine, San Francisco General Hospital, University of California, San Francisco, P.O. Box 0811, San Francisco, CA 94143, Telephone: 415-206-8687, Fax: 415-648-8425, E-mail: grantd@itsa.ucsf.edu
Save