pmc logo image
Logo of nihpaNIHPA bannerabout author manuscriptssubmit a manuscript
Journal List > NIHPA Author Manuscripts

Formats:
Trends Parasitol. Author manuscript; available in PMC 2009 February 23.
Published in final edited form as:
Trends Parasitol. 2008 February; 24(2): 68–73.
Published online 2008 January 7. doi: 10.1016/j.pt.2007.11.003.
PMCID: PMC2646162
NIHMSID: NIHMS68425
Copyright notice and Disclaimer
Can estimates of antimalarial efficacy from field studies be improved?
Michelle L. Gatton1 and Qin Cheng1,2
1 Malaria Drug Resistance and Chemotherapy Laboratory, The Queensland Institute of Medical Research, PO Royal Brisbane Hospital, Herston, Qld 4029, Australia
2 Drug Resistance and Diagnostics, Australian Army Malaria Institute, Gallipoli Barracks, Enoggera, Qld 4051, Australia
Corresponding author: Gatton, M.L. (Email: michelle.gatton/at/qimr.edu.au)
Small right arrow pointing to: The publisher's final edited version of this article is available at Trends Parasitol.
Small right arrow pointing to: See other articles in PMC that cite the published article.
Publisher's Disclaimer
Abstract
Monitoring therapeutic efficacy of antimalarial drugs is important because treatment failure rates are the primary basis for changing antimalarial treatment policy. An important aspect of efficacy studies is the use of PCR genotyping to distinguish recrudescent from new infections. The conclusions reached using this technique might be misleading if there is insufficient parasite diversity or a non-uniform haplotype frequency distribution in the study area. Statistical techniques can be used to overcome this problem, but only when data describing the haplotype frequency distribution are available. Therefore, assessing haplotype frequency and distribution should form an integral part of all studies investigating the therapeutic efficacy of antimalarial treatment regimes.
Plasmodium falciparum and Plasmodium vivax parasites account for most clinical cases of malaria, and over the past two decades both parasites have developed resistance to many of the commonly used and affordable antimalarial treatments. The rapid development and spread of resistance to antimalarial drugs has made providing effective treatment for malaria a major challenge. Drug resistance has also increased the need for monitoring therapeutic efficacy of currently used drugs and drug combinations, and for testing therapeutic efficacy of new drugs and drug combinations.
Monitoring the therapeutic efficacy of antimalarial drugs is usually achieved by assessing the clinical and parasitological outcomes of treatment. Recurrent clinical symptoms and recrudescent parasites indicate reduced parasite sensitivity to the treatment drug. To detect both high and low levels of resistance, the World Health Organization (WHO) recommends that clinical and parasitological outcomes of treatment be assessed after at least 28 days following treatment in areas of high, as well as low to moderate, transmission. A 28-day follow-up period is considered appropriate for amodiaquine, chloroquine and sulfadoxine-pyrimethamine, whereas longer follow-up periods of 42 days and 63 days are recommended for artemether-lumefantrine and mefloquine, respectively [1]. One of the problems associated with such long follow-up periods in areas of continuing transmission is that the recurrence of parasites or clinical symptoms might actually be caused by new infections, not reduced drug efficacy. To attempt to overcome this problem ‘PCR correction’ (see Glossary) of treatment failure data are recommended [1]. This process involves comparing the genotype of parasites in blood samples taken before treatment and during any recurrence.
The concept of using molecular genotyping to distinguish recrudescent from new infections is based on the principle that there is adequate parasite genetic diversity in selected polymorphic markers, such that there is a negligible probability of a new infecting parasite having the same allele or haplotype as the initial infecting parasite. Hence, when an identical allele or haplotype is observed in samples before and after treatment a recrudescence is assumed, whereas a different allele between the paired samples is interpreted as a new infection (for a review, see Ref. [2]). Therapeutic efficacy is determined by the proportion of patients having adequate clinical and parasitological response after adjusting for new infections. The distinction between recrudescent and new parasites is straightforward for monoclonal infections (i.e. only one parasite clone in the sample), but can become complicated when multiple genotypes are present in either the pre-treatment or follow-up sample; several methods of analysing this type of data have been reported [3,4]. Analysis of data from P. vivax infections is also complicated by the possibility of relapses. Previously, it was assumed that parasites causing relapses had the same genotype as the initial infection. However, recent reports suggesting that relapse parasites are often of a different genotype than the initial infection indicate that, at least in some cases, relapse parasites would have a different genotype to the pre-treatment parasites, thereby indicating successful drug treatment [5,6].
Molecular genotyping of Plasmodium parasites is a common procedure in many laboratories around the world. Highly polymorphic antigen genes, such as Pfmsp1, Pfmsp2 and glurp, are common markers for P. falciparum, whereas Pvmsp1, Pvcs and Pvmsp3 are often used to genotype P. vivax. These markers are commonly selected because they are located on different chromosomes, and this reduces the likelihood of linkage disequilibrium. Reports from regions of high transmission indicate that there are more than 20 different alleles for each marker [3,7–11]; the actual alleles detected differ between locations in close proximity and also over time [8]. In regions of lower endemicity genetic diversity is sometimes limited; this seems to be the case in South America for both P. falciparum and P. vivax [12,13].
Genotyping of P. falciparum and P. vivax parasites can also be achieved using polymorphic microsatellite (MS) markers [14–16]. For P. falciparum there are abundant MS loci showing varying degrees of polymorphism [17]. The number of alleles for MS markers differs with up to 18 alleles reported for the Polyα marker in high transmission regions [15]. The relationship observed using MS markers between genetic diversity and transmission intensity is similar to that obtained using antigen markers; namely, that lower transmission areas tend to have low genetic diversity and higher linkage disequilibrium, whereas higher transmission areas show greater diversity and a more random association between marker alleles. There are fewer reported MS loci for P. vivax, but the number is likely to increase with the release of the genome sequence.
When assessing population structure and evolution, it is preferable to type MS markers because they are unlikely to be under selection pressure unless they are located adjacent to drug resistance or antigen loci. However, for the purposes of categorizing post-treatment parasites as the same (recrudescent) or different (new infection) to pre-treatment parasites, either genotyping method is acceptable.
As PCR correction of follow-up data has become standard practice in therapeutic efficacy trials, there has been a decreasing emphasis on determining or reporting parasite diversity and allele or haplotype frequency at the trial site during the period of interest. Generally, it is assumed that use of one or several of the above mentioned markers will be sufficient to adequately differentiate parasites. However, this assumption can be violated if there is insufficient diversity, or if there is a higher frequency of one or a few haplotypes in the parasite population [13,18]. A nonuniform distribution of parasite haplotypes is often more problematic and given less coverage in the literature than the number of different alleles within a population.
The multinomial distribution can be used to calculate the theoretical probability of being infected with the same parasite for any level of diversity or haplotype frequency distribution (Box 1). When undertaking PCR correction of field data, a decision needs to be made about how many markers will be typed. The number of markers selected should aim to achieve a low (e.g. <0.05) theoretical probability of reinfection with a parasite having the same genotype as the pre-treatment parasite. For study sites in sub-Saharan Africa, it has been reported that two markers are sufficient [19]. However, information for other areas is limited. An efficient way to determine how many markers are required is to combine sequential genotyping with probability theory (Figure 1Figure 1). This sequential genotyping involves using the pre-treatment samples to derive a haplotype frequency distribution for the study population that provides ample confidence to distinguish new infections from recrudescence. Figure 2Figure 2 illustrates how this stepwise process works using data reported from different field sites. In each of the examples it was possible to reduce the chance of misclassifying a new infection, with a parasite having the same genetic markers as the pre-treatment parasite, as a recrudescence to less than 0.05. Statistical theory indicates that at least 20 unique alleles (or haplotypes if using more than one marker) are required to reach a threshold of 0.05, but this can only be achieved if each of the 20 alleles is present in equal proportions. Deviation away from this uniform distribution increases the number of alleles required. Data from numerous geographic regions indicate that allele frequency is rarely uniform with >40 alleles for a marker sometimes not being sufficient to reach a threshold of 0.05 (Figure 3Figure 3).
Box 1. Theoretical probability of reinfection with the same parasite
The multinomial distribution can be used to calculate the probability (Pr) of obtaining certain outcomes from a population of different parasites:
equation M1
(1)
where n = the number of independent infections, k = the number of different alleles in the population, Xi = the n umber of infections in a single host having allele i, pi = the probability of being infected with allele equation M2. pi can be approximated by the proportion of the parasite population having allele i.
Using this generic formula, the probability of being infected by two parasites with the same allele is described by:
equation M3
(2)
Figure 1
Figure 1
Figure 1
Proposed protocol for molecular genotyping of parasites to distinguish recrudescence from reinfection. The protocol is an iterative process where individual polymorphic markers are genotyped and the allele (or haplotype) frequency distribution calculated. (more ...)
Figure 2
Figure 2
Figure 2
Determination of the optimal number of genetic markers. Examples of how the stepwise process of genotyping from Figure 1Figure 1 could be implemented using data reported for (a) P. falciparum alleles in Thailand [23], (b) microsatellite markers for P. vivax in (more ...)
Figure 3
Figure 3
Figure 3
Frequency distributions of alleles for individual antigen and microsatellite markers. Data are from P. falciparum isolates in Uganda [3,15] (a, b) and Zimbabwe [15] (c), and P. vivax isolates in India [11] (d). Values of P greater than 0.05 are indicated (more ...)
It is not always possible to reduce the probability of misclassifying a new infection as a recrudescence. This could be caused by limited parasite diversity in the population or genotyping of an insufficient number of markers. Limited diversity has been reported to occur in areas of re-emerging disease or epidemic occurrences of malaria that can be dominated by a few parasite genotypes [20–22]. In such cases, the probability of reinfection with the same parasite should be reported and treatment failure rates statistically adjusted to account for the number of apparent failures that are caused by new infections with the same parasite genotype (Box 2). The statistical adjustment is computationally simple once the haplotype distribution and probability of reinfection by a parasite having the same haplotype have been calculated. It is imperative that parasite diversity is measured during the study period at the field site because it has been demonstrated that the distribution of alleles can vary in the same location over time and also between locations in close proximity [8].
Box 2. Statistical adjustment of treatment failure rates
Following PCR-correction, positive post-treatment samples can be classified into new (D) and recrudescent (S) infections. The number of recrudescent samples is routinely used as the estimate of treatment failure. However, if the parasite diversity and haplotype frequency in the study region has been estimated, the recrudescent group can be further divided into reinfections with a parasite having the same haplotype as the pre-treatment parasite (N) and true treatment failures (R) (Figure IFigure I). This is a statistical adjustment based on the probability of being reinfected with a parasite having the same haplotype (P; see Box 1) and the number of paired pre- and post-treatment samples (n); N = P × n. Bounds on the estimated value of S can be obtained using the binomial distribution (e.g. 90% bounds correspond to the 5th and 95th percentiles of the binomial distribution). The outcome of statistical adjustment combined with PCR-genotyping as compared to PCR-genotyping alone is demonstrated in Table I using data from one field site. The example demonstrates that even in situations where there is a reasonable chance of the post-treatment parasite having the same haplotype as the pre-treatment parasite (e.g. genotyping of only one marker), the estimated treatment failure rate after statistical adjustment is similar to that obtained using several markers. This is in contrast to the results based on genotyping alone, where the estimated number of treatment failures is directly related to the number of genetic markers used.
Figure I
Figure I
Figure I
Segregation of clinical samples into new infections and treatment failures.
Table I
Combining genotyping with statistical adjustment to estimate treatment failure
Genotyping
marker(s)a		nPbDSEstimated
treatment failure rate
(S/n*100)		Statistical adjustment of genotyping data (90% bounds)
NRcAdjusted estimate
of treatment failure
rate (R/n*100)
Pfmsp2610.15243760.79.3 (5–14)27.7 (23–32)45.5 (37.7–52.5)
Pfmsp1 + Pfmsp2650.06323350.83.9 (1–7)29.1 (26–32)44.8 (40.0–49.2)
Pfmsp1 + Pfmsp2 + glurp650.02362944.61.6 (0–4)27.4 (25–29)42.2 (38.5–44.6)
aGenotyping data for one, two or three genetic markers from the Thailand study [19]. Four samples were excluded from the analysis as they had both new and recrudescent genotypes. Abbreviations: n, number of paired pre- and post-treatment samples with results for at least one of the markers genotyped; D, number of paired samples where the pre- and post-treatment samples have different haplotypes (represents new infections by different parasite); S, number of paired samples where the post-treatment sample has the same haplotype as the pre-treatment sample; N, number of paired samples having the same pre- and post-treatment haplotype that have been calculated to be reinfections with a parasite having the same haplotype; R, number of paired samples having the same pre- and post-treatment haplotype that are calculated to be recrudescences (true treatment failures).
bAs reported in Figure 2aFigure 2.
cValues calculated as (S-N).
The issues of parasite diversity and dominance are essential to the accurate determination of treatment failure rates. In light of the new WHO recommendations to change antimalarial treatment policy when the failure proportion exceeds 10% [1], accurate PCR correction of failure data is imperative, because a decision to change first-line treatment might have significant impacts on patient welfare and government health budgets. To achieve this, field studies and trials need to quantify the probability of reinfection based on the parasite diversity within the study area and statistically adjust efficacy rates, if required. Failure to adjust estimated therapeutic efficacy rates might overestimate the occurrence of treatment failure, potentially leading to incorrect conclusions. The molecular markers used to differentiate parasites as new or recrudescent are irrelevant so long as they possess adequate diversity and a relatively uniform distribution. These properties will vary between study sites. Therefore, determination of parasite diversity and allele frequency should form an integral part of all studies investigating the therapeutic efficacy of antimalarial treatment regimes.
Acknowledgments
We thank the participants of the Rieckmann Symposium held at the Menzies Research Institute in August 2005 for their helpful discussions on the topic of PCR-correction of efficacy data. M.L.G. was supported by NIH grant AI-47500–06. The opinions expressed herein are those of the authors and do not necessarily reflect those of the Defence Health Services or any extant policy of the Department of Defence, Australia.
Glossary
Haplotypea combination of alleles at multiple genetic loci
PCR correctionthe use of PCR to compare parasite genotypes and adjust drug efficacy results when the post-treatment parasite represents a new infection
Recrudescencethe recurrence of asexual parasitemia after treatment of the infection with the same infection that caused the original illness

Footnotes
Publisher's Disclaimer: This article was published in an Elsevier journal. The attached copy is furnished to the author for non-commercial research and education use, including for instruction at the author’s institution, sharing with colleagues and providing to institution administration.
Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited.
In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
References
1. WHO/HTM/MAL/2006.1108. World Health Organization Guidelines for the treatment of malaria. World Health Organization; 2006. (available at http://www.who.int/malaria/includes_en/whomalariapublications19982004.htm).
2. Snounou G, Beck HP. The use of PCR genotyping in the assessment of recrudescence or reinfection after antimalarial drug treatment. Parasitol Today. 1998;14:462–467. [PubMed]
3. Cattamanchi A, et al. Distinguishing recrudescence from reinfection in a longitudinal antimalarial drug efficacy study: comparison of results based on genotyping of msp-1, msp-2, and glurp. Am J Trop Med Hyg. 2003;68:133–139. [PubMed]
4. Slater M, et al. Distinguishing recrudescences from new infections in antimalaria clinical trials: major impact of interpretation of genotyping results on estimates of drug efficacy. Am J Trop Med Hyg. 2005;73:256–262. [PubMed]
5. Chen N, et al. P. vivax relapses result from clonal hypnozoites activated at predetermined intervals. J Infect Dis. 2007;195:934–941. [PubMed]
6. Imwong M, et al. Plasmodium vivax relapses usually result from activation of heterologous hypnozoites. J Infect Dis. 2007;195:927–933. [PubMed]
7. Ranford-Cartwright LC, et al. Molecular analysis of recrudescent parasites in a Plasmodium falciparum drug efficacy trial in Gabon. Trans R Soc Trop Med Hyg. 1997;91:719–724. [PubMed]
8. Konate L, et al. Variation of Plasmodium falciparum msp1 block 2 and msp2 allele prevalence and of infection complexity in two neighbouring Senegalese villages with different transmission conditions. Trans R Soc Trop Med Hyg. 1999;93(Suppl 1):21–28. [PubMed]
9. Imwong M, et al. Practical PCR genotyping protocols for Plasmodium vivax using Pvcs and Pvmsp1. Malar J. 2005;4:20. [PubMed]
10. Zakeri S, et al. Circumsporozoite protein gene diversity among temperate and tropical Plasmodium vivax isolates from Iran. Trop Med Int Health. 2006;11:729–737. [PubMed]
11. Kim JR, et al. Genetic diversity of Plasmodium vivax in Kolkata, India. Malar J. 2006;5:71. [PubMed]
12. Bonilla JA, et al. Genetic diversity of Plasmodium vivax Pvcsp and Pvmsp1 in Guyana, South America. Am J Trop Med Hyg. 2006;75:830–835. [PubMed]
13. Montoya L, et al. Plasmodium falciparum: diversity studies of isolates from two Colombian regions with different endemicity. Exp Parasitol. 2003;104:14–19. [PubMed]
14. Nyachieo A, et al. Plasmodium falciparum genotyping by microsatellites as a method to distinguish between recrudescent and new infections. Am J Trop Med Hyg. 2005;73:210–213. [PubMed]
15. Anderson TJ, et al. Microsatellite markers reveal a spectrum of population structures in the malaria parasite Plasmodium falciparum. Mol Biol Evol. 2000;17:1467–1482. [PubMed]
16. Alam MT, et al. Extensive heterozygosity at four microsatellite loci flanking Plasmodium vivax dihydrofolate reductase gene. Mol Biochem Parasitol. 2007;153:178–185. [PubMed]
17. Su X, Wellems TE. Toward a high-resolution Plasmodium falciparum linkage map: polymorphic markers from hundreds of simple sequence repeats. Genomics. 1996;33:430–444. [PubMed]
18. Tami A, et al. Restricted genetic and antigenic diversity of Plasmodium falciparum under mesoendemic transmission in the Venezuelan Amazon. Parasitology. 2002;124:569–581. [PubMed]
19. Mugittu K, et al. Molecular genotyping to distinguish between recrudescents and new infections in treatment trials of Plasmodium falciparum malaria conducted in sub-Saharan Africa: adjustment of parasitological outcomes and assessment of genotyping effectiveness. Trop Med Int Health. 2006;11:1350–1359. [PubMed]
20. Mueller I, et al. The population structure of Plasmodium falciparum and Plasmodium vivax during an epidemic of malaria in the Eastern Highlands of Papua New Guinea. Am J Trop Med Hyg. 2002;67:459–464. [PubMed]
21. Arez AP, et al. A clonal Plasmodium falciparum population in an isolated outbreak of malaria in the Republic of Cabo Verde. Parasitology. 1999;118:347–355. [PubMed]
22. Zakeri S, et al. Population structure analysis of Plasmodium vivax in areas of Iran with different malaria endemicity. Am J Trop Med Hyg. 2006;74:394–400. [PubMed]
23. Brockman A, et al. Application of genetic markers to the identification of recrudescent Plasmodium falciparum infections on the northwestern border of Thailand. Am J Trop Med Hyg. 1999;60:14–21. [PubMed]

PubMed articles by these authors
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph