Critical Evaluation of Molecular Monitoring in Malaria Drug Efficacy Trials and Pitfalls of Length-Polymorphic Markers

Abstract

Estimation of drug efficacy in antimalarial drug trials requires parasite genotyping to distinguish new infections from treatment failures. When using length-polymorphic molecular markers, preferential amplification of short fragments can compromise detection of coinfections, potentially leading to misclassification of treatment outcome. We quantified minority clone detectability and competition among msp1, msp2, and glurp amplicons using mixtures of Plasmodium falciparum strains and investigated the impact of template competition on genotyping outcomes in 44 paired field samples. Substantial amplification bias was detected for all three markers, with shorter fragments outperforming larger fragments. The strongest template competition was observed for the marker glurp Detection of glurp fragments in multiclonal infections was severely compromised. Eight of 44 sample pairs were identified as new infections by all three markers. Ten pairs were defined as new infections based on one marker alone, seven of which were defined by the questionable marker glurp The impact of size-dependent template competition on genotyping outcomes therefore calls for necessary amendments to the current WHO recommendations for PCR correction of malaria drug trial endpoints. Accuracy of genotyping outcomes could be improved by separate amplification reactions per allelic family and basing results on markers msp1 and msp2 first, with glurp only used to resolve discordant results.

Keywords: PCR; Plasmodium falciparum; amplification bias; drug trial; genotyping; glurp; msp1; msp2; template competition.

FIG 1

Agarose gel of glurp nPCR products obtained from mixtures of P. falciparum in vitro culture strains HB3 and 3D7 in different ratios. (A) Strain 3D7 is predominant. (B) Strain HB3 is predominant. NC, negative control. Arrows indicate the sizes of PCR fragments as measured by CE.

FIG 2

Limit of detection of msp1 (A) and msp2 (B) minority clones in mixtures of P. falciparum culture strains. The limit of detection was determined by reciprocal serial dilution of the minority clone. Culture strains carried alleles either of the same allelic family (same tube amplification, left and middle panels) or of different families (amplification in separate tubes, right panels). Green square, allele detected in CE; gray square, allele not detected in CE. Fragments in bp reflect rounded mean allele sizes determined by capillary electrophoresis.

FIG 3

Electropherogram of msp1 and msp2 alleles from the same allelic family, amplified from culture strains mixed at a 1:1 ratio.

FIG 4

glurp minority clone detectability in mixtures of four culture strains. (A) Electropherograms of glurp alleles of four P. falciparum culture strains (HB3, FCB1, K1, and 3D7) mixed at a 1:1:1:1 ratio. (B) Proportion of glurp fluorescent signal detected during capillary electrophoresis for each clone in a four-culture strain mixture, with clone HB3 (shortest glurp allele, left panel) or clone 3D7 (longest glurp allele, right panel) as the increasingly dominant clone.

FIG 5

Patterns in msp1, msp2, and glurp MOI in 44 paired pre- and posttreatment samples. (A) Correlation of msp2/msp1, glurp/msp1, and glurp/msp2 MOI per sample. Correlations were fitted using 2nd-degree polynomials (thick line) and are shown with 95% confidence intervals (thin lines). (B) msp1, msp2, and glurp MOI per sample classified by genotyping outcome. NI, new infection; R, recrudescence. Mean MOI per class and standard deviation are shown.