Sequence variation does not confound the measurement of plasma PfHRP2 concentration in African children presenting with severe malaria

The Plasmodium falciparum histidine-rich proteins PFHRP2 and PFHRP3[1] are soluble proteins containing highly antigenic tandem repeat sequences[2] and are produced in large quantities during the asexual blood stage of the P. falciparum lifecycle[3] where their concentration represents a dynamic equilibrium[4]. Orthologs are not found in other species of human malaria[5]. Antibody detection of PFHRP2 (which also picks up PFHRP3) is, therefore, a sensitive method for the rapid diagnosis of P. falciparum malaria using whole blood[6], as well as an in vitro drug sensitivity readout that has found increasing use in recent years[7]. PFHRP2 is also released into plasma, where its concentration reflects the extent of current and previous cycle parasite sequestration, a process considered critical to the pathology of severe malaria[8].

Since the initial description of these histidine-rich proteins more than two decades ago[1, 9], their physiological role in the lifecycle of P. falciparum has been subject to a number of lines of investigation, with the focus of most attention on haem polymerization[10]. A degree of redundancy exists with parasites lacking one or both of these proteins still able to propagate in asexual stage culture and complete the entire parasite life cycle[11‐13]. Nevertheless, classical falciparum genetic crosses in which one parent had a deletion of Pfhrp3 demonstrated complete[14] or partial[15] bias towards Pfhrp3 inheritance (although the cross-over interval was not fine enough to attribute enhanced fitness specifically to Pfhrp3 as opposed to nearby genes). Perhaps surprisingly, Pfhrp2 did not show inheritance bias in the HB3 x Dd2 cross[15]; low-level transcription in the HB3 parent[16] could potentially explain this in part or in whole.

In contrast to these studies on cultured isolates, prior to 2010 no field isolates had been reported which lacked either Pfhrp2 or Pfhrp3. However, relatively recent evidence of significant numbers of parasites with deletion of one or both genes has emerged in the Peruvian Amazon[17, 18], generating concerns over the reliability of PFHRP2-based rapid diagnostic tests (RDTs) in this region. There has been little evidence that deletion of either gene occurs to a significant degree in Africa. This may reflect both the true rarity of the event (given the predicted deleterious effect on fitness), along with the higher multiplicities of infection found in high transmission settings. It is therefore concerning that isolates from Mali lacking Pfhrp2 have now been described[19].

Both Pfhrp2 and Pfhrp3 also exhibit extensive insertion-deletion polymorphism, and the potential for this to reduce the accuracy of RDTs has been examined at national[20] and global scales[21, 22]. However, whether sequence variation is an important confounder in the assessment of plasma concentration, and hence disease severity, has not been studied. This study examined the relationship between Pfhrp2 sequence and plasma PFHRP2 concentration in samples derived from a large group of African children treated for severe malaria in the AQUAMAT trial[23].

A PCR-based approach was used to answer a number of related questions. First, Pfhrp2 deletions were sought in parasites from patients admitted with a diagnosis of severe malaria but who had very low or undetectable plasma PFHRP2 levels by ELISA. Deletions in Pfhrp3 were also checked since by cross-reacting with PFHRP2, the less abundant PFHRP3 allows RDTs to retain some sensitivity even when Pfhrp2 deletions are present. Finally, the relationship between Pfhrp2 exon 2 length and repeat structure and plasma concentration was explored.

PFHRP2 is released into the circulating plasma[8, 26‐28] where it persists for several days[29]. Because of the exponential rise in parasitaemia that occurs during the development of P. falciparum in the blood, in whole blood a large portion of PFHRP2 is located within asexual parasites[27], and whole-blood PFHRP2 concentrations correlate, therefore, with peripheral blood parasitaemia. By contrast, plasma PFHRP2 provides a summative picture of previous cycles of growth that reflects sequestered parasite biomass rather than circulating biomass[8]. Plasma PFHRP2 measurement can therefore potentially be applied in the management of severely ill patients in high transmission settings who may be ill because of malaria or alternatively have another severe illness, with malaria parasitaemia an incidental finding[30]. Post-mortem reveals that 25% of patients thought to have died of severe malaria have other serious pathology[31]. Confident differentiation of severe malaria from other pathologies where the prevalence of malaria parasitaemia is high is very difficult[30]; observation of malarial retinopathy improves the accuracy of diagnosis for cerebral malaria[32], but requires skill and training and is rarely achievable in practice outside centres of excellence. Since plasma PFHRP2 derives from sequestered parasites that are critical to the pathology of severe malaria, it has potential to differentiate ‘true’ severe malaria from alternative illnesses in parasitaemic patients, as well as provide an independent marker of prognostic significance in severe malaria[8]. Plasma PFHRP2 levels were found to be higher in patients with severe falciparum malaria compared to uncomplicated malaria in Indonesia[33] and Thailand[8] although studies in Kenya[34] and Papua New Guinea[35] did not find a significant difference. Explanations for these apparently discrepant findings include different classifications of disease severity as well as differences in the intensity of transmission. A particular problem in these studies, and in other studies of severe malaria is the classification of children with severe anaemia and low parasitaemia as having “severe malaria”, many of whom have chronic anaemia and incidental parasitaemia. These patients do not have large numbers of sequestered parasites and so do not have high plasma PFHRP2 concentrations.

The antibodies present in PFHRP2 RDT kits detect the repetitive histidine-rich repeats encoded by exon 2 of Pfhrp2 (and Pfhrp3)[2, 36]. This is a strength of PFHRP2-based diagnosis, since the antigen sequence appears many times within the protein and one protein can be bound by several antibody molecules potentially leading to amplification of signal and high assay sensitivity. However, the nature of the antigen detected optimally by each RDT clearly differs between kits[2, 36]. The possibility that this might impair the accuracy of RDTs was reflected for a time by a concern that certain patterns of repeat polymorphism (relating to numbers of type 2 and type 7 repeats) were associated with lower levels of detection, based on a semi-quantitative analysis of kit sensitivity[21]. Subsequent work has not supported this possibility and the performance of a range of RDTs now appears unlikely to be confounded by sequence polymorphisms[22].

This study aimed to address the concern that quantification of plasma PFHRP2 might be confounded by sequence polymorphism in terms of type and number of histidine-containing repeats in PFHRP2. Among parasites from Mozambique and Tanzania, a diverse pattern of repeat types was observed consistent with previous studies from a wide range of African countries[20‐22]. Within this variation, there was no relationship between the number of repeats observed, or overall sequence length, and plasma PFHRP2 level, mirroring the semi-quantitative findings from studies examining rapid diagnostic tests[22]. These findings do not support the idea that sequence polymorphism has a substantial influence on ELISA-based PFHRP2 quantitation in African samples. The results presented here, along with existing data that in Southeast Asian adult plasma PFHRP2 correlates well with severity of disease[8, 33], provide support for further work aimed at developing measurement of plasma PFHRP2 as an aid to the diagnosis and the management of severely ill African children[25].

Previous reports of field isolates with deletion of one or both of these genes examined samples originating in areas with generally low multiplicity of infection (MOI)[17, 18], or focused on samples from individuals with no or few symptoms of malaria and hence likely to have lower MOI[19]. In this survey of around 80 selected parasite samples from patients admitted with a diagnosis of severe malaria but with very low PFHRP2 plasma levels across a number of African countries, there was no evidence for deletion of either Pfhrp2 or Pfhrp3. Multiplicity of infection was not measured, although it is likely that a significant proportion of these samples contained single infections, since around 10% contained insufficient DNA to produce a PCR product in two independent reactions. It is also worth noting that the median asexual parasitaemia in this subgroup was about 9,000 parasites per μl, a figure similar to that associated with false-negative RDTs in the recently published study from Mali. The findings therefore provide a degree of reassurance regarding the prevalence of Pfhrp2 and Pfhrp3 deletion across the sites examined.

PFHRP2 sequence variation in parasites causing severe malaria in Africa does not confound the quantification of PFHRP2 antigen in plasma. This is a key element in the development of plasma PFHRP2 concentration measurement as a method for assessing African children who may have severe falciparum malaria. The evidence supports the concept that sequestered parasite biomass is the major cause of variation of PFHRP2 concentration in patient samples.