Impact of Plasmodium falciparum gene deletions on malaria rapid diagnostic test performance

Antigen-detecting rapid diagnostic tests (RDTs) are recommended diagnostic tools by the World Health Organization (WHO) for malaria case management [1]. The implementation of malaria RDTs has greatly improved access to diagnosis in endemic countries, particularly in Africa [2].

In general, three types of RDTs for detection of Plasmodium falciparum are commercially available: 1) P. falciparum-only RDTs; 2) combination RDTs, which detect and differentiate P. falciparum and some, or all, non-P. falciparum species; and, 3) Pan-only RDTs, which detect but do not differentiate between P. falciparum and non-P. falciparum species. Most P. falciparum-detecting RDTs use histidine-rich protein 2 (HRP2) as it is species-specific and abdundantly produced. Some HRP2-based RDTs may also potentially detect P. falciparum histidine-rich protein 3 (HRP3) due to its structural similarity with HRP2 [3]. RDT bands detecting non-falciparum Plasmodium target Pan or species-specific lactate dehydrogenase (Pan-LDH, Plasmodium vivax (Pv)-LDH or P. vivax, Plasmodium ovale, Plasmodium malariae (Pvom)-LDH), or aldolase. P. falciparum-LDH (Pf-LDH) can also be used specifically to detect P. falciparum. HRP2-based RDTs generally exhibit superior performance, particularly at low parasite densities, and are more heat stable than non-HRP2-based RDTs [4].

The sensitivity of HRP2-based RDTs is seriously threatened by the increasing occurrence of P. falciparum with deleted HRP2 and/or HRP3 antigen-coding genes. Plasmodium falciparum isolates lacking pfhrp2/3 were first reported in Peru with a prevalence of 20 to 90%, depending on location [5‐7]. Subsequently, pfhrp2/3-deleted parasites have been reported in Colombia, Suriname, Bolivia, Brazil, Honduras, Guatemala, and Nicaragua [8‐12]. Parasites lacking pfhrp2 have also been reported around the China-Myanmar border [13] and in India [14] with prevalence up to 25% in some areas [15].

In Africa, parasites lacking one or both pfhrp2 and pfhrp3 have been reported in Mali [16], Ghana [17], Senegal [18], Democratic Republic of the Congo [19], Rwanda [20], Zambia [21], and Kenya [22] with prevalences ranging between 2 and 45%, while up to 80% of symptomatic patients at two regional hospitals in Eritrea had P. falciparum lacking pfhrp2/3 [23].

The emergence of parasites that do not express HRP2 poses a major public health threat due to the heavy reliance of RDTs on this antigen. In response, WHO has released a Response Plan [24]. If HRP2-based P. falciparum-only RDTs are used when a patient is infected solely with parasites lacking HRP2 then a false-negative diagnosis can occur, delaying correct treatment and potentially leading to severe complications and death. In regions where HRP2-pan-LDH combination tests are used, incorrect diagnosis of non-falciparum malaria can occur when individuals are infected with HRP2-negative P. falciparum, potentially impacting the treatment regimen and patient health outcome. In both situations, routine surveillance estimates of malaria incidence will be adversely affected.

The current solution to the diagnostic problem posed by P. falciparum parasites lacking pfhrp2/3 is first to establish prevalence and, based on these results, decide if a replacement RDT or microsopy is needed. Any replacement RDT should not exclusively rely on HRP2 for P. falciparum. However, it can be challenging to maintain access to quality-assured microscopy and to switch RDTs, especially as the number of RDTs targeting alternative antigens is limited. Eighty-nine RDTs that detect P. falciparum, alone or in combination, have undergone WHO product testing within the past 5 years [4]. Of these, 78 use HRP2 exclusively to detect P. falciparum, with nine currently pre-qualified by WHO. The remaining 11 RDTs use Pf-LDH either alone or in combination with HRP2, or pan-LDH only, to detect P. falciparum, with only two of these products (one pan-LDH and one HRP2/Pf-LDH) being WHO pre-qualified [4]. During WHO product testing, the performance of Pf-LDH test bands has been generally poor with only two products having met the WHO performance criteria based on Pf-LDH test line results against wild-type, HRP2-expressing P. falciparum.

Although many studies have evaluated RDT performance in the field, and the WHO and Foundation for Innovative New Diagnostics, in collaboration with the US Centers for Disease Control and Prevention (CDC), have led rigorous performance testing of RDTs over the past decade, there has been no systematic assessment of the performance of RDTs against well-characterized pfhrp2-deleted P. falciparum parasites.

To address this gap, Round 8 WHO Malaria RDT Product Testing Programme included an evaluation of RDTs against a panel of pfhrp2-deleted P. falciparum parasites, with and without pfhrp3. Here, the findings are described and discussed with reference to the implications for future RDT use.

The recent emergence of pfhrp2/3-deleted parasites in several African and South American countries, as well as India, has rapidly escalated the need for RDTs that are not solely reliant on HRP2 for the detection of P. falciparum. Modelling studies have shown that use of RDTs reliant only on HRP2 detection can exert selective pressure on the parasite population to drive the spread of pfhrp2/3-deleted P. falciparum [27, 28]. The WHO recommends that countries do not exclusively rely on HRP2-based RDTs where the prevalence of pfhrp2 deletions causing false-negative RDTs is greater than 5% in symptomatic patients [29]. In many cases this would be operationalized by changing from a HRP2-detecting RDT to a pan-LDH and/or Pf-LDH-detecting RDT, with the assumption that these RDTs perform equally well on HPR2-negative and HRP2-positive parasites. However, this assumption had not been previously tested and the current results suggest that performance of Pf-LDH-detecting RDTs against wild-type samples do not predict performance against pfhrp2/3-deleted parasites (clinical and cultured samples).

There was large variability in the positivity of the nine Pf-LDH RDTs tested against samples equivalent to 200 parasites/µL, but as a group they unexpectedly appeared to detect wild-type P. falciparum at higher rates than pfhrp2-deleted parasites. Indeed, one combination Pf-LDH RDT assessed in this study met the WHO performance criteria against wild-type P. falciparum with a Panel Detection Score (PDS) of 83 (89% positivity), but obtained a PDS of 0 (12% positivity) when assessed against pfhrp2-deleted parasites [4]. Antigen concentration is a potential confounder in the comparison between performance against wild-type and pfhrp2-deleted parasites, so the pfhrp2-deleted panel was prepared to have a similar distribution of Pf-LDH concentration to the wild-type panel, with all ELISAs run in triplicate using the same controls for each panel. Indeed, if the decreased positivity were due to variation in Pf-LDH concentration, it would be expected that the pan-LDH RDTs would show comparable decreases in performance when challenged against the pfhrp2-deleted parasites, which was not the case. Therefore, it is unlikely that differences in antigen concentrations explain the observed results.

The products using Pf-LDH included both dual-band products, with separate test bands detecting Pf-LDH and HRP2, and single-band products, using either Pf-LDH alone or a combination of Pf-LDH and HRP2 on the same band. Interestingly, the reduced performance on the pfhrp2-deleted panel compared to wild-type panel appeared to be restricted to Pf-LDH detecting products that did not contain a separate HRP2 band. This may be a spurious result due to the small number of Pf-LDH-detecting RDTs examined, or the limited size and diversity of the pfhrp2-deleted panel or product specific issues, such as Pf-LDH test lines unexpectedly reacting with HRP2. Reassuringly, all Pf-LDH RDTs were able to detect a small set of double-deleted clinical isolates at the higher parasite density of 2,000 parasites/µL.

Although assessments against larger, more geographically diverse panels are needed, these results suggest that where good quality microscopy is not available and where the prevalence of pfhrp2/3-deleted parasites leading to false-negative RDT results is > 5% [29], the pan-LDH RDTs would be suitable for P. falciparum detection. The two pan-LDH-only products had the best performance against pfhrp2-deleted parasites and well exceeded the minimum WHO RDT performance criteria for P. falciparum, specifically > 75% panel detection score at 200 parasites/µL. However, neither of these products is yet WHO pre-qualified, nor are they in the assessment pipeline, and this may limit procurement by certain agencies [30]. The one pan-LDH-only RDT that is currently WHO pre-qualified has not been evaluated against a pfhrp2-deleted panel and these current results highlight the need for additional assessments.

On the other hand, in areas that require differentiation of P. falciparum from non-P. falciparum infection for treatment decision making, reporting or surveillance, current Pf-LDH-detecting products could have utility until better RDTs become available. The risk–benefit of presumptive treatment of fever versus false-negative Pf-LDH RDTs secondary to parasitaemia below 2,000 parasites/µL would need to be carefully considered.

Although the main focus of this study was to assess the performance of Pf-LDH-detecting RDTs, a variety of RDTs that only detect P. falciparum using HRP2 were also included. This provided the opportunity to review how these products respond with the a priori assumption that HRP2-detecting RDTs would not detect pfhrp2-deleted parasites, an assumption which was confirmed for parasites lacking both pfhrp2 and pfhrp3. The majority of field studies test for the presence of both pfhrp2 and pfhrp3 but to date very few have found pfhrp2-negative/pfhrp3-positive parasites [22]. However, in this study both double- and single-deleted parasites were included, since the structural similarity between HRP2 and HRP3 may provide the opportunity for cross-reactivity [3]. The results demonstrate that some, but not all, HRP2-detecting RDTs return positive results against single-deleted P. falciparum at concentrations equivalent to 200 parasites/µL. Hence, it appears that some products tested are able to detect HRP3, as previously reported, and even at lower concentrations [6, 22]. Therefore, the risk of incorrect diagnosis posed by single-deleted mutants is reduced when certain RDT brands are used. In this study, detection of pfhrp2-negative/pfhrp3-positive parasites for individual HRP2-based RDTs ranged from very limited (1.1%) to almost complete detection (92.0%). This large variation has implications for the detection and surveillance of HRP2-negative parasites, since in some cases only double-deleted parasites will present as RDT-negative, while in other cases both single- and double-deleted parasites will present as RDT-negative. This difference could affect the frequency of false-negative RDT results and also survey estimates of the prevalence of pfhrp2-deleted parasites in symptomatic patients, if HRP2-negative RDT results are used as a first screen for mutant parasites that require genotyping [31]. Therefore, buyers should consider these results when selecting an HRP2 RDT to purchase, as the cross-reactivity afforded by HRP3 will reduce the number of false-negative RDT results. For surveillance, estimation of the prevalence of pfhrp2-deleted parasites may require inclusion of a sub-set of HRP2-positive RDTs from malaria patients, as well as genotyping for both pfhrp2 and pfhrp3. Furthermore, it is not known if pfhrp2 single-deleted mutants are a harbinger for pfhrp3 deletions, and subsequently the double deletions that generate false-negative results on HRP2 test lines.

Combination RDTs that use HRP2 to detect P. falciparum and pan-LDH to detect Plasmodium spp are widely used in areas where P. falciparum and P. vivax co-exist. These tests potentially detect HRP2-negative parasites but are likely to misclassify the result as a non-falciparum infection. The results of this study demonstrate that there is large variability in the rate of this type of false positivity between products, a feature that is possibly related to the specific antibody used and dependent on whether the HRP2 band cross-reacts with HRP3, as well as the sensitivity of the pan-LDH test band. The variability in the positivity of the pan-LDH band noted in this study matches that previously reported [32], and suggests reliance on the pan-LDH band to detect HRP2-negative parasites in regions where P. falciparum dominates may be unreliable.

An important limitation of this study is the small panel size, particularly when separated into double- and single-deleted samples. Unfortunately, all the single-deleted samples were prepared from two culture lines, as no clinical samples were available. There appears to be no difference in RDT performance against clinical and culture samples in the double-deleted parasites used, suggesting that the use of culture-derived samples has not significantly impacted the results of the study.

The current study demonstrates that Pf-LDH detecting RDTs respond strongly to high-density P. falciparum samples lacking pfhrp2, but performance at lower densities is variable. It is recommended that further testing of Pf-LDH detecting RDTs be conducted against a larger and geographically diverse panel of HRP2-negative samples. Surprisingly, many HRP2-detecting RDTs were able to detect single-deleted parasites at low density, a likely positive outcome for clinical management of P. falciparum in areas where pfhrp2-negative/pfhrp3-positive parasites exist, but a potential source of discrepancy for reporting prevalence of HRP2-negative parasites. Ultimately, new targets for P. falciparum detection should be explored, especially since optimizing Pf-LDH RDTs has proven difficult for manufacturers, so clinicians and community health workers can have confidence in RDT results to make clear treatment decisions in all malaria endemic regions.