Performance of a HRP-2/pLDH based rapid diagnostic test at the Bangladesh-India-Myanmar border areas for diagnosis of clinical malaria

Malaria is often lethal with high potential expenditure for health if diagnosis is inaccurate [1]. Accurate diagnosis of malaria is of increasing importance as the prevalence of malaria is declining around the globe, making surveillance and screening more important for programme management [2, 3] and to restrict the use of anti-malarial drugs to restrain the spread of drug resistance [4].

For decades, expert microscopy (EM) of peripheral thick and thin blood smears has been the standard diagnostic test for malaria, however, it is time consuming and requires substantial expertise [1, 5]. Enzyme-linked immunosorbent assay (ELISA) and polymerase chain reaction (PCR)-based diagnostic tests have been introduced which ameliorate sensitivity and specificity of malaria diagnosis, but only in reference settings where well equipped laboratory facilities are available, making it difficult to implement in a field setting [6]. Other nascent molecular methods, such as loop-mediated isothermal amplification (LAMP) [7‐9] and real-time quantitative nucleic acid sequence-based amplification (QT-NASBA) [10] are in use, but the efficacy of each is unproven.

After being introduced in the early 1990s, rapid diagnostic tests (RDTs) have become an attractive alternative to the above-mentioned methods in a resource-limited setting for malaria diagnostics. The antigen-based RDTs detect specific antigens produced by malaria parasites by reaction with bound antibodies on an absorbent nitrocellulose membrane. Among several types of RDTs the two-band tests and three-band tests are most widely used. Two-band tests either detect only one species (Plasmodium falciparum), usually by detecting histidine-rich protein 2 (HRP2), or detect any of the four most common malaria parasites (P. falciparum, Plasmodium vivax, Plasmodium malariae and Plasmodium ovale), typically by detecting pan-Plasmodium-specific lactate dehydrogenase (pLDH), while three-band tests detect both the P. falciparum-specific antigen HRP2 and the pLDH or any one species specific LDH (mostly P. vivax). The third band is the test control band [1, 5, 11].

Southeast Bangladesh, northeast India and southwest Myanmar are similar in geographical characteristics and endemic for malaria. Plasmodium falciparum is the most abundant parasite, followed by P. vivax in these countries [6, 11, 12]. The presence of P. malariae and P. ovale has also been reported in each country [13‐16]. These three countries share their borders, making trans-border malaria transmission plausible. The presence of all four parasites in these mostly remote and resource-limited areas illustrate the importance of a RDT that can detect all malaria parasites. Amongst the locally available RDTs, Parascreen® Pan/Pf Rapid test for malaria (Zephyr Biomedical Systems, India), hereafter noted as Parascreen, is a RDT that has the capability to detect all types of human malaria, as it detects P. falciparum- specific HRP-2 and pan-Plasmodium-specific LDH. It has been evaluated against microscopy and conventional PCR in field and laboratory settings [12, 17‐24]. Here, the assessment of Parascreen in a laboratory setting and its performance compared with EM and qPCR are described.

In total, 327 febrile patients were included in this study from two UHCs. The results of EM, qPCR and Parascreen are provided in Table 1. With EM, there were 207 (63.3%) positive malaria cases, of which 153 (73.9%) were P. falciparum infection and 54 (26.0%) were P. vivax infection. The parasite density for P. falciparum ranged between 16 and 261,480 parasites/μL (IQR: 7,500-50,100) with median value of 19,960 parasites/μL, while the parasite density for P. vivax ranged between 16 and 25,120 parasites/μL (IQR: 320–4,800) with median value of 1,140 parasites/μL. qPCR confirmed 208 (63.6%) positive malaria cases, of which 154 (74.0%) were P. falciparum and 54 (25.9%) were P. vivax. With Parascreen, there were 202 (61.7%) malaria positive cases, of which 150 (74.2%) were P. falciparum and 52 (25.7%) were P. vivax infection.

Table 2 represents the calculated indicators when Parascreen was compared with EM and qPCR. EM being the reference standard, Parascreen had the following results, for any kind of malaria detection, sensitivity: 97.1% (95% CI, 93.5-98.8) and specificity: 99.1% (95% CI, 96.8-99.9); for P. falciparum malaria detection, sensitivity: 96.0% (95% CI, 91.2-98.3) and specificity: 98.2% (95% CI, 94.6-99.5) and for P. vivax malaria detection, sensitivity: 90.7% (95% CI, 78.9-96.5) and specificity: 98.9% (95% CI, 96.5-99.7). When qPCR was used as the reference standard, Parascreen had the following results for any kind of malaria detection, sensitivity: 97.1% (95% CI, 93.5-98.8) and specificity: 100% (96.1-100.0); for P. falciparum malaria detection, sensitivity: 95.4% (95% CI, 90.5-98.0) and specificity: 98.8% (95% CI, 95.4-99.7) and for P. vivax malaria detection, sensitivity: 89.0% (95% CI, 77.0-95.4) and specificity: 98.8% (95% CI, 96.5-99.7).

Parascreen showed higher sensitivity (93.3-100%) in detecting samples with parasite densities >500 parasites/μL for both P. falciparum and P. vivax, whereas for parasite densities ranging from 1–500 parasites/μL, the sensitivity was low (60.0%-83.3%) (Figure 1).

Parascreen showed acceptable performance in this study with overall sensitivity and specificity of 97.1 and 99.1%, respectively, when compared to EM, and 97.1 and 100%, respectively, in comparison with qPCR. Parascreen can detect all types of non-falciparum malaria but in this study only P. vivax was considered as P. malariae and P. ovale cases were not present in the study samples [13, 14]. Parascreen demonstrated varying sensitivity and specificity when compared with EM and qPCR depending on parasite species (P. falciparum and P. vivax) and parasitaemia of infections.

Several evaluation studies of Parascreen in different countries reported overall sensitivity ranging from 47.5 to 95.5% and specificity from 64.3 to 98.5% with varying performance for falciparum and non-falciparum malaria detection [12, 17‐21, 23, 24]. Parascreen has been evaluated by WHO RDT evaluation programme and for P. falciparum detection it showed almost 100% detection rates while for P. vivax it was approximately 30% [27]. Here, in this study, for P. falciparum detection, the sensitivity and specificity was also in concordance with the previous findings [12, 20, 21], while for P. vivax detection, improved sensitivity and specificity are reported. The improved sensitivity and specificity of P. vivax detection compared to previous findings may be due to the increased release of antigen through parasite lysis in the archived sample [1] or due to the improvement in the product. This finding is also corroborated by a meta analysis where mean sensitivity and specificity of 95.0 and 95.2%, respectively, for HRP-2 based assays and 93.2 and 98.5%, respectively, for pLDH based assays were calculated [28].

In a study in India, Parascreen showed 94.0% sensitivity and 72.0% specificity for P. falciparum and for P. vivax 77.2% sensitivity and 98.1% specificity were recorded when compared with EM and similar values observed when compared with PCR [12].

In Myanmar, two RDTs with similar detection properties have been evaluated in field settings [11, 26]. The SD 05FK60 RDT evaluated in the Rakhaine state of Myanmar showed 90.2% sensitivity and 98.5% specificity for P. falciparum and 79.4% sensitivity and 98.7% specificity for non-falciparum malaria [11]. The VIKIA Malaria Ag Pf/Pan™ test showed 98.0 and 100% sensitivity for P. falciparum and non-falciparum malaria, respectively, with specificity of 98.0 and 100%, respectively [26].

Onsite (Pf/Pan), a RDT with similar detection principle recently evaluated in Bangladesh, reported 94.2% sensitivity and 99.5% specificity for falciparum malaria detection and for vivax malaria detection it showed sensitivity and specificity of 97.3 and 98.7%, respectively [5] which showed slightly better sensitivity and specificity compared to Parascreen.

WHO recommends sensitivity ≥95% at ≥100 parasites/uL for RDTs [1]. In this study, for both falciparum and vivax malaria detection, sensitivity was less than the recommended values for low parasitaemia; however, considering fewer low parasitaemia samples, statistically valid conclusions have not been attained.

In this study, Parascreen was unable to detect three microscopically confirmed falciparum malaria samples with parasitaemia ranging from 112 to 2,600 parasites/uL. This might be caused by the degradation of HRP-2 target antigen as the study was carried out with archived samples. Intraspecies sequence variation [29], deletions or mutations of HRP-2 gene [30, 31] among different P. falciparum isolates could also account for false negative tests. The extent of HRP-2 variations in Bangladesh is currently unknown, however variations or deletions in HRP-2 have been reported recently from India, [31] as well as some African countries [30, 32]. In this study, three P. falciparum samples showed no HRP-2 test line but Pan specific test line, thus considered as P. vivax, as other types of malaria were absent in the study samples. The intraspecies variation, mutation or deletions in the HRP-2 gene can cause non-expression of HRP-2 [30] which may explain this. Parascreen identified three P. vivax-positive samples with parasitaemia ranging from 16 to 200 parasites/uL as negative. This might be due to low pLDH level, as pLDH level is directly proportional to parasitaemia [33]. In many studies, a reduced sensitivity for non-falciparum malaria detection, compared to falciparum detection, in combined HRP-2/pLDH RDTs has also been reported [5, 6, 11, 26].

As all four malarial parasites co-exist in the Bangladesh-India-Myanmar border area, an important criterion for selection of an appropriate RDT is the capability to detect all types of malaria. It is advantageous to use Pf/Pan RDTs which can do so. The high predictive values for Parascreen indicate that it is able to detect true malaria cases as well as ruling out non-malaria cases. High sensitivity, specificity and predictive values for Parascreen present it as a viable alternative for malaria diagnostics in Bangladesh-India-Myanmar border areas where malaria is endemic.

The absence of P. malariae and P. ovale samples in this study restricts the findings to the detection performance of falcipaum and vivax malaria. The inclusion of P. malariae and P. ovale in the study samples is needed to assess non-falciparum malaria detection performance.

Parascreen showed acceptable performance for falciparum as well as vivax malaria diagnosis in standard experimental conditions. It can be employed in resource-limited settings to diagnose all types of human malaria.