Use of a highly-sensitive rapid diagnostic test to screen for malaria in pregnancy in Indonesia

Control programmes for malaria will need to account for the changing epidemiology of malaria as countries progress toward elimination. The current goal to eliminate malaria in 31 countries by 2030 will require sensitive diagnostic tools [1] to detect the reservoir of low-density and often sub-patent infections [2‐5]. Although the infectivity of these asymptomatic individuals is low, modelling shows that they contribute significantly to onward transmission because they make up the majority of the population in low-transmission areas [4, 6].

Diagnosis of malaria during pregnancy is complicated by the ability of certain Plasmodium falciparum parasites to sequester in the placenta, which can result in parasite densities in the peripheral blood below the level of detection of existing Rapid Diagnostic Tests (RDTs) and light microscopy [7‐11]. Diagnosis and treatment of these infections may improve pregnancy outcomes [12, 13] by preventing them from developing into patent infections and may control transmission by reducing the parasite reservoir [3, 8, 12, 13]. Lateral-flow rapid diagnostic tests detecting circulating malaria antigen are simple, widely used, and the cheapest point-of-care diagnostic tool for malaria. Many countries in the Asia–Pacific region use “Test and Treat” strategies at the first antenatal booking visit to screen all pregnant women for malaria regardless of symptoms [14‐16], or are exploring intermittent screening and treatment (IST) strategies to test women at each scheduled antenatal visit [13, 17]. However, the current generation of standard RDTs for malaria has a limit of detection of 100 parasites/µl and is not designed to detect low-density infections in asymptomatic pregnant women [2]. Low-cost and field-deployable highly sensitive rapid diagnostics may improve the detection of malaria in these populations. [2, 18].

Currently, only nucleic acid amplification tests (NAATs) are sufficiently sensitive to detect these low-density infections [2]. However, these methods are limited to well-equipped laboratory settings due to their inherent complexity and need for laboratory equipment.

Recently, the new Alere™ Ultra-sensitive Malaria Ag P. falciparum RDT (uRDT) was developed to bridge the gap between high-sensitivity and field-ready diagnostics [19]. To date, studies in Uganda, Ethiopia, Myanmar and Papua New Guinea have shown superior sensitivity of the uRDT in comparison to widely used SD Bioline RDTs in asymptomatic general population. An additional laboratory study reports a log-fold lower limit of detection (LOD) for the HRP2 antigen (80 pg/ml) [19]. The only study in pregnant women did not observe notable differences in sensitivity between the uRDT and existing SD Bioline RDTs. [20].

In total, 270 samples were screened. These included 158 P. falciparum positive samples (including non-falciparum Plasmodium co-infections) and 112 negative samples (by molecular composite). Only 23 positive samples were detected by either uRDT and csRDT (Fig. 1).

This study shows that in a moderate transmission setting in Indonesia, amongst asymptomatic pregnant women, the uRDT performed similarly to the csRDTs, which is a widely used test in the malaria in pregnancy control programme in Indonesia.

These results are supported by a study in Colombia, which also used stored samples from pregnant women. Although they note a trend of increased sensitivity by the uRDT, it was not found to be significantly different from the Standard Bioline RDT (sdRDT) [20]. A Tanzanian study in febrile children and adult outpatients also found no difference between uRDT and sdRDT performance, although febrile patients tend to have higher parasitaemia [29]. The results here differ from other studies conducted in non-pregnant populations that showed a significant increase in sensitivity of the uRDT compared to current RDTs. [19, 26, 30, 31]. The most apparent difference between the conflicting results is the target population (pregnant vs non-pregnant) and the storage conditions. It would be useful to screen pregnant populations in the same settings where differences in RDT performance were observed. One such study has commenced in Papua New Guinea (personal communications with Dr Leanne Robinson).

The uRDT is an antigen capture test specific to the P. falciparum HRP2 antigen, whereas the csRDT captures pLDH specific to P. vivax, P. ovale and P. malariae (indicated by the VOM window), in addition to PfHRP2 (Pf window). Ahmed et al. showed that amongst pregnant women in Indonesia, when using a different CareStart RDT that captures both pan-LDH and PfHRP2, reading the PfHRP2 window alone was less sensitive than using in combination with the pLDH results [18]. These findings and the low sensitivity of both RDTs demonstrated here, may point towards low HRP2 concentrations that fall below the limit of either RDT (uRDT LOD = 80 pg/ml, csRDT = 200 pg/ml). Low concentrations of HRP2 could be due to low parasitaemia, degradation of the protein during storage or deletions in the gene for this protein.

The non-linear relationship between HRP2 concentration/persistence and parasite density [19, 26], complicates analyses between RDT positivity and parasitaemia. Although no trend in sensitivity was observed with increasing Ct values (indicative of parasitaemia), however, most samples fell above Ct 30 indicating very low parasitaemias. Finally, HRP2 deletions have been reported in P. falciparum from several countries in South America, Africa and also India [32‐34], but not yet in SE. Asia. As yet, no full deletions (only polymorphisms [35]) of HRP2 have been reported in Indonesia, although this is an area for further consideration. HRP2 mutations have been implicated in the poor sensitivity of RDTs [33]. Early laboratory analyses of the uRDT using cultured parasites with HRP mutations demonstrated that the uRDT performance varies considerably depending on mutations of the HRP protein [19]. Consequently, investigations around HRP2 polymorphisms in this sample set are underway.

A limitation of this study is the use of different sample types used for the molecular and RDT testing. Molecular testing used stored DNA extracted from DBS, whereas RDTs were performed using pRBC and plasma samples reconstituted to an artificial HCT of 30%. Potential degradation of HRP2 in the stored samples could contribute to poor sensitivity of both RDTs since RDT testing was carried out 1 year after molecular testing, although positive samples remained positive over this time. The sensitivity of the molecular testing could also be enhanced through targeting higher copy genes [36] or RNA [37], or by using improved extraction methods [38]. The performance of both RDTs is anticipated to improve using fresh whole blood at point of contact, i.e. for the intended user scenario. However, the low sensitivity observed here is similar to reported sensitivities using a range of alternative RDTs in pregnant women in Indonesia. [18] The important finding here is that the uRDT performed similarly to the csRDT.

The diagnostic performance of the uRDT was assessed in a diagnostically challenging population: asymptomatic, pregnant women in a moderate transmission setting in Indonesia. This is the first study testing the uRDT in pregnant women in Asia and will build evidence to guide policy around the implementation of this test in these populations. In comparison to the composite molecular reference tests, both the uRDT and csRDT showed low sensitivity. Given the negative outcomes associated with malaria in pregnancy, it is crucial to detect and treat these infections rapidly. Further work is needed to assess the diagnostic performance of the uRDT in pregnant women before consideration of this test for implementation where csRDT are already being used- particularly in asymptomatic pregnant women in moderate transmission settings.