Assessment of two malaria rapid diagnostic tests in children under five years of age, with follow-up of false-positive pLDH test results, in a hyperendemic falciparum malaria area, Sierra Leone

Prompt and accurate diagnosis followed by effective treatment is currently the main malaria control strategy besides preventive measures. Widespread documented resistance to the older common anti-malarial mono-therapies, notably chloroquine, has led to an increased use of the highly effective artemisinin-based combination therapy (ACT) as the first-line treatment for uncomplicated Plasmodium falciparum malaria [1, 2].

Accurate confirmation of malaria diagnosis can reduce the overuse of ACT treatment and, therefore, delay the development of resistance. Moreover, it may both reduce the risk of adverse drug reactions due to unnecessary treatment and increase correct treatment for pathologies other than malaria. This, in turn, increases patient and parents/caretakers confidence in health services by providing accurate information for a proper diagnosis [3, 4]. Currently, the most commonly accepted gold standard diagnostic method is microscopic reading of a stained blood film. This requires laboratory technicians with training and technical expertise, good quality reagents, a well-maintained microscope and is time consuming. From experience it is challenging to have microscopic results in less than two hours, as recommended by WHO, even with well-trained staff. In many populations where malaria is prevalent, no laboratory facilities and/or trained staff are present, making access to microscopically confirmed diagnosis impossible. An alternative to microscopy is to use a rapid diagnostic test (RDT). Malaria RDT technology is attractive to health care settings lacking in human resources especially facing a high caseload of patients with suspected malaria, often including a high proportion of children under five years of age. These tests can be used by personnel including laypersons, and not necessarily laboratory technicians, after limited training. Malaria RDT results are on average available in less than 30 minutes.

The two types of RDT most often used belong to two groups according to the detected antigens: those detecting histidine-rich protein II (HRP2) and those detecting parasite specific lactate dehydrogenase (pLDH).

HRP2 antigen is one of the three histidine-rich proteins produced solely by trophozoites and young gametocytes of Plasmodium falciparum. RDTs based on the detection of HRP2 can only diagnose Plasmodium falciparum infections and thus cannot be used for the detection of Plasmodium vivax or other human malarias (Plasmodium ovale, Plasmodium malariae, and Plasmodium knowlesi, the latter only recently be detected in Asia [5]).

The malaria RDT most commonly used in the field by Médecins Sans Frontières (MSF) at present is Paracheck-Pf^®, which belongs to the HRP2 group. Paracheck-Pf^® has widely proven its field reliability to detect Plasmodium falciparum in various continents, countries and settings [i.e. [6–8]]. However, a limitation of HRP2-based tests is their persistent false positivity after effective treatment of the infection. HRP2 is only slowly eliminated from the blood stream as it is expressed in the erythrocyte membrane [9]. This fact renders inconclusive a positive RDT result with a history of a recently treated infection, especially in areas of high transmission [i.e. [10–13]].

This is less of a problem with the pLDH- based tests. PLDH is an intracellular metabolic enzyme produced by all Plasmodium species that infect humans and is produced by sexual and asexual stages of the parasites. PLDH is cleared more quickly from the bloodstream than HRP2 after starting an effective treatment, as it is an enzyme only produced by viable parasites [i. e. [12, 14–16]]. However, production of pLDH from gametocytes after elimination of asexual stages means some tests will still stay falsely positive for several days [17]. As pLDH-based tests are able to detect all human related Plasmodium, these RDTs could reduce the chance of patients being treated unnecessarily for diseases other than malaria or leave non-falciparum malaria untreated as it would be the case with a HRP2-based test.

Unfortunately, to date, the sensitivity of these tests, under field conditions (mainly using the OptiMal-IT test), has often been reported as lower than their HRP2 counterparts and falling below the internationally recommended standard of 90% [8, 18, 19]. However, two recent studies have shown more promising results with sensitivities over 90% demonstrated with CareStart™ two-line PAN-pLDH and CareStart™ three-line P.f/PAN-pLDH tests [12, 20].

This study evaluated CareStart™ three-line P.f/PAN-pLDH test in comparison to Paracheck-Pf^®. CareStart™ three-line P.f/PAN-pLDH was chosen for three reasons.

First, CareStart™ three-line P.f/PAN-pLDH is able to detect falciparum malaria as well as non-falciparum malaria. In a region where falciparum malaria is predominant but other species are also present, this test may contribute to a more differentiated diagnosis of malaria.

Second, pLDH-based tests are cleared more rapidly from the blood [12, 14–16]. In regions with high malaria transmission, it is especially important that RDTs can be relied upon to test for new malaria infections shortly after effective treatment of the initial diagnosis.

Third, among pLDH-based tests CareStart™ recently showed its capacity to reach a sensitivity of over 90% [12, 20].

Paracheck-Pf^® was chosen for comparison since it is currently in use in the MSF project in Sierra Leone and is considered to have a high sensitivity for Plasmodium falciparum [6–8, 21]. Nevertheless, it has a relatively long false positivity of the test after effective treatment of the patient and only detects falciparum malaria [i.e. [11–21]].

There is an ongoing debate whether or not anti-malarial treatment should be given to children under five years of age only on the basis of clinical suspicion for malaria [4, 22–24]. There is a trend towards laboratory-confirmed, usually RDT based, diagnosis [3, 4]. However, due to the lack of evidence, opinions vary broadly. So far sensitivity and specificity studies for malaria RDTs, with one exception in the Democratic Republic of Congo [11], did not focus on children under five years of age. Operational research is urgently needed to provide evidence for the use of malaria RDTs in this age class.

The two main aims of the study were 1) to compare the sensitivity and specificity of CareStart™ three-line P.f/PAN-pLDH test and Paracheck-Pf^® to diagnose Plasmodium falciparum malaria in children under five years of age using Giemsa-stained blood smears as gold standard reference and 2) to assess the time required for false-positive CareStart™ three-line P.f/PAN-pLDH tests to become negative after successful treatment with ACT.

Secondary aims included assessing ease of use and inter-reader agreement of both malaria RDTs.

Both malaria RDTs, CareStart™ three-line P.f/PAN-pLDH and Paracheck-Pf^®, were highly sensitive for the detection of falciparum malaria monoinfections in children under five years of age. When a parasitaemia of more than 100 parasites per microlitre was considered, both tests were >95% sensitive for the detection of Plasmodium falciparum, meeting the criteria set out by WHO [19, 25]. A literature search produced only two other published studies assessing the sensitivity of CareStart™ RDTs to detect Plasmodium falciparum malaria in Africa. Both studies used the two-line test and did not exclusively focus on children under five years of age. They included a field study in Uganda, which found sensitivity of the CareStart™ two-line PAN-pLDH test to be 95.6% [12], and a study in Madagascar, where sensitivity of CareStart™ two-line PAN-pLDH was 97% [20]. In a survey carried out recently in Myanmar, CareStart™ three-line P.f/PAN-pLDH tests had a sensitivity of 94.7% [26]. The results of these studies confirm the high sensitivity of CareStart™ tests and make this pLDH-based test a strong candidate for consideration when selecting the most suitable RDT in projects in hyperendemic falciparum malaria areas.

Both malaria RDTs were very reliable, having less than one percent invalid tests. Inter-reader reliability was equally good, indicating that wherever the tests are used, readers will be able to give trustworthy results. Both tests were perceived by the study team as similar in terms of ease of use. Therefore, neither RDT was preferable based on these characteristics.

Heat-stability has been a major concern for pLDH tests, especially under field conditions [27]. Due to very basic field conditions no stability testing was carried out in this study. However in a study carried out in Myanmar CareStart™ three-line P.f/PAN-pLDH tests remained stable for up to 90 days at 35°C [26]. In the 2009 WHO's evaluation of RDTs performance, the heat stability of CareStart™ two-line PAN-pLDH remained stable for up to 60 days at 4°C, 35°C and 45°C [28]. Thus, stability of the CareStart™ three-line P.f/PAN-pLDH seems to be satisfactory.

The major advantage of pLDH-based tests over HRP2-based tests is their ability to rapidly become negative after effective treatment with an anti-malarial. This is in contrast to Paracheck-Pf^® that remains false-positive for up to four weeks after successful treatment. In the Democratic Republic of Congo, 73.5% of effectively treated children were still Paracheck-Pf^® false-positive at Day 35 [11]. In Uganda, 69.7% of successfully treated patients remained Paracheck-Pf^® false-positive at Day 14 [12]. In southern Vietnam, the Paracheck-Pf^® false positivity rate was around 30% at Day 28 of follow-up [21]. The first results seen with CareStart™ were very promising - in Uganda 14 days after treatment only 10% of the CareStart™ 2 line PAN-pLDH tests remained false-positive. In that study, the follow-up stopped at Day 14 [12]. This rapid return to negativity was further confirmed in the present study. CareStart™ 3 line P.f/PAN-pLDH became negative after a median of seven days. On the 14^th day of follow-up less than 6% of the study patients still tested false-positive, approximately half the percentage found false-positive during the Uganda study at the same follow-up day. At Day 28, less than one percent of the study patients were still false-positive. Especially in a hyperendemic setting, it is extremely important to have a short parasite antigen clearance time after parasites are removed. This characteristic allows health personnel to confidently interpret positive RDT results without concerns over possible false-positive results from previously treated infections.

At the time this study was planned (early 2008), WHO did not endorse any particular RDT, but listed products submitted to the WHO malaria RDT product testing programme [29]. By April 2009, the outcome of the first round of product testing malaria antigen-detecting RDTs was made public [28]. Both of the RDTs test manufacturers in this study were invited to submit up to three tests for evaluation. AccessBio submitted the CareStart™ two-line PAN-pLDH test while Orchid Biomedical System participated with Paracheck-Pf^®. All tests were evaluated using prepared blood panels of cultured Plasmodium falciparum parasites and patient-derived Plasmodium falciparum and Plasmodium vivax parasites, and a parasite-negative panel.

In the WHO evaluation, CareStart™ two-line PAN-pLDH performed better than Paracheck-Pf^® in terms of detection rate of parasites, inter-reader agreement, and heat stability. However, they both showed an equivalent high level of detection at a high parasite density (2,000 or 5,000 parasites/μL). It has to be noted that the WHO testing was an evaluation of the performance under laboratory conditions. In contrast, this study was carried out in the field and was the first to do so focusing on children under five years of age under these conditions. Future procurement decisions should be based on WHO test results, but need to take into consideration local conditions of malaria transmission and illness where the test will be used. Nevertheless, to centralise RDT product testing in vitro is a very welcome initiative from WHO to facilitate the evaluation of reliable RDTs in the field.

There are some potential limitations in generalizing this promising CareStart™ three-line P.f/PAN-pLDH test as a suitable candidate for malaria diagnosis. The study team was well-trained and constantly supervised in test preparation and interpretation. This might not be the case in all health facilities using this technology. Also, the study cohort only had Plasmodium falciparum malaria species to detect. CareStart™ three-line P.f/PAN-pLDH tests might perform less satisfactorily in settings where other malaria species predominate, which was the case in a study in Myanmar [26]. As the study participants all had a mono-infection with Plasmodium falciparum no conclusions can be drawn in terms of sensitivity of CareStart™ three-line P.f/PAN-pLDH test for the other Plasmodium infections. This confirms the need to carry out RDT assessments in different geographical areas and with different types and levels of transmission in order to find the most suitable test for each region.

Several studies demonstrate inconsistencies in blood smear reading and, therefore, possibly doubtful results for sensitivity and specificity of malaria RDTs [30, 31]. Some studies even suggest that RDTs perform better than microscopy under routine conditions [32] or at parasite densities below the threshold for detection by microscopy [33]. Because blood smears were used as the gold standard reference to compare both RDTs under evaluation, extra emphasis was placed on quality control of the blood smears. All slides were read twice, by two independent well-trained microscopists, who were blinded to both the RDTs results and the findings of their counterpart. In addition, all slides with disconcordant results were read a third time by the Department of Infectious and Tropical Diseases at the London School of Hygiene and Tropical Medicine in London.

Before choosing a malaria RDT for use in the field it must fulfil a number of requirements. In this study CareStart™ three-line P.f/PAN-pLDH fulfilled all the requested criteria: (1) Sensitivity for the diagnosis of falciparum malaria was more than 95% with a parasitaemia greater than 100/μl. (2) All further accuracy parameters (specificity, positive and negative predictive values) were above 90%. (3) The test had an excellent inter-reader agreement and was reliable, easy to use and interpret. (4) False positivity decreased rapidly after successful anti-malarial treatment and became, on average, negative after seven days.

This study confirms the value of RDTs for use in children under five years of age by refuting the idea that RDTs may miss a significant number of diagnoses in this age group. Widespread use of RDTs should improve diagnostic accuracy of malaria in the field and should be adopted by malaria control programmes. These results confirm that, in terms of accuracy, CareStart™ three-line P.f/PAN-pLDH test can be used as an alternative to Paracheck-Pf^® in hyperendemic settings where Plasmodium falciparum malaria is the predominant species.

On the basis of this study, policy makers and governmental and non-governmental health providers can consider CareStart™ three-line P.f/PAN-pLDH test as a valuable tool for diagnosing falciparum malaria in children under five years of age, especially in more remote areas with hyperendemic conditions. This supports the wider use of RDTs to accurately diagnose malaria and avoid over or under treatment.