Low prevalence of Plasmodium malariae and Plasmodium ovale mono-infections among children in the Democratic Republic of the Congo: a population-based, cross-sectional study

In an effort to improve surveillance for epidemiological and clinical outcomes, rapid diagnostic tests (RDTs) have become increasingly widespread as cost-effective and field-ready methods of malaria diagnosis. However, there are concerns that using RDTs specific to Plasmodium falciparum may lead to missed detection of other malaria species such as Plasmodium malariae and Plasmodium ovale. Four hundred and sixty six samples were selected from children under 5 years old in the Democratic Republic of the Congo (DRC) who took part in a Demographic and Health Survey (DHS) in 2013–14. These samples were first tested for all Plasmodium species using an 18S ribosomal RNA-targeted real-time PCR; malaria-positive samples were then tested for P. falciparum, P. malariae and P.ovale using a highly sensitive nested PCR. The prevalence of P. falciparum, P. malariae and P. ovale were 46.6, 12.9 and 8.3 %, respectively. Most P. malariae and P. ovale infections were co-infected with P. falciparum—the prevalence of mono-infections of these species were only 1.0 and 0.6 %, respectively. Six out of these eight mono-infections were negative by RDT. The prevalence of P. falciparum by the more sensitive nested PCR was higher than that found previously by real-time PCR. Plasmodium malariae and P. ovale remain endemic at a low rate in the DRC, but the risk of missing malarial infections of these species due to falciparum-specific RDT use is low. The observed prevalence of P. falciparum is higher with a more sensitive PCR method.

Among the most widely used RDTs are those that detect Plasmodium falciparum histidine-rich protein 2 (PfHRP2). These RDTs are sensitive and specific for parasitaemias above 200 per µl blood [5]. RDTs that use the Plasmodium lactate dehydrogenase (pLDH) antigen are designed to detect all species of malaria, but are less sensitive [6][7][8].
In sub-Saharan Africa, where the vast majority of malaria infections are due to P. falciparum [1], RDTs detecting PfHRP2 are most commonly used. However, mono-infections with non-falciparum malarias (primarily Plasmodium malariae and Plasmodium ovale in sub-Saharan Africa) may go undetected. In this study, a sub-set of samples from a nationally representative, cross-sectional study of children under age 5 years in the Democratic Republic of the Congo (DRC) were used to determine the prevalence of P. malariae and P. ovale mixed and mono-infections, and to assess the risk of missed detection due to the use of falciparum-specific RDTs.

Survey methodology and sample collection
The 2013-14 Demographic and Health Survey (DHS) was a cluster-based household survey in the DRC, which took place between November 2013 and February 2014. As part of the survey, blood samples collected from children under 5 years of age were analysed for malaria infection, without speciation, by light microscopy. The samples were also tested with an RDT targeting PfHRP2 (SD Bioline Malaria Ag P.f., Standard Diagnostics, Gyeonggi-do, Republic of Korea) and used to make dried blood spots (DBS). From DBS, DNA was extracted and tested for P. falciparum infection using a real-time PCR assay targeting the P. falciparum lactate dehydrogenase (pfldh) gene as previously described [9,10]. This research was approved by institutional review boards at the Kinshasa School of Public Health and the University of North Carolina at Chapel Hill. Informed consent was obtained from a parent or responsible adult for all subjects.

All-Plasmodium real-time PCR assay
Samples that initially tested negative by pfldh PCR underwent a real-time PCR assay that detects all species of Plasmodium, targeting the gene encoding the small sub-unit (18S) of the ribosomal RNA gene (heretofore referred to as "All Plasmodium qPCR"). Primer and probe sequences, reaction mixture and cycling conditions for this assay were previously published [12] with the exception that Probe Master qPCR Mix (Roche, Indianapolis, IN, USA) was used (Additional file 1: Table S1). Each sample was tested in duplicate. A sample was considered positive for Plasmodium if both of its replicates amplified or if one replicate amplified with a cycle threshold (C T ) value lower than 38.

Speciation by BLAST
Samples that tested positive in the All Plasmodium qPCR assay were speciated using Sanger sequencing and BLAST. DNA was amplified by nested PCR of Plasmodium 18S rDNA as previously described [13], with the outer primers rPLU1 and rPLU5 and the inner, genusspecific primers rPLU3 and rPLU4. Primer sequences and reaction conditions are listed in Additional file 1: Table S1.
Products from this nested PCR were Sanger sequenced (Eton Bioscience, Durham, NC, USA) and queried in the BLAST database (National Center for Biotechnology Information, Bethesda, MD, USA). If a sequence produced a result with at least 50 % query cover and 94 % identity, it was identified as the species with the best match. Otherwise, if a result produced 10 % query cover and 80 % identity, the cleanest part of the sequence, as determined by the authors, was resubmitted to BLAST and if this search fulfilled the initial criteria the species was defined. All sequences that did not produce such a match were labelled 'indeterminate' by this method.

Speciation by species-specific nested PCR
Samples that were malaria-positive by either the pfldh qPCR or the All Plasmodium qPCR were tested using species-specific nested PCR as in [13]. The Round 1 reaction used the same primers as used in the genus-specific nested PCR but with Qiagen HotStarTaq Master Mix (Qiagen, Hilden, Germany) (Additional file 1: Table S1).
Separate Round 2 reactions were run for P. falciparum (primers rFAL1 and rFAL2), P. malariae (rMAL1 and rMAL2), and P. ovale (rOVA1 and rOVA2) (see Additional file 1: Table S1). The PCR products from each speciation reaction were analysed by gel electrophoresis on a 3 % agarose gel to determine positivity for each species in each sample. Samples that did not amplify for any of the three species were labelled 'indeterminate' by this method.
The species of each sample was determined based on results from nested PCR. If a sample was indeterminate by nested PCR, the BLAST result was used instead. If the BLAST result was also indeterminate, the sample was removed from the analysis.

Analysis
Data were entered and analysed in Microsoft Excel 2007 (Microsoft, Redmond, WA, USA). Weighted prevalence were calculated as follows: for each stratum the proportion of positive samples in the sub-set was multiplied by the number of samples in the stratum. The sum of these was divided by the total number of samples.

Study population
There were 7137 children under 5 years old with known pfldh PCR, microscopy and RDT results. Of these, a total of 466 samples (6.5 %) were chosen from the four strata listed in Table 1 for use in this study. A sample flow diagram is shown in Additional file 1: Figure S1.

All Plasmodium qPCR
Among the 316 samples that were negative by pfldh PCR, 52 samples were malaria-positive by the All Plasmodium qPCR assay-nine microscopy-positive and 43 microscopy-negative. Including these and 150 samples that were positive by pfldh PCR, a total of 202 malaria-positive samples was analysed for speciation. Out of these, four (2.0 %) were indeterminate (Table 2), giving an analysable population of 198 malaria-positive samples and 462 total samples.

Discussion
Using a sub-set of 462 samples from the large, crosssectional DHS, the prevalence of P. malariae and P. ovale among children in the DRC was found to be 12.9 and 8.3 %, respectively, with widespread geographical distributions seen in both species. A recent study of children in Western Kasai, DRC found a similar prevalence of P. malariae (13.8 %) but a lower prevalence of P. ovale (2.4 %) [14], and another study of asymptomatic individuals in six provinces of the DRC found a much lower prevalence of P. malariae (1.0 %) [15]. In 2007 prevalence were 4.9 % for P. malariae and 0.6 % for P. ovale [16]. In general, prevalence found here are higher than those reported in other African countries for P. malariae [17][18][19][20] and P. ovale [17,18,21]. However, all of these differences could be due to normal geographic and temporal variations as well as differences in the PCR and sampling methods.
Mono-infections of P. malariae and P. ovale appear to be rare in the DRC. In this study, the mono-infection prevalence were only 1.0 and 0.6 %, respectively. In Western Kasai, there were no P. malariae or P. ovale monoinfections reported [14].
Because HRP2-based RDTs detect P. falciparum only, they can result in missed detection of P. malariae and P. ovale mono-infections. Of the eight mono-infections found here, six were negative by RDT. The remaining two were likely recently cleared P. falciparum infections in which the HRP2 antigen was still present, as it can remain in the blood stream for up to 1 month after parasite clearance [22]. Overall, the number of non-falciparum infections missed by the RDT was small as most such cases were co-infections with P. falciparum.
Of 316 samples that were initially negative by pfldh PCR, 52 amplified P. falciparum 18S rDNA by nested species-specific PCR. As a result, the PCR prevalence using both tests was higher (46.6 %) than found using a single test (38.6 %). This is likely because the 18S rDNA assay has a lower threshold of detection. Thus, caution must be used when comparing prevalence rates determined using different PCR and survey methodologies.

Conclusions
Plasmodium falciparum remains the most prevalent species of malaria in the DRC, but P. malariae and P. ovale are endemic at a low rate. While RDTs have limitations, the results presented here suggest that the risk of missing malarial infections because they are mono-infections of P. malariae and P. ovale is low. However, the development of new RDTs to cover non-falciparum malaria will improve efforts at elimination.  Fig. 1 Geographical distribution of Plasmodium malariae (a) and Plasmodium ovale (b) cases by DHS cluster. For each cluster, the size of the black dot represents the number of cases tested (out of a total 462) and the size of the red dot represents the number of positive P. malariae (a) or P. ovale (b) cases