Background
Reliable malaria diagnosis including qualitative and quantitative detection of all malarial blood stages remains very important in the campaign against this disease, which is believed to kill one to two million children annually [
1]. Comparing diagnostic approaches that have included conventional blood smear (CBS) light microscopy, enzyme-based rapid diagnostic tests (RDTs) and PCR-based strategies has revealed a greater complexity of malaria infection in individuals and population studies [
2]. Furthermore, direct comparisons of these methods illustrate differences in sensitivity and specificity, cost and efficiency of diagnosis. These comparisons show that no single diagnostic approach can be used to satisfy all malaria diagnostic expectations. Whereas RDTs can be performed rapidly in the field or by highly sensitive and specific PCR-based processing in well-equipped laboratories, these methods lose the ability to evaluate parasite morphology that is exquisitely informative regarding parasite developmental stages of all the human malaria parasites. Quantitative evaluation of gametocytes in particular, could be useful in assessing potential for malaria transmission from humans to vector mosquitoes.
Recently, we developed a novel method, termed magnetic deposition microscopy (MDM) to enhance the detection of malarial blood stages exploiting their paramagnetic characteristics [
3]. Malarial parasites convert erythrocyte haemoglobin to paramagnetic haemozoin [
4]. The strong magnetic field of a permanent magnet is used to isolate infected red blood cells from a cell suspension and accumulate them on transparent polymer slides suitable for light microscopy. In preliminary studies on infected blood from splenectomized non-human primates we demonstrated that it was possible to increase the observable number of erythrocytic malarial blood stages by MDM at least 40-fold [
3]. In this previous study, MDM yielded higher sensitivity than conventional thick blood smears with the clarity of a thin smear for diagnosis of all four human malaria parasite species. However, while trophozoites, schizonts and gametocytes were present in significantly higher numbers, we observed a deficiency of
Plasmodium falciparum ring stages on the MDM slides.
Here we extend observations from our initial study to determine if it is possible to apply the MDM diagnosis of P. falciparum malaria in human infections through a field study in a malaria-endemic region of Papua New Guinea (PNG), conducted between March and June 2007. Moreover, as peripheral blood samples from acute symptomatic P. falciparum infections are characterized by a majority of ring-stage parasites that have yet to accumulate haemozoin, it was critical to evaluate the success in capturing these forms specifically.
Discussion
Our study here demonstrates that MDM [
3] successfully detects
P. falciparum in individuals experiencing symptoms of malarial illness in Papua New Guinea. Although it was surprising that we did not detect non-falciparum species by MDM in this region where we have consistently observed all four human
Plasmodium parasite species in past studies [
7,
9,
10], patient samples were obtained from studies evaluating antimalarial drugs where enrollment criteria avoided complex infections. Moreover, the species-specific detection of
P. falciparum only, by our recently developed post-PCR LDR-FMA assay [
8] for all four human
Plasmodium species parasites, lends strong evidence that study participants were not infected with non-
falciparum species.
Modifications to the MDM device used in this study offer conveniences that facilitate transportation, set up and operation for day-to-day use. More importantly, the introduction of a second interpolar gap increased the overall surface area of the magnetic gradient in the current MDM device. Of additional interest, we also observed that positioning the device at a 45° angle to the bench-top surface led to spreading of the blood volume across the entire surface of the sample flow path to increase the number of cells exposed to the magnetic field per blood volume assayed. Future experiments will be needed to determine how combined interactions between magnetic field, gravity and surface tension influence the flow of Plasmodium infected and uninfected cells through the MDM device. Of practical importance here, these modifications have contributed to our observations reporting routine capture of P. falciparum ring-stage parasites on MDM microscope slides at levels consistent with CBS.
Although our earlier work demonstrated that gametocytes were susceptible to MDM capture [
3], our study here confirms these results in a malaria-endemic field setting using infected human blood samples, and presents observations that are integral to gametocyte biology and transmission of
P. falciparum. This includes evidence that MDM captures both male and female gametocytes, and that many individuals carry numbers of mature gametocytes in blood volumes within the range ingested during female mosquito blood meals. Numerous recent studies have reported on gametocyte prevalence that occurs at, or below, the limits of CBS detection [
11‐
18]. The majority of these studies have relied on the increased sensitivity of reverse transcriptase-PCR methods to detect mRNA expressed specifically during gametocyte stages of development [
11‐
13,
15‐
18]. While these approaches have provided important insight regarding a critical component of
P. falciparum diagnosis, they require laboratory-based procedures that are difficult to perform in a timely fashion in close proximity to the field settings where human to mosquito transmission occurs. Furthermore, quantitative assessments of gametocyte levels by nucleic acid based methods will only provide a rough approximation of mature gametocyte numbers [
19,
20]. Interestingly, the increased prevalence at which gametocytes were observed in patient samples by MDM vs. CBS in this study (6.2-fold increase; CBS gametocyte prevalence similar to previous PNG-based studies [3.3–13%] [
21,
22]) is similar to findings comparing nucleic acid-based diagnostic methods vs. CBS [
11,
13,
16,
18].
As the level at which gametocytes become optimally infectious to mosquitoes is not clear, or may vary based on endemic seasonality, or local vector species preferences [
23‐
25], it is important to improve the capacity for identifying and enumerating gametocytes in human blood samples. A number of field observations now suggest that humans are infectious to mosquitoes even at gametocyte levels below detection limits of CBS light microscopy [
14,
16,
17,
26]. In order to evaluate the impact of current insecticide-treated bed net, antimalarial drug or vaccine control programs, it will be necessary to enumerate accurately mature gametocyte densities circulating in the blood of infected people. As MDM diagnosis provides counts of mature male and female gametocytes within in given blood volume, this diagnostic strategy may help to clarify the relationship between gametocyte density and productive infection of mosquitoes.
In future studies it will be important to expand direct comparisons of malaria diagnosis performed by MDM, CBS (against, 200, 500 and 1000 leukocytes), rapid diagnostic tests (RDTs), and PCR-based techniques targeting
Plasmodium species sequences. An overall comparison of diagnostic strategies will provide opportunity to determine if this new microscopy-based method is capable of delivering sensitivity rivaling PCR, along with the morphological resolution of a thin smear blood film. Also, as PCR-based diagnostic approaches are far superior to other diagnostic strategies for
Plasmodium species identification, it will be important to evaluate the concordance of species diagnosis between MDM and PCR. Direct comparison with the antigen capture-based RDTs would also apply continuing pressure to improve performance of products that are currently expensive, frequently deliver lower sensitivity than CBS [
27], and are susceptible to reduced sensitivity by amino acid sequence polymorphism. Moreover, it is important that all of the diagnostic strategies are forced to compete so that durable, rapid, inexpensive and accurate malaria diagnostic tools are available in settings where the disease is most difficult to control.
Finally, whereas the study performed here compared CBS and MDM diagnosis on people reporting to a health center with malaria symptoms, it will be important to perform a similar comparison study among a series of asymptomatic individuals. A study of this nature will enable evaluation of MDM on a large number of people who are likely to carry very low levels of parasitaemia. As our past studies have identified frequent mixed Plasmodium species infections in cross-sectional surveys, we would expect MDM to detect mixed species infections in asymptomatic individuals more often than CBS. Further, gametocytes are thought to be most prevalent in individuals who are experiencing clinical symptoms, and less common in individuals who have developed acquired immunity against Plasmodium species parasites. A survey of asymptomatic individuals, including MDM diagnosis, will provide new insight regarding the overall prevalence of gametocytes in malaria-endemic populations. This may reveal important new reservoirs contributing to malaria transmission and further refine malaria control strategies.
Acknowledgements
We are grateful for all study volunteers for their willing participation in our ongoing malaria field studies. We thank William Kastens, Livingstone Tavul, Danielle Stanisic, Tim Davis, Harin Karunajeewa and Peter Siba for their technical assistance and support of our field studies. These studies were coordinated with the PNGIMR Madang Molecular Parasitology Unit. This study was supported by grants from the National Institutes of Health (AI36478-04S1 and AI52312). The funding agency played no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Authors' contributions
SK contributed to the experimental and epidemiologic designs, performed all sample preparation, processing and evaluation of CBS and MDM slides. BTG contributed to the experimental design. MD, PM and IM contributed to coordination of the field studies in PNG. SK, LM and MZ designed the MDM device. PAZ developed the Plasmodium species PCR-based diagnostic assay. MZ and PAZ supervised the overall project. SK, MZ and PAZ wrote the manuscript, with assistance from their collaborating authors. All authors read and approved the final manuscript.