Background

Polymerase chain reaction (PCR) detecting Plasmodium DNA is the most sensitive method for diagnosing malaria, and used as reference method for routine diagnostics and in epidemiological surveys. However, PCR can be positive weeks after effective malaria treatment, potentially leading to over-diagnose of recrudescence and re-infections [1]. Residuals of destroyed asexual parasites, and live or destroyed gametocytes, are possible sources of post-treatment Plasmodium DNA in patients cured of their malaria infection. Gametocytes are sexual forms of malaria parasites responsible for transmission between humans after completion of their cycle in mosquitoes. They cause no clinical symptoms in the human host. In Plasmodium falciparum, immature gametocytes sequester in internal host organs, particularly in the bone marrow, and undergo five morphological development stages in the course of 7–12 days, hence only mature gametocytes circulate in the blood until they die of age [2, 3]. It has been estimated from mathematical modelling that P. falciparum gametocyte carriage may persist for up to 55 days after treatment [4]. Primaquine (PQ) is the only drug effective against mature gametocytes, and is added to artemisinins in routine treatment of P. falciparum in many malaria-endemic areas to prevent transmission [5, 6]. Artemisinin also has some effect on gametocyte carriage duration, through the effect on immature gametocytes and rapid killing of asexual parasites [5]. Less is known about Plasmodium vivax gametocytes. It seems that immature P. vivax gametocytes sequester mainly in the bone marrow, similar to P. falciparum [7]. However they mature much faster than P. falciparum gametocytes, they are commonly present before symptoms and before parasite detection by microscopy, and live only for up to three days [2, 8].

Several malaria PCR methods have been introduced applying different amplification targets on the parasite genome. The real-time PCR targeting var gene acidic terminal sequence (varATS) on the chromosomal genome, and cytochrome b (cytb) on the mitochondrial genome, are among the most sensitive ones [9]. The varATS gene exists in 59 copies for each parasite nucleus [10], while the number of copies of the cytb gene varies depending on parasite stage; ring-stage parasites have about 20 copies, and gametocytes about 160 copies [11, 12]. Gametocyte-specific mRNA transcripts can be detected by reverse transcript (RT)-PCR methods [13, 14].

The objective of this clinical observational study was to describe the presence of positive PCR for P. falciparum and P. vivax in follow-up samples post-treatment from returned travellers, and the proportion of positive PCR due to gametocytes, applying PCR methods detecting Plasmodium DNA and gametocyte-specific mRNA.

Methods

Patient material

Blood samples from adult patients, diagnosed with malaria by microscopy and/or rapid diagnostic test at Haukeland University Hospital, Bergen, were collected in the period 2013–2015. Blood samples were collected during hospitalization and at routine follow-up in the outpatient clinic at 11–54 days after treatment. Clinical information and results from routine microscopy was collected retrospectively from patient records.

PCR methods

DNA was extracted from EDTA whole blood applying QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany), while mRNA was collected and extracted applying PAXgene Blood RNA Tubes and Kit (PreAnalytiX, Hombrechtikon, Switzerland), according to the manufacturer’s instructions.

All the DNA samples (N = 25) were analysed with a genus-specific cytb real-time SYBR PCR, and the P. falciparum DNA samples (N = 22) were also analysed with a falciparum-specific varATS real-time TaqMan PCR, following previously published methods [9, 10]. All the mRNA samples (N = 18) were analysed both with a genus-specific 18S rRNA RT-PCR (detecting mRNA transcripts produced by all parasite stages), and a gametocyte-specific Pfs25 (P. falciparum)/Pvs25 (P. vivax) RT-PCR detecting female gametocytes, following previously published methods [14]. In addition, quantitative analysis was performed for the cytb real-time PCR and Pfs25/Pvs25 RT-PCR runs by applying tenfold dilution series of customized plasmids designed with EcoRI linearized q-PCR template and a pUCminusMCS vector backbone (OriGene Technologies, Rockville, MD, USA). All samples were run in triplicates, and a positive result was defined as at least two out of three detections.

Results

The PCR results applying DNA and mRNA template, including mean Ct values as well as quantitative numbers, are presented in Table 1. Genus DNA cytb PCR was positive up to 49 days after effective treatment, and 18S rRNA transcripts from active P. falciparum parasites were detectable for at least 11 days. Microscopy and/or PCR detected gametocytes in 67% (8/12) patients, but gametocyte-specific mRNA was detected at latest only two days after treatment. Late detection of P. vivax was not found by any of the PCR methods.

Table 1 Plasmodium DNA and gametocyte-specific mRNA detections among malaria patients after treatment (N = 13)

All patients were asymptomatic and had negative microscopy at outpatient follow-up visits. The association between positive PCR, microscopy findings and clinical characteristics is presented in Table 2. Six patients with P. falciparum were treated with artesunate intravenously, followed by a full oral course of artemether-lumefantrine (AL) or atovaquone-proguanil (AP), while two received only AL and one only AP. All P. vivax patients, and no P. falciparum patients, received primaquine. Among the six patients with late P. falciparum positive PCR (day 11–49), four had high parasitaemia (6–30%) while two had parasitaemia < 2% at admittance, and three had gametocytes detected. A clear association between low/high P. falciparum parasitaemia and duration of positive PCR detections was not found. One patient with late positive PCR was of Norwegian origin, while five were from sub-Saharan Africa and had lived from 10–46 years in Norway. Two of the patients with late positive PCR were immunodeficient with HIV or sickle cell disease.

Table 2 PCR results associated with microscopy findings and clinical characteristics

Discussion

In this observational study, DNA-based PCR was positive up to seven weeks after effective malaria treatment. No patients had clinical signs of recrudescence, and there was no risk of re-infection. This is in line with a previous report that found positive malaria PCR after six weeks in a group of returned travellers in Sweden [1]. Positive PCR for weeks after effective treatment of infections is a common phenomenon; a study investigating detectable DNA after treatment of Chlamydia trachomatis, Neisseria gonorrhoeae and Trichomonas vaginalis infections, reported that PCR could be positive up to three weeks post-treatment [15].

Molecular studies have shown that most individuals with asexual parasites also have sub-microscopic gametocyte carriage [16]. In general, the level of circulating gametocytes is low, about 5% compared to other parasite-stages. Gametocytes were detected by microscopy and/or RT-PCR in 67% (8/12) of the patients in the present study. However, gametocyte-specific mRNA was not detected in the late follow-up samples, similar to that reported by Vafa Homann et al. in Sweden [1], which might indicate that the detection of Plasmodium DNA origin from residuals of destroyed parasites. The phagocytic system has the potential to remove up to 40–80% of malaria infected red blood cells (RBCs) in a few days, but due to sequestration of infected RBCs in organs, the time of complete clearance of parasite residuals is unknown [17]. Drug treatment may also have contributed to clear gametocytes. All the P. falciparum patients were treated with artemisinin, which has some gametocidal effect, and all P. vivax patients were treated with PQ, which has a strong gametocidal effect. In a study from Kenya and Tanzania estimating gametocyte carriage following treatment with non-artemisinin drugs, artemisinin, and artemisinin in combination with PQ, duration of gametocytaemia was 55, 13 and 6 days, respectively [4].

Collection, handling and analysis of mRNA is challenging [18, 19], and due to the instability of mRNA versus DNA, difference in sensitivity between the methods can be a factor that may underestimate late gametocytaemia in this study. A study investigating a similar gametocyte-specific Pfs25 RT-PCR, reported that in dilution series down to 0.05 and 0.01 gametocytes/µl, the lowest density samples were often negative [20]. In the present study, the quantitative values detected by cytb DNA PCR in samples > 30 days after treatment correspond to ≤ 0.05 gametocytes/µl, so potentially these could have gametocytaemia below detection level of mRNA RT-PCR.

In the study from Kenya and Tanzania investigating gametocytaemia, a nucleic acid sequence based amplification (NASBA) method detecting Pfs25 mRNA was applied, a method slightly more sensitive than Pfs25 RT-PCR; gametocytes were found in 5–78% of the samples at day 14, and 12–48% at day 28 [4]. However, compared to returned travellers in the present study, patients in malaria-endemic areas may have a higher level of gametocyte carriage post-treatment due to reasons such as late admission and delayed treatment, use of less effective drugs, self-medication with sub-optimal regimes, re-infection, or gametocyte carriage that originate from previous undetected low-density infections.

Looking only at the two DNA-based PCR assays, the cytb PCR detected more late follow-up samples than varATS PCR. Hypothetically this could be explained by detection of gametocytes, since the cytb method detect mitochondrial DNA present in large amounts in gametocytes. In regard to copy number, the cytb PCR is about 2.5 times more sensitive in detecting gametocytes than varATS PCR, while the varATS PCR is about three times more sensitive than cytb PCR in detecting asexual parasites. In a study applying field samples from Tanzania, the sensitivity in detecting low level parasitaemia by different real-time malaria PCR methods where compared, and the varATS PCR was then found to be more sensitive than cytb PCR [9].

Two samples 11 days after treatment were positive by 18S rRNA RT-PCR (mRNA), but not by the gametocyte specific Pfs25 RT-PCR, indicating that possible live asexual parasites circulated in the bloodstream in low densities at this point. Although, negative gametocyte specific Pfs25 RT-PCR due to lower sensitivity could also be possible. The quantitative levels by the DNA PCR after 11 days correspond to < 5 parasites/µl. Post-treatment asexual parasites that survive until they are cleared by the immune system continue to produce gametocytes, which also could support the reported phenomenon of gametocyte carriage for several weeks after treatment [4].

For the P. vivax samples, a high level of mRNA from gametocytes was detected by the Pvs25 RT-PCR at day 2 after treatment, but both mRNA assays were negative for the late follow-up samples. The cytb DNA PCR was positive at day 6, though with parasitaemia as low as 0.025 parasites/µl. These results are consistent with the short duration of gametocytaemia in P. vivax [2]. The results for P. vivax also support that persistent positive P. falciparum PCR could be caused by gametocytes; if positive P. falciparum PCR was caused only by DNA residuals, it would be expected that also P. vivax was PCR positive days/weeks after treatment similar in P. falciparum infections. An association between high P. falciparum parasitaemia and persistent PCR would intuitively be expected. However, similar to that reported by Vafa Homann et al. among Swedish travellers [1], no clear association between low/high level of parasitaemia and persistent positive PCR was identified in the present study.

Conclusions

DNA-based PCR can be positive up to at least seven weeks after curative malaria treatment, potentially leading to over-diagnose of recrudescence and re-infections. Based on the observations in this study, it is unclear if the DNA origin from residuals of destroyed parasites or live gametocytes. Further studies of both P. vivax and P. falciparum are needed, since different gametocytaemia biology and treatment-regimes have the potential to give informative answers.