Gametocyte carriage of Plasmodium falciparum (pfs25) and Plasmodium vivax (pvs25) during mass screening and treatment in West Timor, Indonesia: a longitudinal prospective study

This study was conducted in Wewiku subdistrict, Belu district, East Nusa Tenggara province, in eastern Indonesia (June–September 2013). The study area was reported to have the highest malaria endemicity in the province. The annual parasite incidence (API) per 1000 population was 72 and 124 in 2011 and 2012, respectively (Belu district health office, personal communication), whereas in the entire East Nusa Tenggara province it was 25.8 and 21.1 during the same period [30]. Malaria incidence peaked during the months of August and September which coincided with the end of this study [26]. Anopheles barbirostris was reported as the major malaria vector [26]. Other less dominant vectors included Anopheles subpictus and Anopheles vagus [26]. A more detailed description on the study site has previously been reported [26].

The original cluster-randomized trial evaluated the impact of MST conducted (a) twice, and (b) three times, both within 3 months [26]. Samples were collected in June (round-1), July (round-2) and August (round-3) [26]. Positivity for malaria parasites by microscopy was 8% in both groups at baseline, and 9% each at the last round of MST [26]. MST exerted no impact on measured incidence of diagnosed infections in both groups (RR = 0.89, 95%CI 0.43–1.83) [26]. Therefore, in this study, round-1 of MST from both treatment groups were pooled and considered the ‘baseline’, while the last time-point (round-3) from both groups was also pooled and considered the ‘endpoint’.

Finger-pricked blood samples from each of the study subjects were collected for malaria diagnosis using microscopy and 250 µL were also collected in EDTA tubes. Blood slides were air-dried overnight prior to Giemsa staining. Upon staining, microscopic readings were performed on the same day. P. falciparum and P. vivax-positive blood samples were screened for glucose-6-phosphate dehydrogenase (G6PD) deficiency using a fluorescence spot test (Trinity Biotech, Ireland), and subjects with deficient or intermediate G6PD activities were not given primaquine. The time required from the sample collection until treatment administration were 2–3 days. All microscopy-positive subjects were treated with dihydroartemisinin–piperaquine (DHP) and primaquine. Treatment consisted of DHP (each tab contains 40 mg dihydroartemisinin and 320 mg piperaquine) for 3-day with daily doses as follows: ≥ 60 kg = 4, 41–59 kg = 3, 31–40 kg = 2, 18–30 kg = 1.5, 11–17 kg = 1, 6–10 kg = 0.5 and ≤ 5 kg = 0.25 tablet(s). Primaquine administration for P. falciparum were 0.75 mg/kgBW single dose: ≥ 60 kg = 3, 41–59 kg = 2, 31–40 kg = 2, 18–30 kg = 1.5, 11–17 kg = 0.75 tablet(s). For P. vivax, the daily dose for 14 days were 0.25 mg/kgBW: 1, 1, 0.75, 0.5, 0.25 tablet for the same weight group. A pill cutter was used to cut 15 mg of primaquine tablet to meet the required dose according to the body weight. A pill grinder was used to make powder with added saccharine for small children, e.g. < 6 year old, who were unable to swallow tablets. Quality control with cross-checking of microscopic reading was performed post-hoc with kappa value = 0.82 and 0.85 for P. falciparum and P. vivax, respectively. Discordant speciation between microscopy and PCR was not an issue as same drugs were given for both species.

For RNA extractions, 50 µL of blood was mixed with 250 µL of RNAProtect (Qiagen, Germany) within 4 h of collection and stored at − 80 °C until further processing. During five years of storage, temperature was checked daily.

Microscopic and PCR screening results have been reported in the parent study [26]. All P. falciparum and a set of randomly selected P. vivax positive samples underwent RNA extraction (Fig. 1) using the Quick RNA Mini-prep kit (Zymo Research, USA) according to manufacturer’s instruction. DNase treatment (included in the kit) was done during the process. P. falciparum and P. vivax gametocytes were quantified by two-step RT-qPCR of the gametocyte markers pfs25 and pvs25 [31]. Transcriptor First Strand cDNA kit (Roche) was used to generate cDNAs for each sample in triplicates. The presence of mRNA transcripts was verified using RT-qPCR targeting 18S rRNA [32], and only those positive for 18S were analysed for pfs25 and pvs25. The pfs25 and pvs25 qPCR was conducted using FastStart Essential DNA SYBR Green master (Roche) with previously published primers [29]. Details on the assays are provided in the Additional file 1. Melting temperature (T_M) for P. falciparum was 74–75 °C, and 79–80 °C for P. vivax (Additional file 2). For quantification, series of plasmid harbouring target sequence with concentration of 10⁵, 10⁴, 10³, 10², 10, 5, and 1 copy(ies) per reaction were run in triplicate and a standard curve was generated for each run. Negative (no template) controls were included in triplicates. Minus-reverse transcriptase (−RT) qPCR using pfs25 and pvs25 was performed on 23% (19/83) of randomly selected of P. falciparum and 17% (40/231) of P. vivax RNA samples to check for genomic DNA contamination. Sixteen of 19 P. falciparum demonstrated no contamination with DNA while three showed the presence of DNA—with 0.01–1.28% of pfs25 RT qPCR result. Thirty-six of 40 P. vivax demonstrated the absence of DNA, while four showed presence of 0.4%–2.5% DNA with pvs25 RT qPCR results. The limit of detection (LOD) was assessed by running a serial dilution of the plasmid in quintuples and determined as 10 copies/µL for both pfs25 and pvs25. Performance of the assay is described in Additional file 3. All triplicates of the cDNA were run and recorded as positive if a minimum of two of the three demonstrated a positive result. Quantities of the transcript were reported as the average transcript numbers of the replicates to the nearest CT values. Conversion of transcript numbers into gametocyte density was made by extrapolating transcript numbers to the standard curve generated from transcript numbers of the known gametocyte density in the microscopy-positive subjects. Regression coefficient (r) was 0.758 and 0.835 for P. falciparum and P. vivax, respectively. The trendline equation for P. falciparum was: 10^{0.251 [−0.607–1.109]} * (pfs25 transcript numbers/µL)^{0.621 [0.357–0.886]}, whereas for P. vivax was: 10^{0.673 [0.198–1.149]} * (pvs25 transcript numbers/µL)^{0.406 [0.264–0.548]} (Additional file 4). Using this equation, one gametocyte corresponds roughly to 0.4 (0.1–50.2) pfs25 transcripts/µL for P. falciparum and 0.02 (0.01–0.18) pvs25 transcripts/µL for P. vivax. DNA from samples which were tested for gametocytes were quantified to assess the relationship between asexual and sexual stages. This was performed by species-specific 18S qPCR as described previously with some modification [33]. Details of the procedure and performance of the qPCR were shown in Additional files 1 and 3.

All laboratory analyses were performed blindly to subject identification (ID). Upon completion of all laboratory work, the results were linked to the subject ID, MST time point, and treatment status. All laboratory analyses were conducted at the Indonesian Medical Education and Research Institute (IMERI), Faculty of Medicine, University of Indonesia, Jakarta.

This study was unable to be powered to detect any change in gametocyte prevalence during MST as originally planned given the much lower prevalence than estimation in the protocol. Moreover, the design of the study was modified to pre-post instead of a between-group analysis. Categorical variables were analysed using chi square test. Numerical variables were analysed using Student’s t-test when the distribution was normal or Mann–Whitney when it was not normal. 18S rRNA gene copies and pfs25/pvs25 transcripts were log transformed for parametric statistical analyses. Linear regression was conducted to investigate the correlation between parasite density and pfs25/pvs25 numbers. A p-value < 0.05 was considered as statistically significant. All analyses were performed using SPSS version 23 (IBM, Armonk, New York).

A total 866 subjects participated in this study: 811 samples were collected at baseline and 740 at the endpoint. These subjects represented 78% of the total parent study (reference) samples. Age and gender of study subjects were comparable between baseline and endpoint (33% were ≤ 15 year of age and 47% were male). Furthermore, 685 (84%) subjects were sampled at both time points.

Of the 102 P. falciparum positive samples by microscopy and/or PCR from the two time points, RNA extractions were not conducted in three samples due to insufficient blood volume, and 13 samples due to human error. Of the remaining 86 RNA samples, 3 were negative for 18S transcripts, with 83 remaining for pfs25 analysis (Fig. 1). Of the 334 P. vivax positive samples by microscopy and/or PCR from the two time points, 246 were randomly selected for RNA extraction. Of these 246 RNA samples, 15 were negative for 18S transcripts—thus 231 samples were available for pvs25 analysis (Fig. 1).

In P. falciparum, gametocyte transcript numbers were not correlated with parasitaemia density by microscopy (r = 0.198, p = 0.202, Fig. 2a), whereas in P. vivax, positive correlation was demonstrated (r = 0.514, p < 0.001, Fig. 2b). However, gametocyte transcript numbers were positively correlated with 18S gene copy numbers for both species, with a stronger correlation observed for P. vivax than P. falciparum (P. falciparum: r = 0.372, p = 0.003; P. vivax: r = 0.770, p < 0.001, Figs. 3a, b).

Every parasitemic and gametocytemic subject was identified individually at each timepoint, and their infection status was followed or traced backwards. The treatment status was also linked for each identified individual.