Deployment of Molecular Tools to Track the Epidemiology of Plasmodium Vivax in Panama


 Background: As the elimination of malaria in Mesoamerica progresses, detection of Plasmodium vivax asymptomatic patients using conventional diagnostic methods becomes more difficult. Highly sensitive molecular methods are key for the determination of the hidden reservoir of malaria transmission on the road to elimination in countries in the pre-elimination phase such as Panama. Here we describe the clinical validation of a qRT-PCR assay for the detection of P. vivax asexual and sexual stages from low blood volume field samples preserved at ambient temperature. Methods: We collected blood samples from a cross sectional cohort of P. vivax patients in Panama. Different storage formats (room temperature, frozen) and blood volumes were compared to establish the sensitivity of parasite detection including transmission stages (gametocytes) by qRT-PCR and diagnostic microscopy. Results: Study results indicated that blood storage at room temperature using an RNA preservation solution for up to 8 days was sufficient to preserve RNA for subsequent qRT-PCR assays. Detection of gametocytes by qRT-PCR was more sensitive than light microscopy using both our recently established marker PvLAP5 and the gold standard Pvs25, confirming that both markers are suitable for P. vivax gametocyte detection in the field using the above protocol.Conclusions: This study validates a low blood volume qRT-PCR assay system for the detection of P. vivax asexual and sexual stages in field samples preserved at ambient temperature. Results indicate that the assay system is a reliable tool to determine the transmission reservoir of P. vivax in remote areas such as endemic regions of Panama.

symptoms (21,22). The reason for the early transmissibility is the relatively short gametocyte development of approximately 48 hours (14) compared to 10-12 days in P. falciparum. As in P. falciparum, developing (immature) P. vivax gametocytes are predominantly found in the hematopoietic niche of the bone marrow (and possibly spleen) (14,23,24). The detection of mature P. vivax and P. falciparum gametocytes in blood samples by light microscopy is imprecise due to their low levels in circulation (25). Molecular diagnostic tools that detect asymptomatic P. vivax carriers with sub patent infections have been developed (18,26). These assays use the P. vivax 18s ribosomal RNA gene (Pv18SrRNA) or the mature gametocyte marker Pvs25 (25,27). Notably, both markers can amplify from genomic DNA (gDNA). Other P. vivax gametocyte markers such as Pvs28, Pv41, Pvs48 /45, and Pvs230 have been described and characterized (28-32), but none of these genes has been validated as a gametocyte detection tool in eld samples. We recently characterized PvLAP5 as a P. vivax mature gametocyte marker by qRT-PCR assay and using a speci c antiserum (14,33).
Here we present establishment and validation of a eld deployable diagnostic test for detection of P. vivax asexual and gametocyte stages, using nger prick blood, storage and transport at ambient temperature and qRT-PCR ampli cation of Pv18s rRNA, Pvs25 and PvLAP5 markers.
We expect that this assay will contribute to the detection of the asymptomatic P. vivax reservoir, and hence accelerate the elimination of persistent P. vivax malaria transmission from endemic foci in low transmission settings such as Panama.

Methods
Study Design. The aim of the study was to validate a qRT-PCR assay for the detection of P. vivax asexual and sexual stages using small blood volume and no cold chain. Clinical samples were collected during 2017-2019 from Pv malaria positive and negative individuals detected by passive or active surveillance by technicians from the National Vector Control Department of the Ministry of Health (MINSA) of Panama.
Ethics. Study protocol and consent form approval was obtained from The Gorgas Memorial Institutional Bioethics Review Committee (No. 276/CBI/ICGES/16). Written informed consent was obtained from the participants. Animal blood samples used in this study were obtained from the ICGES malaria strains repository, or from animals inoculated for use as donors in other protocols. Collection of malaria naïve Aotus monkey blood was carried out as part of a routine animal health program. All animals were maintained and treated in accordance with the Guide for the Use for the Care and Use of Laboratory Animals, eighth edition 2011, National Research Council, Washington, DC.
Epidemiology of Plasmodium vivax in Panama during 2017-2020. P. vivax malaria incidence maps at the level of corregimiento (smallest political division) for the years 2017-2020 were prepared with data obtained from the National Vector Control Program of the Ministry of Health of Panama using ArcMap 10.6.1. software (Esri, Redlands, CA). Epidemiological curves by year, month and age groups for the years 2017-2020, as well as the ethnicity distribution of study participants for the years 2017-2019 were prepared using the Prism 6.0 (GraphPad Software, Inc, La Jolla, CA, USA) and Excel (Microsoft, Seattle, WA) software. Spatial, demographic, and socioeconomic characteristics of the study population. Spatial, demographic, and socioeconomic information was gathered from each study participant using an epidemiological survey form developed with the Survey123 for ArcGIS online survey software (Esri, Redlands, CA).
Blood sampling. Thin and thick blood smears were prepared from a nger-prick made with a lancet. Blood smears were then dried and transported to the laboratory for staining with Giemsa, parasite density determination, species identi cation and stage differential counts. Additionally, 60-120 µL of nger prick blood were collected into 1.8 ml NUNC® cryovials containing 500 µL of RNAprotect® (RNAp) (Qiagen, Germany) for RNA isolation and qRT-PCR assay. Samples were transported to the laboratory at ambient temperature and the cryovials stored at -80 Celsius at arrival. In total, ~ 150 µL of blood were obtained from each volunteer including blood smears.
Microscopy. Giemsa stained thick and thin blood smears were examined by light microscopy for parasitemia density determination, Plasmodium species con rmation and stage differential counts. Parasite densities were calculated by quantifying the number of malaria infected red blood cells (iRBCs) among 500-2000 RBCs on a thin blood smear and expressing the results as % parasitemia (% parasitemia = parasitized RBCs /total RBCs) x 100), or quantifying parasites against white blood cells (WBCs) on the thick smear until 500 or 1000 WBCs were counted (parasitized RBCs x µL of blood, assuming 8,000 WBC/µL of blood). Stage differential counts were expressed as percentage of total parasite stages counted.
qRT-PCR assay Parasites. P. vivax SAL-1 Aotus infected whole blood from experimentally inoculated and malaria naive Aotus, kept at the Gorgas Memorial Institute in Panama, were used as positive and negative controls for the qRT-PCR assay as described (33). Heparin anticoagulated whole blood from fteen male and female Aotus monkeys was used as negative controls to determine the cut-off point Cycle Threshold (Ct) value of the qRT-PCR assay. P. vivax SAL-1 infected anticoagulated (Sodium Citrate 4% Solution, Sigma, St. Louis, MO) whole blood obtained from a donor Aotus animal MN12939 was used as positive control.
Primers. We used forward and reverse primers sets for PvLAP5 and Pvs25 and Pv18SrRNA as previously described (33) (Supplementary Table S1). PvLAP5 primers were designed to span exon-exon junctions to minimize ampli cation from gDNA. As gold standard control, we used primers for the gametocyte marker Pvs25. Primer sets including PvLAP5, Pvs25 and Pv18SrRNA were synthesized by Genscript (Piscataway, NJ, USA).
RNA extraction and cDNA synthesis. RNA was isolated from RNAp preserved blood samples with the Qiagen RNAeasy® Plus kit including a gDNA eliminator column (Qiagen, Germany) per the manufacturer's instructions. RNA concentration was measured in a NanoDrop® ND spectrophotometer (Thermo Fisher Scienti c Inc, USA) and the nucleic acid treated with DNA-free™ kit (Ambion, Life Technologies, USA) for removal of residual DNA. The treated RNA was then transcribed to cDNA with the QuantiTect® Reverse Transcription Kit (Qiagen, Germany) following the manufacturer's instructions.
Procedure for the qRT-PCR assay. Assay reactions were performed in a QuantStudio™ 5 Real-Time PCR 384 well plate system (Applied Biosystems™) as described (14). Each Fast SYBR Green reaction ( nal volume of 20 µL) consisted of Master Mix Fast SYBR Green (Applied Biosystems™) forward and reverse primers mix at 300 nm concentration and 2 µL of cDNA. Thermal cycle conditions were as follows: 10 min at 95 C, followed by 40 cycles at 95 C for 15 s, 60 C for 1 min. A melting curve analysis was added at the end of the reaction cycle. Samples were analysed in duplicate. Each plate included a positive and negative control (uninfected sample) and a negative ampli cation control. A Ct value of ≤ 38 for the endogenous Pv18SrRNA gene marker was used as the positive threshold for P. vivax detection. The Ct cut-off point of ≤ 38 was calculated from the mean Ct value of sixteen malaria smear negative human and fteen Aotus monkey controls minus two standard deviations as shown on Supplementary Tables S2 and S3. Field validation of the qRT-PCR assay qRT-PCR assay of eld samples. To validate the qRT-PCR assay and sample preservation system in the eld, we determined the mean negative Ct value threshold using 16 smear negative samples for each marker. The negative Ct value threshold was de ned as the mean Ct value -2 standard deviations from the mean. We subsequently tested 45 smear positive P. vivax samples for PvLAP5, Pvs25 and Pv18SrRNA as described (14). Representative qRT-PCR assay ampli cation and melt curve plots of a positive P. vivax sample is shown in Supplementary Figure S1.
Assay validation. Using the open web based tool "Diagnostic Test Evaluation Calculator" (https://www.medcalc.org/calc/diagnostic_test.php) (MedCalc Software Ltd, Osten, Belgium) we determined the following parameters: i) the sensitivity (Se, probability that a test result will be positive when the disease is present (true positive rate)); ii) the speci city (Sp, probability that a test result will be negative when the disease is not present (true negative rate)); iii) the positive likelihood ratio (PLR, ratio between the probability of a positive test result given the presence of the disease and the probability of a positive test result given the absence of the disease (True positive rate / False positive rate = Sensitivity / (1-Speci city)); iv) the negative likelihood ratio (NLR, ratio between the probability of a negative test result given the presence of the disease and the probability of a negative test result given the absence of the disease (False negative rate / True negative rate = (1-Sensitivity) / Speci city)); v) the positive predictive value (PPV, probability that the disease is present when the test is positive); and vi) the negative predictive value (NPV, probability that the disease is not present when the test is negative). These two last de nitions depend on the disease prevalence (34,35).
The data was then tabulated on a series of 2 x 2 tables as follows: a) the number of P. vivax microscopic eld positive slides (disease present), b) negative control smears (disease absent), c) the number of qRT-PCR positive samples (test positive) and d) number of negative control samples (test negative) for each gametocyte gene marker (PvLAP5 and Pvs25) and the endogenous marker Pv18SrRNA as described (36-38). For validation we calculated the theoretical minimum number of positive and negative samples necessary to achieve a level of sensitivity of 97% and speci city of 99% with a margin of error of 2-5% and a con dence level of 95% as described (37).

Results
The overall goal of this study was to validate a qRT-PCR assay for the detection of P. vivax using clinical samples from Panama. Speci cally, we aimed to i) compare the microscopic parasite detection rate (gold standard) to detection by qRT-PCR assay using mature gametocyte markers Pvs25 and PvLAP5 and constitutive marker Pv18SrRNA, and ii) validate the assay protocol for ongoing elimination efforts in Panama and Mesoamerica.

Epidemiology of Plasmodium vivax in Panama during 2017-2020
During the period between 2017-2020 the highest P. vivax incidence occurred in individuals living in the indigenous comarcas and the provinces of Panama and Darien (Fig. 1a), with most of the cases (> 70%) reported occurring in subjects less than 29 years old (Fig. 1b) and of Amerindian ethnicity (Fig. 1c). The epidemiological curve for the year of 2017 shows cases peaking in February during the middle of the dry season (> 90 cases) and again in December at the beginning of the next dry season (> 75 cases). A similar pattern can be observed in 2018. In contrast, during 2019 the number of cases increased to 1646 (plus 43% compared to 2018) with a peak of more than 300 cases reported during dry season in February/March and a similar trend in 2020, suggesting the occurrence of an epidemic outbreak during these two years (Fig. 1d).
Demographic characteristics of the study population Study participants comprised of volunteers that were residents of the provinces of Darien, Panama, Veraguas and the Indigenous Comarca of Guna Yala (Fig. 2a). In total 73 participants were enrolled in the study: 45 subjects with P. vivax based on Giemsa smear positivity, 8 with P. falciparum and 16 malaria smear negative controls. Four subjects were later excluded from the study due to insu cient blood sample volume. 64 % of the participants were male and 36 % female, with a median age of 24 for male (range: 0.5 to 76) and 21 for female (range: 0.5, 53) years. Most participants (42%) were Amerindians and all combined, 93% were residents of the provinces of Darien (26 %), Panama (40 %), and the Comarca Guna Yala (27 %). Including all age groups, 26 % of the participants were unemployed at the time of the survey, 14 % illiterate, and 44 % lived in type 2 and 3 houses as de ned previously (4), with 6 (range: 1, 18) dwellers on average per household (Supplementary Tables  S4 and S5).

Parasite characteristics by light microscopy
To determine the proportion of asexual and sexual stages in the eld samples, we examined Giemsa thin blood smears from each study participant. Representative images of P. vivax asexual stages, including rings, trophozoites, schizonts and gametocytes are shown in Fig. 2b. All stages were detected at similar prevalence (rings: 76%, trophozoites: 88%, schizonts: 76%) except for the less abundant gametocytes (62%). As previously reported schizont and gametocyte stages are present at signi cantly lower levels in the peripheral blood than rings and trophozoites, presumably due to their tissue enrichment (14) ( Figs. 2c,d).
Sample processing and molecular detection of P. vivax by qRT-PCR To optimize the blood volume and processing of eld samples for parasite stage analysis by qRT-PCR, we designed an experiment simulating eld conditions prior to the start of this study. For this purpose, we ampli ed the reference strain P. vivax SAL-1 in the Aotus non-human primate model. Fifteen days after infection, when parasitemia reached 51,080 parasites x µL, blood was collected. A total volume of 60 or 120 µL of P. vivax-infected blood, respectively, was preserved in RNAp for eight days at ambient temperature (~ 27 degrees Celsius), or frozen immediately at -80 degrees Celsius. Samples across conditions were then processed for RNA isolation and subsequent cDNA synthesis and qRT-PCR. Comparison using ANOVA revealed no statistically signi cant differences across conditions using the 3 markers (Pvx18s rRNA, Pvs25, PvLAP5) ( Table  1). For this eld study, we therefore decided to collect 60 µL of sample and store in RNAp media at room temperature for up to 8 days before processing.
Molecular assays were performed on thesubset of 45 P. vivax smear positive eld cases (Supplementary Table S6). The assay detection rate for all P. vivax parasites using Pv18SrRNA was 44/45 (98 %), and for sexual stages 41/45 (91 %) for both Pvs25 and PvLAP5 (Fig.3a). In contrast, microscopic examination only detected gametocytes in 26/42 (62 %) of available smears. This represents a 47 % increase in the detection rate of gametocytes by qRT-PCR assays over microscopy. Comparison of relative transcript expression for PvLAP5vsPvs25 revealed signi cantly higher relative expression of PvLAP5 (Fig.3b,c). The analytical sensitivity of the qRT-PCR assays had been established previously (14). In this study the clinical limit of detection (LOD) of the qRT-PCR assay was established by limiting dilution of clinical samples where we had determined parasite stage concentration (see Fig.2d). PvLAP5 was detected at a minimum concentration of 1.44 gametocytes x µL and Pvs25 at 0.144 gametocytes x µL (Fig.3d). Hence, the PvLAP5 and Pvs25qRT-PCR assays are estimated to be 5-50 fold more sensitive than the theoretical qRT-PCR LOD that was previously established at 9.6 gametocytes x µL (39).
To further compare the detection rate of the qRT-PCR assays to the gold standard of gametocyte detection (microscopy), we analysed the association of study variables age, sample days in transit to the laboratory, whole RNA concentration, mean % parasitemia and marker Ct values with microscopy detection of gametocytes (gametocyte positive = 1 and negative = 0). A statistically signi cant association between categories was detected for mean parasitemia % (p = 0.03) and all qRT-PCR markers (PvLAP5: p = 0.02, Pvs25: p = 0.007; Pv18SrRNA: p = 0.02), but not for any other variable (Table 2 and Fig. 3e). Multivariate analysis demonstrated strong positive correlation between gametocyte markers PvLAP5 and Pvs25 (r = 0.8507; p < 0.001) and between Pv18SrRNA and PvLAP5 (r = 0.5034; p < 0.001) and Pvs25 (r = 0.8891; p = < 0.001) ( Table 3). To assess the validity of the qRT-PCR assays at detecting P. vivax in eld samples preserved in RNAp at ambient temperature, we determined sensitivity (Se), speci city (Sp) as well as positive and negative likelihood ratios (PLR, NLR) and predictive values (PPV, NPV) using microscopy as the gold standard. Indeed, all qRT-PCR assays have high Se and Sp (80% or greater), with PvLAP5 having slightly higher probability of detecting a true positive gametocyte sample compared to Pvs25 (Table 4  and supplementary Tables S7-9).

Discussion
As malaria continues to decline (10), elimination from endemic foci where residual transmission persists is a constant challenge for countries on the verge of elimination by WHO standards (8). To closely monitor advances towards elimination, it is important to maintain a robust malaria molecular epidemiological surveillance program, especially in remote areas that lack the infrastructure to maintain a cold chain. Previous genomic studies have found extremely high clonality in the Panamanian P. vivax population, with clonal lineage 1 (CL1) persisting for at least the past 10 years and CL2 circulating mainly in the Darien province (19). In this study, we deployed and eld validated molecular tools for detection of P. vivax using eld samples preserved and transported at ambient temperature from remote areas of Panama.
We observed an increase in malaria cases during the period 2019-2020 with respect to the previous two years, suggesting an epidemic outbreak during the dry season. Reasons for this outbreak are currently unclear but extreme weather conditions (Tropical storms and hurricanes that impacted Central America and the Caribbean during the 2019-2020) (40), or prolonged con nement for the control of the COVID-19 pandemic might have contributed to it.
Hurricanes (41) and other extreme weather events such as the "El Niño Southern Oscillation (ENSO)", which was particularly strong during 2018-2019 in the region (42), have been associated with changes in malaria transmission in Panama and the Caribbean (43,44). However, we cannot rule out other factors such as increased detection to the introduction of Malaria RDTs in 2017 by MINSA (1) (see supplementary Table S10), changes in vectorial transmission e ciency (45,46), reintroduction of parasites (19,33), waning immunity due to lack of exposure (47) and socioeconomic factors that affect these communities (4).
P. vivax qRT-PCR assays based on detection of ribosomal RNA from low blood volume eld samples stored without cold chain have previously been validated for molecular epidemiological studies (18). The method takes advantage of abundant 18SrRNA transcripts present in blood stage parasites circulating in peripheral blood. Similar approaches for P. vivax gametocytes detection by qRT-PCR using Pvs25 has been described elsewhere (25,48). P. vivax gametocytes represent a small fraction of the total parasite mass found in an infected individual, especially in asymptomatic patients with generally low parasite load. Therefore, detection by conventional microscopy has limited application for P. vivax gametocytes. We have previously demonstrated that PvLAP5 detects P. vivax gametocytes in infected non-human primates, both using speci c antibodies and by qRT-PCR. Here, we deployed a qRT-PCR assay for the detection of PvLAP5, the gametocyte gold standard Pvs25 and Pv18s rRNA using low blood volume and sample storage at ambient temperature. Unlike Pvs25, PvLAP5 qRT-PCR detection uses exon-spanning primers thereby minimising spurious ampli cation from DNA. We validated the usefulness of the assay as a molecular epidemiological tool for the determination of the hidden transmission reservoir in individuals living in remote areas of Panama (14). Results of our study indicate that 60 µL of blood obtained by nger prick and preserved in RNAp at ambient temperature provided similar qRT-PCR results compared to controls stored at -80 degrees Celsius.
Both PvLAP5 and Pvs25 qRT-PCR assays showed at least 57% improvement for detection of gametocytes over light microscopy in eld samples. Markers show similarly high sensitivity and speci city, con rming their suitability as gametocyte markers for molecular epidemiological surveys (38). Our assay system can be used as a screening tool to determine the P. vivax transmission reservoir in asymptomatic carriers and low transmission settings and to maintain a robust molecular epidemiological surveillance program in remote areas.