Dionne C. Argyropoulos

1 Summary: The persistence of Plasmodium vivax is driven by the hidden reservoirs of infection, presenting a key obstacle to elimination. Antibodies persist after asexual infections are cleared from peripheral blood and therefore can indicate current and recent past infections. Here, we present a machine learning algorithm that classifies recent P. vivax infections using serological markers to identify likely hypnozoite carriers. Using serological measurements from year-long observational cohort studies conducted in three low-transmission settings (including negative controls, N =2,635), we selected optimal subsets of markers by balancing sero-diagnostic performance against assay complexity and scalability. We initially trained a random forest classifier and then subsequently we compared several machine learning classifiers. Tree-based methods consistently performed best, although differences were marginal. An online R Shiny application (PvSeroApp) was developed to automate data processing, quality control, and serostatus classification. This algorithm underpins the P. vivax serological testing and treatment (PvSeroTAT) strategy, enabling targeted anti-hypnozoite therapy and strengthening elimination efforts.

Data wrangling and cleaning, quality control checks and implementation of machine learning classification algorithm.

Plasmodium vivax presents a significant obstacle to malaria elimination due to its capacity to form dormant liver-stage hypnozoites that can cause relapses. Universal radical cure, which administers hypnozoite-targeting treatment to patients with P. falciparum malaria living in co-endemic areas, has potential to reduce P. vivax relapses. However, its implementation is hindered by the lack of a diagnostic tool for detecting hypnozoite-carriage. P. vivax serological exposure markers (SEMs) were evaluated as a screening tool in P. falciparum patients for predicting risk of P. vivax microscopy-detected episodes over the following 63 days. Analysis was performed using samples from participants in the PRIMA study from Ethiopia, Indonesia, and Bangladesh (NCT03916003). An existing random forest serological classification algorithm was used and evaluated for sensitivity and specificity, then further optimized by re-training a study-specific model. IgG antibodies were measured in 244 P. falciparum participant samples, of which 22 had a vivax microscopy-detected episode during follow-up. SEMs showed high sensitivity (82%) but low specificity (27%), likely due to sustained antibody responses in moderate-to-high transmission areas or undetected sub-microscopic P. vivax infections during follow-up. A study-specific algorithm increased specificity to 68%, but with a corresponding drop in sensitivity to 68%. Although P. vivax SEMs showed limited suitability for guiding radical cure in P. falciparum patients in this study (where the outcome was microscopy-detected vivax episodes), they could play a valuable role in building community trust and acceptance of universal radical cure when supported by strong communication and implementation strategies.

Here, we introduce a new endpoint ‘census population size’ to evaluate the epidemiology and control of Plasmodium falciparum infections, where the parasite, rather than the infected human host, is the unit of measurement. To calculate census population size, we rely on a definition of parasite variation known as multiplicity of infection (MOI _var ), based on the hyper-diversity of the var multigene family. We present a Bayesian approach to estimate MOI _var from sequencing and counting the number of unique DBLα tags (or DBLα types) of var genes, and derive from it census population size by summation of MOI _var in the human population. We track changes in this parasite population size and structure through sequential malaria interventions by indoor residual spraying (IRS) and seasonal malaria chemoprevention (SMC) from 2012 to 2017 in an area of high, seasonal malaria transmission in northern Ghana. Following IRS, which reduced transmission intensity by >90% and decreased parasite prevalence by ~40–50%, significant reductions in var diversity, MOI _var , and population size were observed in ~2000 humans across all ages. These changes, consistent with the loss of diverse parasite genomes, were short-lived and 32 months after IRS was discontinued and SMC was introduced, var diversity and population size rebounded in all age groups except for the younger children (1–5 years) targeted by SMC. Despite major perturbations from IRS and SMC interventions, the parasite population remained very large and retained the var population genetic characteristics of a high-transmission system (high var diversity; low var repertoire similarity), demonstrating the resilience of P. falciparum to short-term interventions in high-burden countries of sub-Saharan Africa.

Current interventions targeting malaria control in sub-Saharan Africa (SSA) are focused on Plasmodium falciparum , the most prevalent species infecting humans. Despite renewed efforts for malaria elimination in SSA, little attention has been paid to the neglected parasites P. malariae and P. ovale spp. and the impact of interventions like long-lasting insecticidal nets (LLINs), indoor residual spraying (IRS) with non-pyrethroid insecticides, and/or seasonal malaria chemoprevention (SMC) on these minor Plasmodium spp. To address this research gap, this study was undertaken to assess the efficacy of two sequential interventions, IRS and SMC combined with LLINs, on minor- Plasmodium spp. infections in Bongo District, an area characterized by high seasonal transmission in the Northern Sahelian belt of Ghana. Using an interrupted time-series study, five age-stratified surveys, each of ∼2,000 participants, were undertaken at the end of the wet seasons between 2012 and 2022. Across this 10-year study period, infections with P. malariae and P. ovale spp. were detected using a species-specific PCR targeting the 18S rRNA gene. In 2015, following IRS, the prevalence of the minor Plasmodium spp. declined in all ages from baseline in 2012, with participants being significantly less likely to be infected with P. malariae (13.7% vs 1.4%) and P. ovale spp. (5.7% vs 0.4%). Despite this decline, in 2017, 32 months after IRS was discontinued and SMC was introduced, the prevalence of P. malariae (2.9%) and P. ovale spp. (4.0%) rebounded 2- and 10-fold, respectively. This rebound in the minor species was observed in all age groups, except for the younger children (< 5 years) targeted by SMC. Finally, when we examined this population in 2020 and 2022 after sustained deployment of SMC, the prevalence of P. malariae continued to increase (7.4% and 5.8%), while the prevalence of P. ovale spp. declined (2.6% and 1.3%). Results show that both IRS and SMC were effective not only against P. falciparum but also reduced the prevalence of P. malariae and P. ovale spp. in Bongo District. Going forward, molecular diagnostics will be critical to identify changes in the submicroscopic reservoir of the minor Plasmodium spp. found in adolescents and adults and to achieve malaria elimination in this region of Ghana. AUTHOR SUMMARY: To achieve malaria elimination, the use of molecular diagnostics for surveillance of the Plasmodium spp. reservoir is essential. Such diagnostics are critical to detect hidden reservoirs of submicroscopic minor-species infections with P. malariae and P. ovale spp. Here we examine the impact of two sequential malaria control interventions, indoor residual spraying (IRS) with non-pyrethroid insecticides and seasonal malaria chemoprevention (SMC) with sulfadoxine-pyrimethamine plus amodiaquine administered to all children < 5 years, combined with long-lasting insecticidal nets (LLINs) impregnated with pyrethroids to reduce the reservoir of Plasmodium spp. infections where P. falciparum is dominant and the minor species, P. malariae and P. ovale spp., were also found. This interrupted time-series study, undertaken over 10 years (2012-2022) in Bongo District, located in the Northern Sahelian belt of Ghana, identified that IRS and SMC reduced the prevalence of all Plasmodium spp. IRS significantly reduced the prevalence of single- and mixed- Plasmodium spp. infections across all age groups. However, following its discontinuation a rebound was observed for P. falciparum and minor-species infections, especially in adolescents and adults. Although SMC maintained a near-zero prevalence of all Plasmodium spp. infections, its impact was limited to the target age group (< 5 years). Furthermore, this study showed that following the discontinuation of IRS, the older children, adolescents, and adults, not eligible for SMC, continued to harbour Plasmodium spp. infections, including minor species, thus sustaining the reservoir. These results show that by combining vector control inventions as well as targeting all age groups with chemoprevention, high-burden countries in sub-Saharan Africa, including Ghana, could bring the Plasmodium spp. reservoir closer to elimination more rapidly.

A major motivation for developing molecular methods for malaria surveillance is to measure the impact of control interventions on the population genetics of Plasmodium falciparum as a potential marker of progress towards elimination. Here we assess three established methods (i) single nucleotide polymorphism (SNP) barcoding (panel of 24-biallelic loci), (ii) microsatellite genotyping (panel of 12-multiallelic loci), and (iii) varcoding (fingerprinting var gene diversity, akin to microhaplotyping) to identify changes in parasite population genetics in response to a short-term indoor residual spraying (IRS) intervention. Typical of high seasonal transmission in Africa, multiclonal infections were found in 82.3% (median 3; range 1-18) and 57.8% (median 2; range 1-12) of asymptomatic individuals pre- and post-IRS, respectively, in Bongo District, Ghana. Since directly phasing multilocus haplotypes for population genetic analysis is not possible for biallelic SNPs and microsatellites, we chose ~200 low-complexity infections biased to single and double clone infections for analysis. Each genotyping method presented a different pattern of change in diversity and population structure as a consequence of variability in usable data and the relative polymorphism of the molecular markers (i.e., SNPs < microsatellites < var). Varcoding and microsatellite genotyping showed the overall failure of the IRS intervention to significantly change the population structure from pre-IRS characteristics (i.e., many diverse genomes of low genetic similarity). The 24-SNP barcode provided limited information for analysis, largely due to the biallelic nature of SNPs leading to a high proportion of double-allele calls and a view of more isolate relatedness compared to microsatellites and varcoding. Relative performance, suitability, and cost-effectiveness of the methods relevant to sample size and local malaria elimination in high-transmission endemic areas are discussed.

At a time when effective tools for monitoring malaria control and eradication efforts are crucial, the increasing availability of molecular data motivates their application to epidemiology. The multiplicity of infection (MOI), defined as the number of genetically distinct parasite strains co-infecting a host, is one key epidemiological parameter for evaluating malaria interventions. Estimating MOI remains a challenge for high-transmission settings where individuals typically carry multiple co-occurring infections. Several quantitative approaches have been developed to estimate MOI, including two cost-effective ones relying on molecular data: i) THE REAL McCOIL method is based on putatively neutral single nucleotide polymorphism loci, and ii) the varcoding method is a fingerprinting approach that relies on the diversity and limited repertoire overlap of the var multigene family encoding the major Plasmodium falciparum blood-stage antigen PfEMP1 and is therefore under selection. In this study, we assess the robustness of the MOI estimates generated with these two approaches by simulating P. falciparum malaria dynamics under three transmission conditions using an extension of a previously developed stochastic agent-based model. We demonstrate that these approaches are complementary and best considered across distinct transmission intensities. While varcoding can underestimate MOI, it allows robust estimation, especially under high transmission where repertoire overlap is extremely limited from frequency-dependent selection. In contrast, THE REAL McCOIL often considerably overestimates MOI, but still provides reasonable estimates for low and moderate transmission. Regardless of transmission intensity, results for THE REAL McCOIL indicate that an inaccurate tail at high MOI values is generated, and that at high transmission, an apparently reasonable estimated MOI distribution can arise from some degree of compensation between overestimation and underestimation. As many countries pursue malaria elimination targets, defining the most suitable approach to estimate MOI based on sample size and local transmission intensity is highly recommended for monitoring the impact of intervention programs.

At a time when effective tools for monitoring malaria control and eradication efforts are crucial, the increasing availability of molecular data motivates their application to epidemiology. The multiplicity of infection (MOI), defined as the number of genetically distinct parasite strains co-infecting a host, is one key epidemiological parameter for evaluating malaria interventions. Estimating MOI remains a challenge for high-transmission settings where individuals typically carry multiple co-occurring infections. Several quantitative approaches have been developed to estimate MOI, including two cost-effective ones relying on molecular data: i) THE REAL McCOIL method is based on putatively neutral single nucleotide polymorphism loci, and ii) the var coding method is a fingerprinting approach that relies on the diversity and limited repertoire overlap of the var multigene family encoding the major Plasmodium falciparum blood-stage antigen PfEMP1 and is therefore under selection. In this study, we assess the robustness of the MOI estimates generated with these two approaches by simulating P. falciparum malaria dynamics under three transmission conditions using an extension of a previously developed stochastic agent-based model. We demonstrate that these approaches are complementary and best considered across distinct transmission intensities. While var coding can underestimate MOI, it allows robust estimation, especially under high-transmission where repertoire overlap is extremely limited from frequency-dependent selection. In contrast, THE REAL McCOIL often considerably overestimates MOI, but still provides reasonable estimates for low- and moderate-transmission. As many countries pursue malaria elimination targets, defining the most suitable approach to estimate MOI based on sample size and local transmission intensity is highly recommended for monitoring the impact of intervention programs. Author Summary: Despite control and elimination efforts, malaria continues to be a serious public health threat especially in high-transmission regions. Molecular tools for evaluating these efforts include those seeking to estimate multiplicity (or complexity) of infection (MOI), the number of genetically distinct parasite strains co-infecting a host, a key epidemiological parameter. MOI estimation remains challenging in high-transmission regions where hosts typically carry multiple co-infections by Plasmodium falciparum . THE REAL McCOIL and the var coding are two cost-effective methods relying on distinct parts of the parasite genome, those respectively under neutrality and selection. The more recent var coding approach relies on the var multigene family encoding for the major blood-stage antigen and contributing to a complex immune evasion strategy of the parasite. We compare the performance of the two methods by simulating disease dynamics under different transmission intensities with a stochastic agent-based model tracking infection by different parasite genomes and immune memory in individual hosts, then sampling resulting infections to estimate MOI. Although THE REAL McCOIL provides reasonable estimates for low- and moderate-transmission, var coding allows more robust estimates especially under high-transmission. Defining the most suitable approach to estimate MOI based on local transmission intensity is highly recommended for hyper-diverse pathogens such as malaria.

High-malaria burden countries in sub-Saharan Africa are shifting from malaria control towards elimination. Hence, there is need to gain a contemporary understanding of how indoor residual spraying (IRS) with non-pyrethroid insecticides when combined with long-lasting insecticidal nets (LLINs) impregnated with pyrethroid insecticides, contribute to the efforts of National Malaria Control Programmes to interrupt transmission and reduce the reservoir of Plasmodium falciparum infections across all ages. Using an interrupted time-series study design, four age-stratified malariometric surveys, each of ~2,000 participants, were undertaken pre- and post-IRS in Bongo District, Ghana. Following the application of three-rounds of IRS, P. falciparum transmission intensity declined, as measured by a >90% reduction in the monthly entomological inoculation rate. This decline was accompanied by reductions in parasitological parameters, with participants of all ages being significantly less likely to harbor P. falciparum infections at the end of the wet season post-IRS (aOR = 0.22 [95% CI: 0.19–0.26], p-value < 0.001). In addition, multiplicity of infection (MOI_var) was measured using a parasite fingerprinting tool, designed to capture within-host genome diversity. At the end of the wet season post-IRS, the prevalence of multi-genome infections declined from 75.6% to 54.1%. This study demonstrates that in areas characterized by high seasonal malaria transmission, IRS in combination with LLINs can significantly reduce the reservoir of P. falciparum infection. Nonetheless despite this success, 41.6% of the population, especially older children and adolescents, still harboured multi-genome infections. Given the persistence of this diverse reservoir across all ages, these data highlight the importance of sustaining vector control in combination with targeted chemotherapy to move high-transmission settings towards pre-elimination. This study also points to the benefits of molecular surveillance to ensure that incremental achievements are not lost and that the goals advocated for in the WHO’s High Burden to High Impact strategy are realized.

Here, we report the first population genetic study to examine the impact of indoor residual spraying (IRS) on Plasmodium falciparum in humans. This study was conducted in an area of high seasonal malaria transmission in Bongo District, Ghana. IRS was implemented during the dry season (November–May) in three consecutive years between 2013 and 2015 to reduce transmission and attempt to bottleneck the parasite population in humans towards lower diversity with greater linkage disequilibrium. The study was done against a background of widespread use of long‐lasting insecticidal nets, typical for contemporary malaria control in West Africa. Microsatellite genotyping with 10 loci was used to construct 392 P. falciparum multilocus infection haplotypes collected from two age‐stratified cross‐sectional surveys at the end of the wet seasons pre‐ and post‐IRS. Three‐rounds of IRS, under operational conditions, led to a >90% reduction in transmission intensity and a 35.7% reduction in the P. falciparum prevalence (p < .001). Despite these declines, population genetic analysis of the infection haplotypes revealed no dramatic changes with only a slight, but significant increase in genetic diversity (H_e: pre‐IRS = 0.79 vs. post‐IRS = 0.81, p = .048). Reduced relatedness of the parasite population (p < .001) was observed post‐IRS, probably due to decreased opportunities for outcrossing. Spatiotemporal genetic differentiation between the pre‐ and post‐IRS surveys (D = 0.0329 [95% CI: 0.0209 – 0.0473], p = .034) was identified. These data provide a genetic explanation for the resilience of P. falciparum to short‐term IRS programmes in high‐transmission settings in sub‐Saharan Africa.