Advertisement
Research Article

A High Force of Plasmodium vivax Blood-Stage Infection Drives the Rapid Acquisition of Immunity in Papua New Guinean Children

  • Cristian Koepfli,

    Affiliations: Swiss Tropical and Public Health Institute, Basel, Switzerland, University of Basel, Basel, Switzerland, Walter and Eliza Hall Institute, Parkville, Victoria, Australia

    X
  • Kathryn L. Colborn,

    Affiliations: Walter and Eliza Hall Institute, Parkville, Victoria, Australia, University of California, Berkeley, Department of Biostatistics, Berkeley, California, United States of America

    X
  • Benson Kiniboro,

    Affiliation: Papua New Guinea Institute of Medical Research, Goroka, Eastern Highland Province, Papua New Guinea

    X
  • Enmoore Lin,

    Affiliation: Papua New Guinea Institute of Medical Research, Goroka, Eastern Highland Province, Papua New Guinea

    X
  • Terence P. Speed,

    Affiliation: Walter and Eliza Hall Institute, Parkville, Victoria, Australia

    X
  • Peter M. Siba,

    Affiliation: Papua New Guinea Institute of Medical Research, Goroka, Eastern Highland Province, Papua New Guinea

    X
  • Ingrid Felger mail,

    ingrid.felger@unibas.ch (IF); ivomueller@fastmail.fm (IM)

    Affiliations: Swiss Tropical and Public Health Institute, Basel, Switzerland, University of Basel, Basel, Switzerland

    X
  • Ivo Mueller mail

    ingrid.felger@unibas.ch (IF); ivomueller@fastmail.fm (IM)

    Affiliations: Walter and Eliza Hall Institute, Parkville, Victoria, Australia, Papua New Guinea Institute of Medical Research, Goroka, Eastern Highland Province, Papua New Guinea, Barcelona Centre for International Health Research, Barcelona, Spain

    X
  • Published: September 05, 2013
  • DOI: 10.1371/journal.pntd.0002403

Abstract

Background

When both parasite species are co-endemic, Plasmodium vivax incidence peaks in younger children compared to P. falciparum. To identify differences in the number of blood stage infections of these species and its potential link to acquisition of immunity, we have estimated the molecular force of blood-stage infection of P. vivax (molFOB, i.e. the number of genetically distinct blood-stage infections over time), and compared it to previously reported values for P. falciparum.

Methods

P. vivax molFOB was estimated by high resolution genotyping parasites in samples collected over 16 months in a cohort of 264 Papua New Guinean children living in an area highly endemic for P. falciparum and P. vivax. In this cohort, P. vivax episodes decreased three-fold over the age range of 1–4.5 years.

Results

On average, children acquired 14.0 new P. vivax blood-stage clones/child/year-at-risk. While the incidence of clinical P. vivax illness was strongly associated with molFOB (incidence rate ratio (IRR) = 1.99, 95% confidence interval (CI95) [1.80, 2.19]), molFOB did not change with age. The incidence of P. vivax showed a faster decrease with age in children with high (IRR = 0.49, CI95 [0.38, 0.64] p<0.001) compared to those with low exposure (IRR = 0.63, CI95[0.43, 0.93] p = 0.02).

Conclusion

P. vivax molFOB is considerably higher than P. falciparum molFOB (5.5 clones/child/year-at-risk). The high number of P. vivax clones that infect children in early childhood contribute to the rapid acquisition of immunity against clinical P. vivax malaria.

Author Summary

In areas where P. vivax and P. falciparum parasite species co-occur, immunity to P. vivax seems to be acquired more rapidly. This difference could be caused either by generic differences in the way immunity is acquired or by a relatively higher exposure to P. vivax blood-stage infections in early life. We found that children experienced an average of 14 new P. vivax blood-stage infections per year, and that the number of new infections acquired predicted how often children fell ill with vivax malaria by genotyping all P. vivax infections that occurred in a group of 264 children 1–4 years of age followed for 16 months. The burden of blood-stage infections caused by P. vivax was therefore at least twice as high as that caused by P. falciparum. This higher force-of-blood-stage infection (molFOB) caused by P. vivax is at least partially due to the ability of P. vivax hypnozoites to relapse from long-lasting liver stages. A high exposure to P. vivax blood-stage infection resulted in more rapid decrease in the incidence of P. vivax malaria. The high number of P. vivax clones that infect children in early childhood is thus likely to contribute substantially to the rapid acquisition of immunity against clinical P. vivax malaria.

Introduction

Exposure to malaria – i.e. the number of blood-stage infections acquired over time – determines to a large extent the frequency of disease and can also explain seasonal patterns and age trends. The number of infections acquired over time depends on transmission intensity, with frequent bites by infected mosquitoes resulting in high numbers of infections in a short period of time. In P. vivax, additional blood-stage infections are caused by relapses from dormant hypnozoites.

As people who live in malaria endemic areas achieve immunity to disease after several years of exposure, newly acquired infections do not always result in clinical episodes. The speed of acquisition of immunity depends on transmission intensity [1][3] and also differs between parasite species, with immunity to P. vivax appearing to be acquired faster than immunity to P. falciparum [4], [5]. In numerous field studies conducted in areas co-endemic for both species, the burden of P. vivax infections and disease was found to peak at a younger age than that of P. falciparum [6][10].

The mechanisms underlying the immunity to malaria are not entirely understood [2], [11]. Whereas semi-immune people still develop blood stage parasitemia, parasite densities are considerably lower and rarely cause fever. Differences in the rate of natural acquisition of immunity to P. falciparum and P. vivax may be the result of differences in the immune responses induced by either species, and/or a consequence of different numbers of infections acquired over time.

Malaria infections consist of different parasite clones that can infect individuals successively or simultaneously as multiple clone infections. Clones differ in their polymorphic surface antigens as well as in neutral genetic markers. Genotyping of such markers allows discrimination of individual parasite clones during multiple clone infections. The diversity of antigens is well documented in P. falciparum [12][14] and P. vivax [15], [16]. The large number of antigen variants is thought to allow parasites to escape the immune system, as new infections likely express alleles different from previous clones, and thus are not recognized by the present humoral response [17], [18]. Immunity is thus assumed to be largely clone-specific, providing little cross-protection against heterologous clones [4], [19][22].

Plasmodium falciparum clones transmitted by a mosquito bite appear in the blood stream within 7–10 days. Albeit some of the transmitted sporozoites might not establish blood-stage infections, the number of genetically distinct clones detected in the blood-stream over time (molFOB, as determined by genotyping) is a direct measure of the molecular force of infection (molFOI, defined as number of distinct clones entering the body over time). It is thus closely linked to intensity of transmission [23].

In tropical and sub-tropical areas, P. vivax infectious mosquito bites also lead to primary blood stage infections within 10–14 days. In addition, a proportion of parasites remain dormant in the liver. Relapses from such hypnozoites can then occur months later and result in delayed blood stage infections. In high transmission regions, such relapsing parasites are mostly genetically different from parasites detected during the last acute blood stage infection [24], [25]. The relapsing clones might have been present earlier as blood stage infection [26], or the primary and relapsing infections are genetically different clones that were jointly transmitted in a single mosquito bite with only one of these clones having emerged from the liver. Thus, relapses may not only boost existing immune responses by repeated exposure to the same parasite clone, but also lead to a broader immune repertoire. No genotyping approach allows differentiating a relapse from a primary blood stage infection. The number of P. vivax parasites detected in the peripheral blood in a given interval (molFOB) is therefore a combination of primary blood-stage infections and relapses from the same and/or earlier mosquito bites.

In order to assess the association between (individual) exposure and risk of malaria, we followed 264 children aged 1 to 3 years at enrolment in an area of high endemicity of P. falciparum and P. vivax in Papua New Guinea (PNG) over 16 months [23], [27], [28]. As in earlier studies in PNG [6], [10], [29], P. vivax incidence in this cohort peaked in younger children compared to P. falciparum incidence [27]. While P. vivax incidence decreased throughout the age group, P. falciparum incidence increased between the ages of 1 to 3.5 years with little change thereafter [27]. Previously we had estimated the P. falciparum molFOI by msp2 genotyping, and shown that children acquired 5.9 new P. falciparum infections per year-at-risk [23]. P. falciparum molFOI increased significantly with age and was highly predictive of incidence patterns [23].

The diversity and multiplicity of P. vivax in this cohort was published previously [28]. In this paper, we present estimates of the P. vivax molecular force of blood-stage infection (molFOB).

Methods

Ethics statement

Informed written consent was obtained from all parents or guardians prior to recruitment of each child. Scientific approval and ethical clearance for the study was obtained from the Medical Research and Advisory Committee (MRAC 05.19 and 09.24) of the Ministry of Health in PNG and from the Ethikkommission beider Basel in Switzerland (no 03/06).

Field survey and patients

This study was conducted in Ilaita, a rural area near Maprik, East Sepik Province, Papua New Guinea. A detailed description of the study was given elsewhere [27]. Briefly, 264 study participants were enrolled at an age of 10 to 38 months between March and September 2006, and followed actively every 2 weeks to determine malaria morbidity for a period of up to 16 months (until July 2007). In addition, children were actively checked every 8 to 9 weeks for the presence of malarial infections. Except for the first and last round of active case detection, two consecutive blood samples were collected by finger prick 24 hours apart from each study participant at each follow-up visit. An individual thus contributed up to 16 samples, 14 of which were paired samples collected 24 hours apart. A passive case detection system was maintained at the local health center and aid post throughout the entire study period. At each episode of febrile illness, a blood sample was collected from the participant and a rapid diagnostic test (RDT) was performed and haemoglobin measured. Antimalarial treatment with arthemeter-lumefantrine (AL) or in a few cases with amodiaquine plus sulphadoxine-pyrimethamine was administered upon a positive RDT or if haemoglobin levels were less than 7.5 g/dl. In children with negative RDT, blood slides were read within 24 hours, and microscopy positive children were treated with AL.

Laboratory procedures

For genotyping individual P. vivax clones, the molecular markers msp1F3 and MS16 were typed using capillary electrophoresis for highly precise fragment sizing, which is a precondition for longitudinal follow up of individual parasite clones. Both markers proved to be highly polymorphic in the cohort with an expected heterozygosity (He) of 97.8% for MS16 and 88.1% for msp1F3. Details of the genotyping have been described previously [28].

Data analysis

In a previous analysis of the samples collected 24 hours apart, we found that not all clones present in a host were detected within a single sample. Twenty-one percent of all msp1F3 alleles and 28% of all MS16 clones were missed on a single day [30]. Thus, we used the combined genotyping data from both day 1 and day 2, except for samples from enrolment and final visits, where only 1 venous blood sample was taken.

The force of new P. vivax blood-stage infections (molFOB) was generated by counting the number of genotypes in each interval that had not been present in the preceding interval. An 8 to 9 weeks interval started on the first day after a regular cross-sectional visit and included all samples collected during passive case detection over two months plus the samples collected at the end of the interval. molFOB was also determined for both markers combined, msp1F3 and MS16. In case of discrepancy between the markers for an 8-weeks interval the higher estimate from either marker was used. This approach corrected for imperfect resolution and detectability of a single marker.

Genotyping cannot directly identify relapses; molFOB measures the combination of primary blood-stage infections and those caused by relapsing hypnozoites. Thus, homologous relapses occurring in two subsequent 2-month intervals would be misclassified as persisting clones. New Guinean P. vivax strains are known to relapse rapidly [31], however in regions of high transmission, relapsing clones are usually of a different genotype than the initial blood stage infection [24]. As a consequence the number of homologous relapses that were not detected is expected to be relatively small.

In line with the pharmacokinetic properties of the drugs [32][34], children were not considered at risk for two weeks after treatment with artemether-lumefantrine and four weeks after treatment with amodiaquine (AQ) plus sulphadoxine-pyramethamine (SP). The force of blood-stage infection for each child and interval was subsequently converted into the number of new clones acquired per year-at-risk.

Similar to previous analyses of P. falciparum molFOI [23], generalized linear mixed models (GLMMs) were used for analyses of force of blood-stage infection as well as for incidence of P. vivax episodes. These models were chosen because they allowed the fixed effects to be specified separately from the random effects (i.e. repeated measurements from the same child over time and unmeasured village factors). Furthermore, the random-effects model allowed for decomposition of the error into between-village and within-village variation.

We fit a Poisson GLMM model with a log link to relate the fixed and (Gaussian) random effects to the number of clinical episodes experienced during a two month interval (defined as febrile illness plus P. vivax >500 parasites/µl). Covariates were selected based on earlier analyses of the same data [27]. Seasonality was characterized by two readily interpretable parameters: the amplitude, which was half the range between the peak and trough, and the phase, which was the location of the first zero crossing in a cycle relative to the origin in time (in this case, the first week of the year). For computational convenience, they were replaced by sine and cosine terms with fixed phases. For all outcomes except prevalence, an offset was fit to adjust for years at risk. Estimation of these models was done using the LME4 package in R version 2.12 [35]. All point estimates provided throughout the text (except those for seasonal effects) were obtained from cubic splines fit using generalized additive models (Figure 1) using the MGCV package in R version 2.12 [36]. For a more detailed description of the statistical approaches see [23]. Point estimates for seasonal peaks and troughs were obtained from the GLMMs by setting all other values of the covariates at their means. For the analyses of the effect of exposure on the relationship between age and incidence of P. vivax malaria, children were stratified into terciles according to the average molFOB during the entire follow-up.

thumbnail

Figure 1. Age and seasonal patterns of molecular force of blood-stage infection (molFOB, Panel A) and incidence of clinical P. vivax malaria (Panel B), and effects of ITN use on these parameters.

Smooth splines from generalized additive models with single predictors and 95% confidence intervals.

doi:10.1371/journal.pntd.0002403.g001

Results

A total of 264 children aged 0.9 to 3.2 years at baseline were enrolled and followed up for 69 weeks. Out of 264 children, 248 (93.9%) were retained until the end of the study with 96.0%–100.0% of children seen at each scheduled two-monthly survey. Over the entire follow-up period, the age ranged from 0.9 to 4.5 years. A detailed description of this cohort was published previously [27]. All but five children had at least one P. vivax msp1F3 or MS16 PCR positive sample during the 16 months of follow-up. Of all samples collected, 51.6% and 52.7% were positive for msp1F3 and MS16, respectively, and 54.8% were positive for either marker. In a total of 1448 P. vivax positive samples, 2,305 and 3,372 distinct clones were detected by msp1F3 (65 different alleles) and MS16 (113 alleles) genotyping, respectively.

Force of blood-stage infection

Excluding any period with residual drug levels from the time at risk, each child was at risk of acquiring new infections for an average of 0.93 years during the cohort (95% confidence interval (CI95) [0.91, 0.96] range: 0.11–1.32). On average, 8.7 P. vivax msp1F3 (CI95 [8.1, 9.4], range: 0–30) and 12.8 MS16 clones (CI95 [11.8, 13.7], range: 0–35) were found per child over the entire follow-up period. When both markers were combined, an average of 14.0 P. vivax clones were observed per child (CI95 [13.0, 15.0], range 0–38).

The average molFOB was 9.4 new P. vivax infections per child per year-at-risk by msp1F3 (CI95 [8.7, 10.0]), 13.8 by MS16 (CI95 [12.8, 14.8]) and 15.1 by both markers combined (CI95 [14.1, 16.2]). All further analyses were done for both markers combined. In addition, children acquired an average of 5.1 different P. falciparum msp2 clones during the cohort (CI95 [4.6, 5.6], range: 0–19), resulting in a corresponding molFOB of 5.5 new P. falciparum infections per child per year-at-risk (CI95 [5.0, 6.1]).

P. vivax molFOB showed a very pronounced seasonality (Figure 1, Table 1, P<0.0001), peaking in early January (week 1, 17.1 clones/year-at-risk) and was lowest in early July (week 27, 11.3 clones/year-at-risk, Figure 2). molFOB was also significantly lower in 2007 compared to 2006 (incidence rate ratio (IRR) 0.84, CI95 [0.77, 0.92], P = 0.0002, Table 1, Figure 1). Regular ITN use was associated with a significant reduction in acquisition of new clones (IRR 0.66, CI95 [0.56, 0.77], P<0.0001, Table 1). Children with antimalarial treatment in the preceding four weeks had a higher molFOB than those that were not treated (IRR 1.24, CI95 [1.15, 1.33], P<0.0001, Table 1). molFOB did not vary significantly with age (Figure 1, P (GLMM) = 0.6). There was significant variation in molFOB between children in a village (P<0.001) but not between villages (P = 0.3, Table 1).

thumbnail

Figure 2. Seasonality of P. vivax clinical episodes (as compared to the first week of January, where incidence peaked) excluding molFOB (blue) and adjusted for molFOB (red).

molFOB accounts for approximately 50% of the seasonal variation in incidence. IRR = incidence rate ratio.

doi:10.1371/journal.pntd.0002403.g002
thumbnail

Table 1. Parameter estimates from GLMMs predicting the molecular force of P. vivax blood-stage infections (molFOB) and the number of incident clinical episodes of P. vivax malaria with density >500 parasites/µl with and without adjustment for molFOB.

doi:10.1371/journal.pntd.0002403.t001

Predictors of clinical P. vivax illness

Over the 69 weeks of follow-up, a total of 1134 febrile episodes with parasitemia of any parasite species and any parasite density by light microscopy were observed, resulting in an incidence rate (IR) of 4.60 episodes/year-at-risk [27]. P. vivax was the second most common cause of malarial illness, causing 605 episodes (IR = 2.46) with any parasite density (supplementary Table 1). Of these, 391 episodes (IR = 1.59) fulfilled the more specific definition of P. vivax malaria (i.e. febrile illness plus P. vivax parasitemia >500 parasites/µl) [37]. P. falciparum caused slightly more clinical episodes (any density: 630 episodes (IR = 2.56); >2,500 parasites/µl: 472 episodes (IR = 1.92)).

As in earlier analyses [27], age and season were significant predictors of clinical episodes of P. vivax malaria (Table 1). The incidence of P. vivax malaria decreased log-linearly with age (Figure 1, Table 1, P<0.0001) from 2.9 episodes/year-at-risk at 1 year of age to a minimum of 0.6 episodes at 3.5 years of age. It peaked at the beginning of the rainy season (early December, week 49, 1.7 clinical episodes, Figure 1) and was lowest in the early dry season (early June, week 23, 0.8 clinical episodes, Figure 1). Insecticide treated net (ITN) use was not significantly associated with incidence of P. vivax malaria. The incidence of P. vivax malaria varied significantly between villages (P<0.0001) and between children living in the same village (P<0.0001).

When molFOB was added to the model (defined as the rate of new clones acquired per year-at-risk), it was highly significantly associated with the incidence of P. vivax malaria (Table 1, P<0.0001). Adjusting for molFOB resulted in a 45% decrease in seasonal differences in P. vivax incidence (Table 1, Figure 2). Contrary to what was observed for P. falciparum [23], adding P. vivax molFOB did not significantly alter the association of age or ITN use with incidence of clinical P. vivax episodes. Comparable results were seen when P. vivax episodes with any parasite density were considered (Table S1).

In order to determine the association of exposure with the rate of immune acquisition we stratified the cohort in children with high (molFOB: 18.1–39.0/year-at-risk), medium (molFOB: 10.7–18.0/year-at-risk) and low exposure (molFOB: 0–10.7/year-at-risk) based on their average P. vivax molFOB. The reduction of incidence of P. vivax episodes >500/ul with age was less pronounced in the one third of children with the lowest average molFOB (IRR = 0.63/year increase in age, p = 0.02) than those with the highest exposure (IRR = 0.49/year increase in age, p<0.001) (Figure 3). Even stronger differences were observed in associations of age with all P. vivax episodes, where no significant reduction in incidence with increasing age was found for children with low exposure (IRR = 0.93, p = 0.6), but strong reductions in children with medium and high exposure (IRR = 0.50–0.52, p<0.001, Figure 3). These differences in IRR for P. vivax episodes of all densities with age were statistically significant for children with low versus medium (p = 0.03) and high exposure (p = 0.02), respectively. Due to overall lower number of episodes with a parasitemia >500 parasites/ul (and thus reduced power), the differences between children with low and medium exposure (p = 0.40) and low and high exposure (p = 0.15) did not reach statistical significance.

thumbnail

Figure 3. Association of incidence of P. vivax episodes with age in children with low (molFOB: 0–10.7), medium (molFOB: 10.7–18.0) and high average exposure (molFOB: 18.1–39.0) to P. vivax infections.

Incidence rate ratio and 95% confidence intervals for changes associated with 1 year increase in age.

doi:10.1371/journal.pntd.0002403.g003

Discussion

By genotyping all blood-stage parasites detected over 16 months of follow-up, this study provides the first direct estimate of the molecular force of P. vivax blood-stage infections. Children aged 0.9 to 4.5 years acquired 14.0 P. vivax clones per year-at-risk, thus approximately twice as many as P. falciparum blood-stage clones [23]. Differences between P. vivax and P. falciparum were not only evident in the absolute number of infections, but also in the associations between molFOB, incidence of clinical episodes and age. While P. falciparum molFOB increased with age and thus paralleled the trend in incidence [23], P. vivax molFOB did not change with age, but incidence of disease decreased dramatically over the age range of the cohort.

Minor differences in the typing techniques applied for these two Plasmodium species could account for some differences in the estimates. Typing was based on length polymorphic marker genes. Their diversity was high, yet slightly differed: P. falciparum was typed using msp2 (expected heterozygosity HE = 0.933) [38], while msp1F3 (HE = 0.881) and MS16 (HE = 0.978) were used for P. vivax typing [28]. For all three markers, we previously determined the clone detectability, i.e. the proportion of clones detected in both of two bleeds collected 24 hours apart. Detectability differed between markers: 79% for Pfmsp2, 61% for PvMS16 and 73% for Pvmsp1F3 [30]. The overall diversity of the Pfmsp2 marker was therefore intermediate to that of the two P. vivax markers, while Pfmsp2 detectability was highest.

The most obvious difference in typing strategies was that our analysis of P. vivax was based on two loci with the combined molFOB determined from maximal number of alleles per sample observed by any marker. Thus, the ability to detect clones of both P. vivax markers combined was higher than by the single P. falciparum marker.

To assess the effect of the use of 2 markers, P. vivax molFOB was also calculated using a single marker only. The average molFOB was 8.7 and 12.8 clones/year-at-risk for msp1F3 and MS16, respectively. Both of these values are substantially higher than the 5.5 clones/year-at-risk detected by Pfmsp2 genotyping. It is worth noting that a 1.6 times higher molFOB was obtained with P. vivax marker msp1F3 compared to P. falciparum despite its lower diversity and detectability. Therefore, the differences in allelic diversity and detectability of genotyping markers can not account for the large differences between P. vivax and P. falciparum molFOB.

Given imperfect detectability and resolution of markers plus infrequent sampling, our estimates of molFOB of both species were likely underestimates of the true burden and complexity of Plasmodium spp. infection. However, for 97.5% (1638/1680) of all regular follow-up bleeds (except baseline and the final round), two paired samples taken 24 hours apart were available, thus the probability of missing a clone in both of two paired samples dropped to 0.07 for P. vivax msp1F3, 0.15 for P. vivax MS16 and 0.044 for P. falciparum msp2 [30] (assuming independence in the chance to detect a clone from any one sample). In addition, clones detected during passive case detection between regular visits were also included, further reducing the chance of missing a clone during an 8-week interval. This suggests that the estimates of molFOB were biased downwards as a result of imperfect detectability only by between 6 and 13% for P. vivax and 4% for P. falciparum.

In PNG, P. vivax and P. falciparum are transmitted by the same mosquito vectors. While no entomological studies were conducted in parallel with this cohort, earlier studies in different PNG lowland populations reported comparable sporozoite rates for P. falciparum and P. vivax [6], [39], [40]. However, the likelihood that individuals become infected with multiple P. vivax clones by a single mosquito bite is high, because 75% of all P. vivax positive individuals from our cohort carried multiple clone infections [28]. It is therefore likely that a mosquito takes up different P. vivax gametocyte clones during a blood-meal, resulting in sexual recombination and the transmission of a genetically diverse inoculum to a new host. This contrasts to only 33% of multi-clone infections in the P. falciparum positive children (Schoepflin, Mueller & Felger unpublished results).

P. vivax parasites detected in the blood stream do not always derive from a recent mosquito bite; they can also result from relapsing hypnozoites. New Guinean P. vivax strains were reported to relapse rapidly, and 63% of 1 to 5 year old children had a recurrent parasitemia within six weeks after treatment of blood-stage parasites [41]. Two thirds of those recurrent infections (all post day 28) were of a different genotype [42]. Although genotyping cannot differentiate between true new infections and relapses of an unrelated genotype, P. vivax relapses likely contribute significantly to the higher molFOB. Evidence for this comes from a recent cohort of Papua New Guinean children aged 1 to 5 years, in which the contribution of relapses to the burden of infection was assessed directly by randomising one third of children to receive anti-hypnozoite drug therapy with a 14 day course of high-dose primaquine. While the primaquine treatment was not 100% efficacious, it did nevertheless result in a 34–57% reduction in the incidence of new P. vivax infections [43].

As a consequence of relapses, molFOB measures different processes in P. vivax and P. falciparum. P. falciparum molFOB is directly linked to transmission intensity in a given interval, whereas molFOB of P. vivax represents a composite measure of both transmission intensity and frequency and genetic complexity of relapsing parasites (some of which may have been acquired months or years earlier). P. vivax molFOB is therefore not a direct measure of molFOI (i.e. the number of new parasites acquired by the human host), but the difference between the two measures is small if the total follow-up period is substantially longer than the average relapse frequency. In the Southwest Pacific, P. vivax infections are thought to relapse very rapidly (i.e. within a few weeks [31]), thus P. vivax molFOB is a good surrogate marker for molFOI if calculated over the entire 16 months of follow-up.

Interestingly, P. vivax molFOB did not change with age, while P. falciparum molFOI was strongly age dependent and increased from 3 to 8 clones per year-at-risk over the age range of 1–4.5 years of this cohort [28]. As a consequence, molFOI largely explained the age trend in P. falciparum incidence (increasing from 1 episode per year in children one year of age to 2.5 in children three years of age) [23]. In contrast, P. vivax molFOB did not explain the decrease in P. vivax incidence from over 3 episodes per year in children one year of age to less than 1 episode in children older than three years.

The age shift in incidence in clinical disease caused by P. vivax and P. falciparum observed in this study – i.e. P. vivax incidence peaking in younger children - parallels earlier findings in Papua New Guinea [6], [10], [29] and other regions where both species are co-endemic [7][9]. This indicates a rapid acquisition of immunity to P. vivax in individuals with life-long exposure to both P. falciparum and P. vivax. As immunity to malaria builds up gradually and is thought to be strain-specific [4], [5], it is likely that the number of distinct infections acquired over an individual's lifetime is a major driving force for acquisition of immunity [20].

In their first years of life, the children in our cohort were estimated to have acquired three times more genetically distinct P. vivax than P. falciparum infections. In both species a higher molFOB was itself associated with a significant increase in incidence of clinical disease. However, a more rapid decrease in incidence of P. vivax malaria was observed in children exposed to high compared to low levels of P. vivax infections (Figure 3). Although notable against P. vivax episodes with >500 parasites/µl, this effect of exposure on age-specific risk of P. vivax malaria was more pronounced when all P. vivax episodes (irrespective of parasitaemia) were considered. While the incidence of P. vivax malaria >500 parasites/µl decreased, even in children with low exposure, the incidence of P. vivax of any density did not change significantly with age in the children with the low exposure. This indicates that, in children with low exposure, those >3 years acquired some immunity against high-density clinical P. vivax episodes, but not against those associated with low levels of parasitaemia. Together these observations suggest that high molFOB, and consequently the overall genetic diversity to which children were exposed in early childhood, contribute substantially to the rapid acquisition of clinical immunity to P. vivax across the entire age range [27].

The high exposure to P. vivax may even be sufficient for children to acquire a certain degree of immunity against P. vivax infection. With at least some malaria vectors, biting rates increase with host body size [44]. Therefore, as children grow older, their exposure to malarial infection also increases, as supported by the strong age-dependence of P. falciparum molFOI [28]. The lack of an association of P. vivax molFOB with age suggests that, in older children, some of the P. vivax sporozoites transmitted by mosquitoes do not succeed in establishing detectable blood stage infections.

Conclusions

We propose that the high number of genetically distinct P. vivax blood stages infections acquired in the first 4 years of life (as measured by molFOB) is a significant contributor to the rapid acquisition of immunity against clinical P. vivax malaria. Albeit less closely linked to transmission (i.e. force of infection (FOI)) than in P. falciparum [23], molFOB is nevertheless a measure of individual exposure to P. vivax blood-stage infection and is significantly linked to the observed burden of P. vivax malaria. As such, it could be used as both a surrogate maker for exposure and as a parameter for monitoring the impact of antimalarial interventions.

Supporting Information

Checklist S1.

STROBE checklist.

doi:10.1371/journal.pntd.0002403.s001

(DOC)

Table S1.

Parameter estimates from GLMMs predicting the number of incident clinical episodes of P. vivax malaria with any parasites density with and without adjustment for molFOB.

doi:10.1371/journal.pntd.0002403.s002

(DOCX)

Acknowledgments

We would like to thank the study participants and their parents or legal guardians, and the field teams of the PNG Institute of Medical Research. We thank Thomas A. Smith for critical discussion of the manuscript.

Author Contributions

Conceived and designed the experiments: IM IF. Performed the experiments: CK BK EL. Analyzed the data: CK KLC TPS IM. Contributed reagents/materials/analysis tools: PMS. Wrote the paper: CK IM IF KLC TPS.

References

  1. 1. Koch R (1900) Dritter Bericht über die Tätigkeit der Malariaexpedition. Deutsche Medizinische Wochenschrift 17. doi: 10.1055/s-0029-1203982
  2. 2. Doolan DL, Dobano C, Baird JK (2009) Acquired immunity to malaria. Clin Microbiol Rev 22: 13–36 Table of Contents. doi: 10.1128/cmr.00025-08
  3. 3. Snow RW, Omumbo JA, Lowe B, Molyneux CS, Obiero JO, et al. (1997) Relation between severe malaria morbidity in children and level of Plasmodium falciparum transmission in Africa. Lancet 349: 1650–1654. doi: 10.1016/s0140-6736(97)02038-2
  4. 4. Ciuca M, Ballif L, Chelarescu-Vieru M (1934) Immunity in malaria. Transactions of the Royal Society of Tropical Medicine and Hygiene 27: 4. doi: 10.1016/s0035-9203(34)90111-5
  5. 5. Jeffery GM (1966) Epidemiological significance of repeated infections with homologous and heterologous strains and species of Plasmodium. Bull World Health Organ 35: 873–882.
  6. 6. Michon P, Cole-Tobian JL, Dabod E, Schoepflin S, Igu J, et al. (2007) The risk of malarial infections and disease in Papua New Guinean children. Am J Trop Med Hyg 76: 997–1008.
  7. 7. Gunewardena DM, Carter R, Mendis KN (1994) Patterns of acquired anti-malarial immunity in Sri Lanka. Mem Inst Oswaldo Cruz 89 Suppl 2: 63–65. doi: 10.1590/s0074-02761994000600015
  8. 8. Maitland K, Williams TN, Bennett S, Newbold CI, Peto TE, et al. (1996) The interaction between Plasmodium falciparum and P. vivax in children on Espiritu Santo island, Vanuatu. Trans R Soc Trop Med Hyg 90: 614–620. doi: 10.1016/s0035-9203(96)90406-x
  9. 9. Phimpraphi W, Paul RE, Yimsamran S, Puangsa-art S, Thanyavanich N, et al. (2008) Longitudinal study of Plasmodium falciparum and Plasmodium vivax in a Karen population in Thailand. Malar J 7: 99. doi: 10.1186/1475-2875-7-99
  10. 10. Mueller I, Widmer S, Michel D, Maraga S, McNamara DT, et al. (2009) High sensitivity detection of Plasmodium species reveals positive correlations between infections of different species, shifts in age distribution and reduced local variation in Papua New Guinea. Malar J 8: 41. doi: 10.1186/1475-2875-8-41
  11. 11. Langhorne J, Ndungu FM, Sponaas AM, Marsh K (2008) Immunity to malaria: more questions than answers. Nat Immunol 9: 725–732. doi: 10.1038/ni.f.205
  12. 12. Tanabe K, Mackay M, Goman M, Scaife JG (1987) Allelic dimorphism in a surface antigen gene of the malaria parasite Plasmodium falciparum. J Mol Biol 195: 273–287. doi: 10.1016/0022-2836(87)90649-8
  13. 13. Fenton B, Clark JT, Khan CM, Robinson JV, Walliker D, et al. (1991) Structural and antigenic polymorphism of the 35- to 48-kilodalton merozoite surface antigen (MSA-2) of the malaria parasite Plasmodium falciparum. Mol Cell Biol 11: 963–971.
  14. 14. Marshall VM, Zhang L, Anders RF, Coppel RL (1996) Diversity of the vaccine candidate AMA-1 of Plasmodium falciparum. Mol Biochem Parasitol 77: 109–113. doi: 10.1016/0166-6851(96)02583-2
  15. 15. Rosenberg R, Wirtz RA, Lanar DE, Sattabongkot J, Hall T, et al. (1989) Circumsporozoite protein heterogeneity in the human malaria parasite Plasmodium vivax. Science 245: 973–976. doi: 10.1126/science.2672336
  16. 16. Tsuboi T, Kappe SH, al-Yaman F, Prickett MD, Alpers M, et al. (1994) Natural variation within the principal adhesion domain of the Plasmodium vivax duffy binding protein. Infect Immun 62: 5581–5586.
  17. 17. Bull PC, Marsh K (2002) The role of antibodies to Plasmodium falciparum-infected-erythrocyte surface antigens in naturally acquired immunity to malaria. Trends Microbiol 10: 55–58. doi: 10.1016/s0966-842x(01)02278-8
  18. 18. Osier FH, Fegan G, Polley SD, Murungi L, Verra F, et al. (2008) Breadth and magnitude of antibody responses to multiple Plasmodium falciparum merozoite antigens are associated with protection from clinical malaria. Infect Immun 76: 2240–2248. doi: 10.1128/iai.01585-07
  19. 19. Collins WE, Jeffery GM (1999) A retrospective examination of secondary sporozoite- and trophozoite-induced infections with Plasmodium falciparum: development of parasitologic and clinical immunity following secondary infection. Am J Trop Med Hyg 61: 20–35. doi: 10.4269/tropmed.1999.61-020
  20. 20. Smith T, Felger I, Tanner M, Beck HP (1999) Premunition in Plasmodium falciparum infection: insights from the epidemiology of multiple infections. Trans R Soc Trop Med Hyg 93 Suppl 1: 59–64. doi: 10.1016/s0035-9203(99)90329-2
  21. 21. Lyon JA, Angov E, Fay MP, Sullivan JS, Girourd AS, et al. (2008) Protection induced by Plasmodium falciparum MSP1(42) is strain-specific, antigen and adjuvant dependent, and correlates with antibody responses. PLoS ONE 3: e2830. doi: 10.1371/journal.pone.0002830
  22. 22. Cole-Tobian JL, Michon P, Biasor M, Richards JS, Beeson JG, et al. (2009) Strain-specific duffy binding protein antibodies correlate with protection against infection with homologous compared to heterologous plasmodium vivax strains in Papua New Guinean children. Infect Immun 77: 4009–4017. doi: 10.1128/iai.00158-09
  23. 23. Mueller I, Schoepflin S, Smith TA, Benton KL, Bretscher MT, et al. (2012) Force of infection is key to understanding the epidemiology of Plasmodium falciparum malaria in Papua New Guinean children. Proc Natl Acad Sci U S A 109: 10030–5. doi: 10.1073/pnas.1200841109
  24. 24. Imwong M, Snounou G, Pukrittayakamee S, Tanomsing N, Kim JR, et al. (2007) Relapses of Plasmodium vivax infection usually result from activation of heterologous hypnozoites. J Infect Dis 195: 927–933. doi: 10.1086/512241
  25. 25. Restrepo E, Imwong M, Rojas W, Carmona-Fonseca J, Maestre A (2011) High genetic polymorphism of relapsing P. vivax isolates in northwest Colombia. Acta Trop 119: 23–29. doi: 10.1016/j.actatropica.2011.03.012
  26. 26. Imwong M, Boel ME, Pagornrat W, Pimanpanarak M, McGready R, et al. (2011) The First Plasmodium vivax Relapses of Life Are Usually Genetically Homologous. J Infect Dis 205: 680–3. doi: 10.1093/infdis/jir806
  27. 27. Lin E, Kiniboro B, Gray L, Dobbie S, Robinson L, et al. (2010) Differential patterns of infection and disease with P. falciparum and P. vivax in young Papua New Guinean children. PLoS ONE 5: e9047. doi: 10.1371/journal.pone.0009047
  28. 28. Koepfli C, Ross A, Kiniboro B, Smith TA, Zimmerman PA, et al. (2011) Multiplicity and Diversity of Plasmodium vivax Infections in a Highly Endemic Region in Papua New Guinea. PLoS Negl Trop Dis 5: e1424. doi: 10.1371/journal.pntd.0001424
  29. 29. Kasehagen LJ, Mueller I, McNamara DT, Bockarie MJ, Kiniboro B, et al. (2006) Changing patterns of Plasmodium blood-stage infections in the Wosera region of Papua New Guinea monitored by light microscopy and high throughput PCR diagnosis. Am J Trop Med Hyg 75: 588–596.
  30. 30. Koepfli C, Schoepflin S, Bretscher M, Lin E, Kiniboro B, et al. (2011) How much remains undetected? Probability of molecular detection of human plasmodia in the field. PLoS ONE 6: e19010. doi: 10.1371/journal.pone.0019010
  31. 31. White NJ (2011) Determinants of relapse periodicity in Plasmodium vivax malaria. Malar J 10: 297. doi: 10.1186/1475-2875-10-297
  32. 32. Salman S, Page-Sharp M, Griffin S, Kose K, Siba PM, et al. (2011) Population pharmacokinetics of artemether, lumefantrine, and their respective metabolites in Papua New Guinean children with uncomplicated malaria. Antimicrobial agents and chemotherapy 55: 5306–5313. doi: 10.1128/aac.05136-11
  33. 33. Hombhanje FW, Hwaihwanje I, Tsukahara T, Saruwatari J, Nakagawa M, et al. (2005) The disposition of oral amodiaquine in Papua New Guinean children with falciparum malaria. Br J Clin Pharmacol 59: 298–301. doi: 10.1111/j.1365-2125.2004.02257.x
  34. 34. Salman S, Kose K, Griffin S, Baiwog F, Winmai J, et al. (2011) The pharmacokinetic properties of standard and double-dose sulfadoxine-pyrimethamine in infants. Antimicrob Agents Chemother 55: 1693–1700. doi: 10.1128/aac.01075-10
  35. 35. Bates D, Maechler M, Bolker B (2011) lme4: Linear mixed-effects models using S4 classes.
  36. 36. Wood SN (2006) Generalized Additive Models: An Introduction with R: Chapman and Hall/CRC.
  37. 37. Muller I, Genton B, Rare L, Kiniboro B, Kastens W, et al. (2009) Three different Plasmodium species show similar patterns of clinical tolerance of malaria infection. Malar J 8: 158. doi: 10.1186/1475-2875-8-158
  38. 38. Schoepflin S, Valsangiacomo F, Lin E, Kiniboro B, Mueller I, et al. (2009) Comparison of Plasmodium falciparum allelic frequency distribution in different endemic settings by high-resolution genotyping. Malar J 8: 250. doi: 10.1186/1475-2875-8-250
  39. 39. Hii JL, Smith T, Vounatsou P, Alexander N, Mai A, et al. (2001) Area effects of bednet use in a malaria-endemic area in Papua New Guinea. TransRSocTropMedHyg 95: 7–13. doi: 10.1016/s0035-9203(01)90315-3
  40. 40. Benet A, Mai A, Bockarie F, Lagog M, Zimmerman P, et al. (2004) Polymerase chain reaction diagnosis and the changing pattern of vector ecology and malaria transmission dynamics in Papua New Guinea. Am J Trop Med Hyg 71: 277–284.
  41. 41. Karunajeewa HA, Mueller I, Senn M, Lin E, Law I, et al. (2008) A trial of combination antimalarial therapies in children from Papua New Guinea. N Engl J Med 359: 2545–2557. doi: 10.1056/nejmoa0804915
  42. 42. Barnadas C, Koepfli C, Karunajeewa HA, Siba PM, Davis TME, et al. (2011) Characterization of Treatment Failure in Efficacy Trials of Drugs against Plasmodium vivax by Genotyping Neutral and Drug Resistance-Associated Markers. Antimicrob Agents Chemother 55: 4479–4481. doi: 10.1128/aac.01552-10
  43. 43. Betuela I, Rosanas-Urgell A, Kiniboro B, Stanisic DI, Samol L, et al. (2012) Relapses contribute significantly to the risk of Plasmodium vivax infection and disease in Papua New Guinean children 1–5 years of age. J Infect Dis 206: 1771–1780. doi: 10.1093/infdis/jis580
  44. 44. Port GR, Boreham PFL, Bryan JH (1980) The relationship of host size to feeding by mosquitos of the Anopheles gambiae Giles complex (Diptera, Culicidae). Bull Entomol Res 70: 133–144. doi: 10.1017/s0007485300009834