Introduction
Pregnant women are at an increased risk of both Plasmodium falciparum and Plasmodium vivax infections compared with their nonpregnant counterparts.1 It is estimated that over 85 million pregnancies each year are at risk of P. falciparum infection and 93 million are at risk of P. vivax infection.2 The increased risk of P. falciparum infection has largely been attributed to the ability of the P. falciparum-infected erythrocyte (IE) to bind and sequester in the placenta.3,4 Plasmodium vivax does not sequester to the same extent in the placenta, and reasons for altered risk of P. vivax infection are less clear, but immunological changes that occur during pregnancy may play a role.1 There is emerging evidence that the increased risk during pregnancy may not immediately return to normal after delivery,5–7 and a recent systematic review demonstrated that postpartum women may be another population at high risk of Plasmodium infection and clinical malaria episodes.8 The factors responsible for an altered risk of malaria in the postpartum period are unknown.
Individuals living in malaria-endemic regions develop naturally acquired immunity to P. falciparum and P. vivax with repeated infections. Antibodies are an important component of naturally acquired immunity against malaria.9,10 Antibodies targeting transmission stages (sporozoites, gametocytes) can prevent liver infection or transmission to mosquitoes and antibodies targeting blood stages (merozoites, IEs) can control parasitemia and prevent the development of clinical symptoms.11–13 The predominant antigen expressed on the surface of the P. falciparum IE is PfEMP114 and a PfEMP1 variant, VAR2CSA, mediates sequestration in the placenta via adherence to chondroitin sulphate A.3,4 Evidence from population studies suggests that antibody responses of sufficient breadth and magnitude can achieve protection against clinical malaria15–18 while also acting as biomarkers of past exposure.19 Thus, in populations experiencing relatively high homogenous exposure and high levels of immunity, high levels of antibodies against Plasmodium spp. have been reported as protective against clinical disease.20,21 Conversely, in areas where transmission is low and heterogenous, antibodies serve as a marker of increased risk with high-risk exposed individuals generating higher antibody responses compared with individuals with low risk of Plasmodium spp. exposure.22
Despite extensive literature documenting the increased risk of malaria and Plasmodium spp. infection in pregnancy, relatively little is known about the risk of malaria in the postpartum period. There is emerging evidence for an altered susceptibility to P. falciparum and P. vivax during the postpartum period.8 Studies undertaken in Senegal and Gabon reported that postpartum women were at an increased prospective risk of P. falciparum infection (relative risk = 1.8 and 2.7, respectively) and clinical falciparum malaria (relative risk = 4.1 and 9.8, respectively) relative to nonpregnant controls.5,6 Only one study, conducted on the Thailand–Myanmar border, has compared the prospective risk of both P. falciparum and P. vivax infection in postpartum and nonpregnant women and found that postpartum women experienced significantly less P. falciparum infections and significantly more P. vivax infections than nonpregnant controls (hazard ratio [HR] = 0.39, 95% confidence interval [CI] = 0.21, 0.72 and HR = 1.34, 95% CI: 1.05, 1.72, respectively).7 Despite these epidemiological observations, there have been few immunological investigations on the association of acquired immune responses and risk of infection during the postpartum period.1 We previously demonstrated that levels of antibodies against P. falciparum and P. vivax targets were reduced in postpartum women compared with nonpregnant women, but that these antibody levels recover to normal levels.23 The present study sought to investigate the relationship between antibodies specific for P. falciparum and P. vivax antigens and prospective risk of microscopically confirmed species-specific infection in these postpartum and control (nonpregnant and nonpostpartum) women.7
Materials and Methods
Ethics statement.
Ethics approval was sought and provided by The Alfred Hospital Human Research and Ethics Committee, Melbourne, Australia (88/13) and the Faculty of Tropical Medicine Ethics Committee, Mahidol University Bangkok, Thailand (MUTM 2007-023) and Oxford Tropical Medicine Ethical Committee, Oxford University, United Kingdom (002-07). All participants gave written, informed, or thumb print, if illiterate, consent before enrollment in the study.
Study design and population.
This study investigated 201 postpartum women and 201 nonpostpartum women (controls) over a 12-week period. These women represent a subset of women (described previously23) from a larger cohort study.7 Briefly, pregnant women attending Shoklo Malaria Research Unit (SMRU) antenatal clinics from November 2007 to September 2009 were invited to participate and nonpregnant females of similar age and from same location were recruited as controls. During follow-up, women underwent weekly blood smears and completed questionnaires on behavior. The first serological measurement (baseline) was available at first postpartum visit with additional serological measurements obtained approximately monthly thereafter. Microscopically confirmed P. falciparum infections were treated with mefloquine and artesunate. Microscopically confirmed P. vivax infections were treated with chloroquine.
Antibody determination.
Data were available for the levels of total IgG against a variety of P. falciparum (PfEBA140RIII–V, PfEBA175RII, PfEBA175RIII–V, PfAMA-1, PfMSP2, PfRh2, PfDBLα, and PfCSP) and P. vivax (PvAMA-1, PvMSP119, PvDBP, and PvCSP) antigens as assessed by enzyme-linked immunosorbent assay as well as total IgG binding to the surfaces of PfVAR2CSA-expressing IEs as assessed by flow cytometry.23 There were a total of 1,462 serological assessments from 402 women. Seropositivity to a given antigen was defined as an optical density (OD) (or in the case of PfVAR2CSA mean fluorescence intensity [MFI]) greater than the mean + 3 standard deviations derived from 37 negative controls (nonexposed Melbourne donors).
Statistical analysis.
All statistical analyses were performed using Stata Version 13.1 (StataCorp, College Station, TX). To determine the association between antibody levels and time to first microscopically confirmed P. falciparum or P. vivax infection, we constructed survival curves comparing high, medium, and low responders (tertiles based on all data) at baseline in postpartum and control subgroups and conducted log-rank tests as a preliminary analysis. Cox proportional hazards regressions were then performed using days after first antibody measurement as time, censoring at the recording of a microscopic P. falciparum or P. vivax infection; the woman was lost to follow-up; or the woman attended her final visit. Coinfections of P. vivax and P. falciparum (N = 2) were included as a P. vivax infection in P. vivax survival analyses and a P. falciparum infection in P. falciparum survival analyses. If a P. falciparum infection occurred, the 3-week period after treatment was omitted from the P. vivax exposure period, as treatment of P. falciparum with mefloquine and artesunate prevents P. vivax infection.24 If a P. vivax infection preceded a P. falciparum infection, the analysis was unchanged as chloroquine is not effective against P. falciparum in the study area.25 Individuals with a P. vivax or P. falciparum infection on the day of, or before, first antibody measurement but after delivery were excluded from relevant survival analyses. Antibody levels were analyzed as log-transformed continuous variables (log2([OD or MFI] + 0.001)) or as seropositive/seronegative (in analyses presented in Supplemental Tables 1 and 2).
The assumption of a linear association between antibody levels and the log hazard of infection was tested by comparing Cox regression models with categorical (quartile groupings) and pseudo-continuous antibody variables by likelihood ratio tests. Antibodies were modeled as time-varying exposures, where the most recent measurement was utilized for the following period of exposure. For adjusted analyses, confounders were selected a priori based on a causal diagram26 and included clinic visited (Wang Pha/Mawker Tai/Walley/Mu Ler Chai), maternal age (years), use of bednets every night during follow-up (yes/no), slept outside at any time during follow-up (yes/no), and work undertaken outside at any time during follow-up (yes/no).
To determine whether associations between antibody levels and time to first microscopically confirmed infection varied according to postpartum status a likelihood ratio test was performed comparing the model with and without an interaction term for postpartum status (postpartum/control). Where significant interactions were not observed, output from the simpler model was considered for the overall interpretation of results.
Results
Characteristics of postpartum and control women.
A total of 201 postpartum women and 201 control women were included in the immunological study. Eleven postpartum women and 12 control women experienced a microscopically confirmed P. falciparum infection during the follow-up period after their first antibody measurement. Forty-eight postpartum women and 36 control women experienced a microscopically confirmed P. vivax infection in the same period. One postpartum woman and one control woman experienced a coinfection of P. falciparum and P. vivax. Most individuals with microscopically confirmed infections did not have fever at the time of testing; there were only six febrile P. falciparum infections and six febrile P. vivax infections. Postpartum women were of a higher gravidity (median of 3 versus 2, P < 0.001) and were less likely to sleep and work outside (13.4% versus 19.9%, P = 0.08; and 18.4% versus 58.7%, P < 0.001) than control women (Table 1).
Characteristics of postpartum women and control women
| Postpartum (N = 201) | Controls (N = 201) | P value | |
|---|---|---|---|
| At enrollment | |||
| Age (years) | 27.5 (22–32) [18–45.5] | 28 (23–35) [18–50] | 0.40 |
| Gravidity | < 0.001 | ||
| Nulligravid | 0 (0) | 38 (18.9) | |
| 1–2 | 78 (38.8) | 75 (37.3) | |
| 3+ | 123 (61.2) | 88 (43.8) | |
| History of malaria* | |||
| Plasmodium falciparum in last 9 months | 35 (17.4) | 10 (5.0) | < 0.001 |
| Plasmodium vivax in last 9 months | 93 (46.3) | 11 (5.5) | < 0.001 |
| During follow-up | |||
| P. falciparum† | |||
| Any infection | 11 (5.8) | 12 (6.7) | 0.72 |
| Number of infections | 1 (1–2) [1–3] | 1 (1–1.5) [1–2] | 0.53 |
| P. vivax‡ | |||
| Any infection | 48 (26.3) | 36 (19.5) | 0.12 |
| Number of infections | 1 (1–1) [1–2] | 1 (1–1) [1–3] | 0.48 |
| Behavior | |||
| Use of bednets | 174 (86.6) | 184 (91.5) | 0.11 |
| Slept outside | 27 (13.4) | 40 (19.9) | 0.08 |
| Worked outside | 37 (18.4) | 118 (58.7) | < 0.001 |
Data presented as median (interquartile range) [minimum–maximum] or n (%). Wilcoxon rank-sum tests were performed on continuous data; χ2 tests were performed on categorical data.
Malaria history data collected differently between groups: postpartum women had experienced weekly smears during their pregnancy; control women had not been monitored before inclusion.
Among women included in P. falciparum survival analysis (189 postpartum, 178 controls).
Among women included in P. vivax survival analysis (182 postpartum, 185 controls).
Plasmodium falciparum antibodies and prospective risk of P. falciparum infection in the postpartum period.
Survival curves indicated that women with high antibody responses specific for P. falciparum antigens (PfEBA140RIII–V, PfEBA175RII, PfEBA175RIII–V, PfAMA-1, PfMSP2, PfRh2, PfDBLα, PfCSP, and PfVAR2CSA) at baseline had shorter time to first microscopic P. falciparum infection compared with those with medium and low antibody responses in both postpartum and control women (PfEBA140RIII–V, Figure 1A, P = 0.03 and 1B, P = 0.004, respectively; see Supplemental Figures 1 and 2 for other P. falciparum antigens).
Kaplan–Meier survival curves showing prospective risk of Plasmodium falciparum infection among baseline tertiles of PfEBA140RIII–V responders in postpartum (A) and controls (B) and the prospective risk of Plasmodium vivax infection among baseline tertiles of PvAMA-1 responders in postpartum (C) and controls (D).
Citation: The American Society of Tropical Medicine and Hygiene 96, 5; 10.4269/ajtmh.16-0690
Univariable and multivariable Cox regression models (adjusting for exposure behaviors, age, and antenatal clinic attended) were used to assess the relationship between antibodies (log2(units), time varying) specific for P. falciparum antigens with prospective risk of microscopic P. falciparum infection in the postpartum period (Table 2). Multivariable analyses showed that antibodies to each P. falciparum antigen were associated with an increased adjusted prospective risk of P. falciparum infection similarly in postpartum women and controls; for each 2-fold increase in OD there was a 3–79% increase (depending on antigen) in adjusted prospective risk of P. falciparum infection in postpartum women and a 14–75% increased adjusted prospective risk in control women (Table 2). The relationship between P. falciparum antibodies and prospective risk of infection did not significantly vary between postpartum and control women (likelihood ratio test, all P > 0.17). Similar results were observed when comparing seropositive women to seronegative women (Supplemental Table 1). Plasmodium falciparum seropositive women had a 2.8- to 17.3-fold (depending on antibody) increase in the prospective risk of P. falciparum infection than a woman who was seronegative (P < 0.05 for all antibodies). Overall, antibodies specific for P. falciparum antigens were associated with prospective risk of infection similarly in postpartum and control women.
Antibodies and prospective risk of Plasmodium falciparum infection
| Antibody (log2(units*)) | All women | Postpartum† | Control† | LR test for interaction‡ | |||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | P value | |
| Unadjusted | |||||||
| PfEBA140RIII–V | 1.41 (1.21, 1.65) | < 0.001 | 1.44 (1.13, 1.83) | 0.003 | 1.41 (1.15, 1.72) | 0.001 | 0.89 |
| PfEBA175RII | 1.37 (1.15, 1.62) | < 0.001 | 1.37 (1.06, 1.77) | 0.02 | 1.36 (1.08, 1.70) | 0.01 | 0.95 |
| PfEBA175RIII–V | 1.52 (1.26, 1.83) | < 0.001 | 1.44 (1.11, 1.87) | 0.007 | 1.61 (1.23, 2.11) | 0.001 | 0.56 |
| PfAMA1 | 1.41 (1.17, 1.69) | < 0.001 | 1.20 (1.01, 1.68) | 0.04 | 1.51 (1.15, 1.99) | 0.003 | 0.43 |
| PfMSP2 | 1.46 (1.17, 1.81) | 0.001 | 1.42 (1.04, 1.96) | 0.03 | 1.48 (1.10, 2.00) | 0.01 | 0.86 |
| PfRh2 | 1.94 (1.50, 2.50) | < 0.001 | 1.99 (1.33, 3.00) | 0.001 | 1.91 (1.37, 2.68) | < 0.001 | 0.88 |
| PfDBLα | 1.45 (1.07, 1.95) | 0.02 | 1.12 (0.68, 1.83) | 0.66 | 1.78 (1.18, 2.68) | 0.006 | 0.14 |
| PfVAR2CSA | 1.88 (1.36, 2.59) | < 0.001 | 2.14 (1.41, 3.27) | < 0.001 | 1.58 (0.93, 2.67) | 0.09 | 0.36 |
| PfCSP | 1.57 (1.11, 2.21) | 0.01 | 1.75 (1.09, 2.81) | 0.02 | 1.38 (0.84, 2.27) | 0.20 | 0.50 |
| Adjusted§ | |||||||
| PfEBA140RIII–V | 1.33 (1.15, 1.54) | < 0.001 | 1.32 (1.05, 1.65) | 0.02 | 1.35 (1.07, 1.65) | 0.003 | 0.87 |
| PfEBA175RII | 1.29 (1.10, 1.51) | 0.001 | 1.26 (0.99, 1.60) | 0.06 | 1.32 (1.06, 1.63) | 0.01 | 0.79 |
| PfEBA175RIII–V | 1.46 (1.21, 1.76) | < 0.001 | 1.36 (1.03, 1.80) | 0.03 | 1.55 (1.18, 2.03) | 0.002 | 0.53 |
| PfAMA1 | 1.34 (1.12, 1.61) | 0.001 | 1.22 (0.95, 1.57) | 0.12 | 1.47 (1.13, 1.92) | 0.01 | 0.32 |
| PfMSP2 | 1.37 (1.10, 1.71) | 0.01 | 1.36 (0.97, 1.89) | 0.07 | 1.39 (1.04, 1.86) | 0.03 | 0.92 |
| PfRh2 | 1.67 (1.29, 2.16) | < 0.001 | 1.54 (1.03, 2.32) | 0.04 | 1.75 (1.25, 2.44) | 0.001 | 0.64 |
| PfDBLα | 1.37 (1.03, 1.83) | 0.03 | 1.03 (0.61, 1.76) | 0.90 | 1.63 (1.11, 2.40) | 0.01 | 0.17 |
| PfVAR2CSA | 1.88 (1.29, 2.74) | 0.001 | 1.96 (1.21, 3.17) | 0.01 | 1.77 (0.97, 3.20) | 0.06 | 0.78 |
| PfCSP | 1.45 (0.97, 2.16) | 0.07 | 1.79 (1.06, 3.02) | 0.03 | 1.14 (0.64, 2.04) | 0.65 | 0.26 |
CI = confidence interval; HR = hazard ratio; LR = likelihood ratio.
Mean fluorescence intensity for PfVAR2CSA, optical density for all other antibody measurements.
Model with interaction term between antibody level and postpartum status.
LR test (P value) comparing model with interaction term between antibody level and postpartum status to model without interaction term.
Adjusted for exposure behaviors, age, postpartum status, and clinic attended.
Plasmodium vivax antibodies and prospective risk of P. vivax infection in the postpartum period.
Survival curves for P. vivax antibody responses at baseline showed that high, medium, and low responders against PvCSP, PvDBP, and PvAMA1 tended to have similar times to first microscopically confirmed P. vivax infection in postpartum and control women (PvAMA-1, Figure 1C and D, P > 0.45; see Supplemental Figure 3 for other P. vivax antigens). In contrast, PvMSP119 survival curves suggested that high responders to PvMSP119 had a slightly shorter time to first P. vivax infection (P = 0.04 for control women and P = 0.37 for postpartum women).
Multivariable Cox regression showed that levels (log2(OD), time varying) of antibodies specific for PvCSP, PvDBP, and PvAMA1 were not associated with prospective risk of P. vivax infection in all (postpartum and control) women (estimated HRs ranged from 0.99 to 1.14; Table 3). However, a 2-fold increase in PvMSP119 levels (OD) was associated with a modest 20% (HR = 1.20, 95% CI = 1.02 to 1.40) and 11% (HR = 1.11, 95% CI = 0.94 to 1.33) increased prospective risk of P. vivax infection among postpartum and control women, respectively, after adjustment. The relationship between P. vivax antibodies and prospective risk of infection did not significantly vary between postpartum and control women (likelihood ratio test, all P > 0.16). Similar results were observed when comparing seropositive women to seronegative women (Supplemental Table 2). In this population, antibodies specific for most P. vivax antigens were not associated with P. vivax infection in all women.
Antibodies and prospective risk of Plasmodium vivax infection
| Antibody (log2(units*)) | All women | Postpartum† | Control† | LR test for interaction‡ | |||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | P value | |
| Unadjusted | |||||||
| PvDBP | 1.02 (0.86, 1.21) | 0.82 | 1.05 (0.84, 1.33) | 0.76 | 1.00 (0.76, 1.31) | 0.99 | 0.76 |
| PvAMA-1 | 1.05 (0.97, 1.13) | 0.23 | 1.09 (0.99, 1.21) | 0.07 | 0.99 (0.87, 1.12) | 0.84 | 0.20 |
| PvMSP119 | 1.14 (1.02, 1.28) | 0.02 | 1.17 (1.01, 1.36) | 0.03 | 1.12 (0.95, 1.33) | 0.18 | 0.70 |
| PvCSP | 1.04 (0.87, 1.24) | 0.67 | 1.06 (0.84, 1.33) | 0.63 | 1.06 (0.80, 1.37) | 0.72 | 0.98 |
| Adjusted§ | |||||||
| PvDBP | 1.03 (0.87, 1.23) | 0.71 | 1.07 (0.85, 1.35) | 0.56 | 0.99 (0.76, 1.29) | 0.92 | 0.65 |
| PvAMA-1 | 1.06 (0.98, 1.15) | 0.15 | 1.12 (1.01, 1.24) | 0.04 | 0.99 (0.87, 1.12) | 0.85 | 0.12 |
| PvMSP119 | 1.16 (1.03, 1.30) | 0.01 | 1.20 (1.02, 1.40) | 0.03 | 1.11 (0.94, 1.33) | 0.22 | 0.55 |
| PvCSP | 1.11 (0.92, 1.33) | 0.28 | 1.14 (0.89, 1.46) | 0.31 | 1.07 (0.82, 1.40) | 0.62 | 0.75 |
CI = confidence interval; HR = hazard ratio; LR = likelihood ratio.
Mean fluorescence intensity for PfVAR2CSA, optical density for all other antibody measurements.
Model with interaction term between antibody level and postpartum status.
LR test (P value) comparing model with interaction term between antibody level and postpartum status to model without interaction term.
Adjusted for exposure behaviors, age, postpartum status, and clinic attended.
Associations between antibody levels and prospective risk of infection attenuate when adjusting for documented retrospective infections.
To assess whether antibodies were acting as biomarkers of retrospective exposure, we investigated the association of antibodies specific for Plasmodium spp. at baseline with retrospective Plasmodium spp. infections in postpartum women who had been screened weekly during pregnancy for the presence of Plasmodium spp. parasites. Antibody levels and seroprevalence to P. falciparum antigens were higher among postpartum women who experienced a P. falciparum infection during pregnancy (N = 35) compared with postpartum women who remained P. falciparum free during pregnancy (N = 166) (Figure 2). Similarly, postpartum women who experienced a P. vivax infection during pregnancy (N = 93) tended to have higher levels and seroprevalence of antibodies against P. vivax targets than postpartum women who had no P. vivax infections during pregnancy (N = 108). However, the magnitude of difference in P. vivax antibody responses was smaller than that observed for P. falciparum antibody responses between pregnancy exposure groups (Figure 2).
Antibodies to Plasmodium species antigens at baseline in postpartum women (N = 201) with species-specific infections during pregnancy (N = 35 and N = 93 for Plasmodium falciparum and Plasmodium vivax infections, respectively) and without species-specific infections during pregnancy (N = 166 and N = 108 for absence of P. falciparum and P. vivax infections, respectively). (A) Seroprevalence against P. falciparum and P. vivax antigens in women with species-specific infection during pregnancy (black circles) and women who did not experience a species-specific infection during pregnancy (gray triangles). Bars indicate 95% confidence intervals. P < 0.05 for all antibodies except PfDBLα (P = 0.09), PvDBP (P = 0.89), PvMSP119 (P = 0.09), and PvCSP (P = 0.20). (B) Box and whiskers plots of IgG levels (log2(MFI) for PfVAR2CSA, log2(OD) for all other antibodies) against P. falciparum and P. vivax antigens in women with species-specific infection during pregnancy (black) and women who did not experience a species-specific infection during pregnancy (gray). Horizontal line in box indicates median, box indicates the interquartile range, whiskers indicate the highest and lowest values within 1.5 × interquartile range of the first and third quartiles, dots represent outliers. P < 0.05 for all antibodies except PfVAR2CSA (P = 0.22), PvDBP (P = 0.35), and PvCSP (P = 0.14). MFI = mean fluorescence intensity; OD = optical density.
Citation: The American Society of Tropical Medicine and Hygiene 96, 5; 10.4269/ajtmh.16-0690
Adjusting for a documented P. falciparum infection during pregnancy in the postpartum women reduced the magnitude of association between P. falciparum antibodies and prospective risk of P. falciparum infection (median [minimum–maximum] of all HR estimates unadjusted for history = 1.36 [1.01–1.81], adjusted for history = 1.26 [0.85–1.52]) (Supplemental Table 3). Similarly, adjusting for a documented P. vivax infection during pregnancy in postpartum women reduced the magnitude of association between PvMSP119 and prospective risk of P. vivax infection (Supplemental Table 3) (HR unadjusted for history = 1.18, 95% CI = 1.00, 1.38; adjusted for history = 1.12, 95% CI = 0.95, 1.32). However, adjusting for retrospective infections as assessed by questionnaire in the control women did not alter the magnitude of associations (Supplemental Table 4). Positive associations between antibodies and prospective risk of Plasmodium spp. infection are likely reflective of a subset of women at high risk of Plasmodium spp. exposure (both retrospective and prospective).
Discussion
In the first study investigating humoral immunity and prospective risk of P. falciparum and P. vivax infection postpartum, all P. falciparum antibodies investigated showed a positive association with prospective risk of P. falciparum infection, and there was no statistical evidence that this association differed for postpartum and control women. Only PvMSP119 antibodies, but not PvAMA-1, PvDBP, or PvCSP antibodies, showed a positive association with prospective risk of P. vivax infection. Associations between antibody levels and prospective risk of infection were broadly comparable in postpartum and control women. In this low-endemic setting, antibodies specific for P. falciparum, but not for P. vivax, acted as biomarkers of both retrospective and prospective risk of infection, in both postpartum and control women.
Antibodies specific for P. falciparum antigens were associated with an increased prospective risk of P. falciparum infection in both postpartum and controls. Antibodies specific for P. falciparum antigens were considerably higher in postpartum women who had experienced a P. falciparum infection in pregnancy compared with those who had not and were therefore indicative of retrospective P. falciparum exposure in this population. The association of P. falciparum antibodies with prospective risk of infection is in line with studies from other populations in low-transmission settings, which have also reported positive associations between P. falciparum antibodies and risk of infection.21,27,28
Unlike P. falciparum antibody responses, antibodies specific for P. vivax antigens showed no association with prospective risk of P. vivax, with the exception of antibodies against PvMSP119. Antibodies against P. vivax antigens were typically higher in women who experienced an infection during pregnancy than those who were uninfected but the effect was less pronounced than that observed in response to P. falciparum targets. The association of antibodies against PvMSP119, but not other P. vivax targets, with prospective risk of P. vivax infection suggests that PvMSP119 may be more immunogenic than other P. vivax targets. Indeed, our earlier research indicated that antibodies against PvMSP119 were more responsive to P. vivax infection than the other antibody responses investigated.23 Another key difference between P. falciparum and P. vivax is that P. vivax possesses the ability to form hypnozoites in the liver, a dormant stage that can lead to relapses of blood-stage infections.29 Relapses are responsible for a high proportion of observed P. vivax infections,30,31 so any reduction in vector exposure that occurs postpartum will have a negligible impact on prospective risk of a blood-stage P. vivax infection. These data suggest that P. vivax antibodies against targets investigated in this study have limited utility in detecting retrospective exposure to P. vivax or predicting future exposure to P. vivax in this population.
Antibodies specific for P. falciparum and P. vivax were present at higher levels in women who had experienced a retrospective infection. Given that postpartum and control women were monitored for malaria differently in this period (weekly testing by microscopy versus a one-off retrospective questionnaire), we were not able to directly compare the effect of a retrospective infection across all women. Instead we ran subgroup analyses within postpartum and control women. Adjusting for infection during pregnancy reduced the association between antibodies and prospective risk of Plasmodium spp. infection in postpartum women, but adjustment for retrospective history of malaria by questionnaire in control women did not alter the magnitude of associations most likely because the questionnaire, unlike microscopy, did not accurately define prior malaria episodes. Our results suggest that serology may be a useful tool to predict and monitor individuals and populations at increased risk of P. falciparum, particularly in the absence of a detailed history of retrospective infections.
In the two other studies that have investigated the risk of P. falciparum infection in postpartum women compared with nonpregnant control women, an increased risk of infection in the postpartum women was observed (relative risk = 1.8 for P. falciparum infection, 4.1 for clinical malaria5; and incident rate ratio = 2.7 for P. falciparum infection, 9.8 for clinical malaria6). These studies, performed in high-transmission areas in Africa, are in stark contrast to unadjusted analyses of the present study population in a low-transmission area of Thailand where postpartum women were at reduced prospective risk of P. falciparum infection (HR = 0.39).7 The study settings in African and Asia centers differ in key respects. At SMRU, testing for the presence of Plasmodium spp. infection by light microscopy occurs weekly, with infected women receiving prompt treatment. This treatment is typically before the commencement of symptomatic malaria, precluding an assessment of clinical immunity. As such, associations investigated in this study were measuring the effect of antibodies on an individual having an infection with parasitemia greater than 50 parasites/μL (detection threshold of light microscopy).32 Both studies in Africa were conducted in areas of high P. falciparum endemicity5,6 where asymptomatic P. falciparum infections were not treated routinely. They found that, for a given P. falciparum infection, postpartum women were more likely to develop clinical symptoms compared with control women.5,6 Whether impaired humoral immunity in postpartum women reduces their ability to control an infection below a clinical pyrogenic threshold is unknown, but past research in other populations has suggested that even where increased levels of antibodies indicate an increased risk of infection, higher levels of antibodies are associated with a reduced likelihood of developing symptoms given infection.27 A transient reduction in antibody levels post pregnancy23 may be responsible for an increased risk of infection and disease in the African studies, but these studies did not investigate serology. As such, studies of immunity to malaria in postpartum women are warranted in high-transmission settings.
This study provides a comprehensive analysis of the relationship between antibodies toward two Plasmodium spp. and prospective risk of infection postpartum. Antibodies against numerous P. falciparum targets acted as biomarkers of exposure; only antibodies specific for PvMSP119 demonstrated any association with prospective risk of P. vivax infection. Further studies are required in high-transmission, high-immunity settings, particularly exploring the association between immunity and symptomatic episode endpoints. In addition, given that this is the only population where prospective risk of P. vivax postpartum has been investigated, further immunoepidemiological studies of P. vivax in other postpartum populations are needed to assess the generalizability of findings. While this study focused on antibody responses, additional research should be conducted into the relationship between antibody-independent mechanisms, such as innate and cell-mediated immunity, and susceptibility to malaria in the postpartum period.
ACKNOWLEDGMENTS
We thank Rosanna Powell, Gaoqian Feng, Andrew Guy, Xi Zen Yap, Vashti Irani and Kerryn Moore for technical assistance; Robin Anders, David L. Narum, Joseph Smith, and Annie Mo for provision of proteins; all the study participants for their participation; the Shoklo Malaria Research Unit staff for their contributions.
- 1.↑
McLean AR, Ataide R, Simpson JA, Beeson JG, Fowkes FJ, 2015. Malaria and immunity during pregnancy and postpartum: a tale of two species. Parasitology 142: 999–1015.
- 2.↑
Dellicour S, Guerra CA, ter Kuile FO, Snow RW, Tatem AJ, 2010. Quantifying the number of pregnancies at risk of malaria in 2007: a demographic study. PLoS Med 7: e1000221.
- 3.↑
Fried M, Duffy PE, 1996. Adherence of Plasmodium falciparum to chondroitin sulfate A in the human placenta. Science 272: 1502–1504.
- 4.↑
Salanti A, Staalsoe T, Lavstsen T, Jensen AT, Sowa MP, Arnot DE, Hviid L, Theander TG, 2003. Selective upregulation of a single distinctly structured var gene in chondroitin sulphate A-adhering Plasmodium falciparum involved in pregnancy-associated malaria. Mol Microbiol 49: 179–191.
- 5.↑
Diagne N, Rogier C, Sokhna CS, Tall A, Fontenille D, Roussilhon C, Spiegel A, Trape JF, 2000. Increased susceptibility to malaria during the early postpartum period. N Engl J Med 343: 598–603.
- 6.↑
Ramharter M, Grobusch MP, Kiessling G, Adegnika AA, Moller U, Agnandji ST, Kramer M, Schwarz N, Kun JF, Oyakhirome S, Issifou S, Borrmann S, Lell B, Mordmuller B, Kremsner PG, 2005. Clinical and parasitological characteristics of puerperal malaria. J Infect Dis 191: 1005–1009.
- 7.↑
Boel ME, Rijken MJ, Leenstra T, Pyae Phyo A, Pimanpanarak M, Keereecharoen NL, Proux S, Laochan N, Imwong M, Singhasivanon P, White NJ, McGready R, Nosten FH, 2013. Malaria in the post-partum period; a prospective cohort study. PLoS One 8: e57890.
- 8.↑
Boel ME, Rijken MJ, Brabin BJ, Nosten F, McGready R, 2012. The epidemiology of postpartum malaria: a systematic review. Malar J 11: 114.
- 9.↑
Cohen S, McGregor IA, Carrington S, 1961. Gamma-globulin and acquired immunity to human malaria. Nature 192: 733.
- 10.↑
Sabchareon A, Burnouf T, Ouattara D, Attanath P, Bouharoun-Tayoun H, Chantavanich P, Foucault C, Chongsuphajaisiddhi T, Druilhe P, 1991. Parasitologic and clinical human response to immunoglobulin administration in falciparum malaria. Am J Trop Med Hyg 45: 297–308.
- 11.↑
Cutts JC, Powell R, Agius PA, Beeson JG, Simpson JA, Fowkes FJ, 2014. Immunological markers of Plasmodium vivax exposure and immunity: a systematic review and meta-analysis. BMC Med 12: 150.
- 12.
Fowkes FJI, Richards JS, Simpson JA, Beeson JG, 2010. The relationship between anti-merozoite antibodies and incidence of Plasmodium falciparum malaria: a systematic review and meta-analysis. PLoS Med 7: 1–20.
- 13.↑
Doolan DL, Dobano C, Baird JK, 2009. Acquired immunity to malaria. Clin Microbiol Rev 22: 13–36, Table of Contents.
- 14.↑
Chan JA, Howell KB, Reiling L, Ataide R, Mackintosh CL, Fowkes FJ, Petter M, Chesson JM, Langer C, Warimwe GM, Duffy MF, Rogerson SJ, Bull PC, Cowman AF, Marsh K, Beeson JG, 2012. Targets of antibodies against Plasmodium falciparum-infected erythrocytes in malaria immunity. J Clin Invest 122: 3227–3238.
- 15.↑
Osier FHA, Fegan G, Polley SD, Murungi L, Verra F, Tetteh KKA, Lowe B, Mwangi T, Bull PC, Thomas AW, Cavanagh DR, McBride JS, Lanar DE, Mackinnon MJ, Conway DJ, Marsh K, 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.
- 16.
Richards JS, Stanisic DI, Fowkes FJ, Tavul L, Dabod E, Thompson JK, Kumar S, Chitnis CE, Narum DL, Michon P, Siba PM, Cowman AF, Mueller I, Beeson JG, 2010. Association between naturally acquired antibodies to erythrocyte-binding antigens of Plasmodium falciparum and protection from malaria and high-density parasitemia. Clin Infect Dis 51: e50–e60.
- 17.
Rono J, Osier FH, Olsson D, Montgomery S, Mhoja L, Rooth I, Marsh K, Farnert A, 2013. Breadth of anti-merozoite antibody responses is associated with the genetic diversity of asymptomatic Plasmodium falciparum infections and protection against clinical malaria. Clin Infect Dis 57: 1409–1416.
- 18.↑
Richards JS, Arumugam TU, Reiling L, Healer J, Hodder AN, Fowkes FJ, Cross N, Langer C, Takeo S, Uboldi AD, Thompson JK, Gilson PR, Coppel RL, Siba PM, King CL, Torii M, Chitnis CE, Narum DL, Mueller I, Crabb BS, Cowman AF, Tsuboi T, Beeson JG, 2013. Identification and prioritization of merozoite antigens as targets of protective human immunity to Plasmodium falciparum malaria for vaccine and biomarker development. J Immunol 191: 795–809.
- 19.↑
Elliott SR, Fowkes FJ, Richards JS, Reiling L, Drew DR, Beeson JG, 2014. Research priorities for the development and implementation of serological tools for malaria surveillance. F1000Prime Rep 6: 100.
- 20.↑
Murungi LM, Kamuyu G, Lowe B, Bejon P, Theisen M, Kinyanjui SM, Marsh K, Osier FH, 2013. A threshold concentration of anti-merozoite antibodies is required for protection from clinical episodes of malaria. Vaccine 31: 3936–3942.
- 21.↑
Stanisic DI, Fowkes FJ, Koinari M, Javati S, Lin E, Kiniboro B, Richards JS, Robinson LJ, Schofield L, Kazura JW, King CL, Zimmerman P, Felger I, Siba PM, Mueller I, Beeson JG, 2015. Acquisition of antibodies against Plasmodium falciparum merozoites and malaria immunity in young children: influence of age, force of infection, and magnitude of response. Infect Immun 83: 646–660.
- 22.↑
Bejon P, Warimwe G, Mackintosh CL, Mackinnon MJ, Kinyanjui SM, Musyoki JN, Bull PC, Marsh K, 2009. Analysis of immunity to febrile malaria in children that distinguishes immunity from lack of exposure. Infect Immun 77: 1917–1923.
- 23.↑
McLean AR, Boel ME, McGready R, Ataide R, Drew D, Tsuboi T, Beeson JG, Nosten F, Simpson JA, Fowkes FJ, 2016. Antibody responses to Plasmodium falciparum and Plasmodium vivax blood-stage and sporozoite antigens in the postpartum period. Sci Rep 6: 32159.
- 24.↑
Tjitra E, Baker J, Suprianto S, Cheng Q, Anstey NM, 2002. Therapeutic efficacies of artesunate-sulfadoxine-pyrimethamine and chloroquine-sulfadoxine-pyrimethamine in vivax malaria pilot studies: relationship to Plasmodium vivax dhfr mutations. Antimicrob Agents Chemother 46: 3947–3953.
- 25.↑
Villegas L, McGready R, Htway M, Paw MK, Pimanpanarak M, Arunjerdja R, Viladpai-Nguen SJ, Greenwood B, White NJ, Nosten F, 2007. Chloroquine prophylaxis against vivax malaria in pregnancy: a randomized, double-blind, placebo-controlled trial. Trop Med Int Health 12: 209–218.
- 26.↑
Williamson EJ, Aitken Z, Lawrie J, Dharmage SC, Burgess JA, Forbes AB, 2014. Introduction to causal diagrams for confounder selection. Respirology 19: 303–311.
- 27.↑
Greenhouse B, Ho B, Hubbard A, Njama-Meya D, Narum DL, Lanar DE, Dutta S, Rosenthal PJ, Dorsey G, John CC, 2011. Antibodies to Plasmodium falciparum antigens predict a higher risk of malaria but protection from symptoms once parasitemic. J Infect Dis 204: 19–26.
- 28.↑
Bejon P, Cook J, Bergmann-Leitner E, Olotu A, Lusingu J, Mwacharo J, Vekemans J, Njuguna P, Leach A, Lievens M, Dutta S, von Seidlein L, Savarese B, Villafana T, Lemnge MM, Cohen J, Marsh K, Corran PH, Angov E, Riley EM, Drakeley CJ, 2011. Effect of the pre-erythrocytic candidate malaria vaccine RTS,S/AS01E on blood stage immunity in young children. J Infect Dis 204: 9–18.
- 29.↑
Krotoski WA, Collins WE, Bray RS, Garnham PC, Cogswell FB, Gwadz RW, Killick-Kendrick R, Wolf R, Sinden R, Koontz LC, Stanfill PS, 1982. Demonstration of hypnozoites in sporozoite-transmitted Plasmodium vivax infection. Am J Trop Med Hyg 31: 1291–1293.
- 30.↑
Robinson LJ, Wampfler R, Betuela I, Karl S, White MT, Suen CSLW, Hofmann NE, Kinboro B, Waltmann A, Brewster J, Lorry L, Tarongka N, Samol L, Silkey M, Bassat Q, Siba PM, Schofield L, Felger I, Mueller I, 2015. Strategies for understanding and reducing the Plasmodium vivax and Plasmodium ovale hypnozoite reservoir in Papua New Guinean children: a randomised placebo-controlled trial and mathematical model. PLoS Med 12: e1001891.
- 31.↑
Betuela I, Rosanas-Urgell A, Kiniboro B, Stanisic DI, Samol L, de Lazzari E, Del Portillo HA, Siba P, Alonso PL, Bassat Q, Mueller I, 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.