- Email: [email protected]
Variant surface antigen-specific IgG and protection against clinical consequences of pregnancy-associated Plasmodium falciparum malaria
MECHANISMS OF DISEASE
Mechanisms of disease
Variant surface antigen-specific IgG and protection against clinical consequences of pregnancy-associated Plasmodium falciparum malaria Trine Staalsoe, Caroline E Shulman, Judith N Bulmer, Ken Kawuondo, Kevin Marsh, Lars Hviid
Summary Background Pregnancy-associated malaria caused by Plasmodium falciparum adherence to chondroitin sulfate A in the placental intervillous space is a major cause of low birthweight and maternal anaemia in areas of endemic P falciparum transmission. Adhesion-blocking antibodies that specifically recognise parasite-encoded variant surface antigens (VSA) are associated with resistance to pregnancyassociated malaria. We looked for a possible relation between VSA-specific antibody concentrations, placental infection, and protection from low birthweight and maternal anaemia. Methods We used flow cytometry to measure VSA-specific IgG concentrations in plasma samples taken during child birth from 477 Kenyan women selected from a cohort of 910 women on the basis of HIV-1 status, gravidity, and placental histology. We measured VSA expressed by one placental P falciparum isolate and two isolates selected or not selected for chondroitin sulfate A adhesiveness in-vitro. Findings Concentrations of plasma IgG specific for VSA, expressed by chondroitin sulfate A-adhering parasites (VSA in pregnancy-associated malaria or vsa-pam), increased with gravidity and were associated with placental histological findings. Women with chronic pregnancy-associated malaria and low or absent VSA-PAM-specific IgG had lower haemoglobin values (reduced by 17 g/L; 95% CI 8·1 –25·2) and delivered smaller babies (birthweight reduced by 0·26 kg; 0·10–0·55) than did corresponding women with high VSA-PAM-specific IgG. No such relation was shown for concentrations of IgG with specificity for non-pregnancy-associated malaria VSA. Interpretation VSA-PAM-specific IgG protects against low birthweight and maternal anaemia. Our data indicate an important mechanism of clinical protection against malaria and raise hope for the clinical effectiveness of a potential VSA-based vaccine against pregnancy-associated malaria. Lancet 2004; 363: 283–89
Centre for Medical Parasitology at Department of Infectious Diseases and Department of Clinical Microbiology, Copenhagen University Hospital (Rigshospitalet) and Institute for Medical Microbiology and Immunology, University of Copenhagen, Copenhagen, Denmark (T Staalsoe PhD, L Hviid PhD); London School of Hygiene and Tropical Medicine, London, UK (C E Shulman MRCGP); Centre for Geographical Medicine Research Coast, Kenya Medical Research Institute, Kilifi, Kenya (C E Shulman MRCGP, K Kawuondo HND, Prof K Marsh FRCP); and Department of Pathology, University of Newcastle upon Tyne, Royal Victoria Infirmary, Newcastle upon Tyne, UK (J N Bulmer MRCPath) Correspondence to: Dr Lars Hviid, Department of Infectious Diseases M7641, Rigshospitalet, Blegdamsvej 9, 2100 Copenhagen, Denmark. (e-mail: [email protected])
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
Introduction In areas of intense parasite transmission, severe Plasmodium falciparum malaria is largely a childhood disease due to gradual acquisition of protective immunity, although pregnant women—and in particular primigravidae— constitute an important exception. PREGNANCY-ASSOCIATED MALARIA, which is characterised by accumulation of infected erythrocytes in the placental intervillous space,1–3 is a major cause of low birthweight and maternal anaemia in areas of endemic parasite transmission.4–6 A detailed understanding of the pathogenesis and immunology of pregnancy-associated malaria has emerged only after results of investigations showing that infected erythrocytes obtained from the placenta of women with malaria are unique in their ability to adhere to chondroitin sulfate A in vitro and in their absence of adhesiveness to other receptor molecules commonly exploited by non-placental infected erythrocytes.2 These findings suggest that parasites responsible for pregnancy-associated malaria express a distinct set of molecules mediating adhesion to receptors located only in placental tissue,3 explaining the sudden susceptibility to malaria in otherwise clinically immune women at the time of their first pregnancy.4 Parasite-encoded VARIANT SURFACE ANTIGENS (VSA) inserted into the membrane of the infected erythrocyte are the primary mediators of the characteristic adhesion of erythrocytes infected by P falciparum. The best characterised of these adhesins is P falciparum erythrocyte membrane protein 1 (PfEMP1); a family of polymorphic, high molecular weight proteins encoded by var genes present at about 60 loci per genome.7 VSA can mediate adhesion to a number of receptors in the host vasculature, including chondroitin sulfate A.8 Several results strengthen the hypothesis that parasites in pregnancy-associated malaria express unique VSA (VSA-PAM) that are antigenically distinct from VSA present on non-placental infected erythrocytes. Thus, men exposed to malaria never have VSA-PAM-specific IgG despite high values of other VSA-specific antibodies, a finding that we have termed sex-specificity.9,10 Also, women exposed to malaria do not acquire VSA-PAM-specific antibodies earlier than halfway through their first pregnancy.10,11 Finally, plasma concentrations of IgG with specificity for VSA-PAM, but not for other VSA, are highly correlated with gravidity in pregnant, malaria-exposed women (parity-dependency).9,10 The ability of VSA-PAM-specific IgG to interfere with chondroitin sulfate A-specific adhesion in vitro,9,10,12 and the association between VSA-PAM IgG values and protection against placental parasitemia,10 indicate that the parity-dependent increase in VSA-PAM-specific IgG concentrations and in resistance to pregnancy-associated malaria are causally related, but reliable evidence has not been forthcoming. We thus set out to investigate the hypothesis that VSA-PAM-specific IgG protects against maternal anaemia and low birthweight, which are the most important clinical consequences of pregnancy-associated malaria. 283
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
GLOSSARY PREGNANCY-ASSOCIATED MALARIA
A specific syndrome seen only in pregnant women and characterised by accumulation of parasitised erythrocytes in the placenta. Such malaria is associated with substantial maternal and perinatal morbidity and mortality. PfEMP1
Plasmodium falciparum erythrocyte membrane protein 1. A family of high molecular weight parasite proteins expressed on the surface of parasitised erythrocytes and mediating their adhesion to a range of host receptors. VARIANT SURFACE ANTIGEN
Parasite-encoded, clonally-variant antigens, inserted into the surface membrane of the infected erythrocyte and thought to be an important target of protective immunity to P falciparum malaria.
VSA-PAM-specific IgG reactivity was used neat, and diluted 1 in 4 and 1 in 16 as positive controls. Sample size was determined by the need to undertake all measurements on a specific parasite isolate in one experiment, imposing an upper limit of about 500 on total plasma sample size. Parasites and in-vitro selection for adhesion to chondroitin sulfate A We used three P falciparum isolates cultured in vitro by standard methodology.10 One isolate (EJ24) was obtained at delivery from the placenta of a woman with pregnancyassociated malaria. The isolation, adhesion properties, and other characteristics of EJ24 are described in detail elsewhere.14 The second isolate (Busua) was obtained from the peripheral blood of a non-immune man, and was
VSA-PAM
A
Variant surface antigen in pregnancy-associated malaria.
NC
EJ24 EJ24
1.
Methods
2. 3–4. 5–7. 8–15. 50
B
75
100
125
NC
150 Busua Busua
1. Gravidity
Plasma samples and study population We studied plasma samples taken during delivery from 477 pregnant women, giving birth between Jan 1, 1996, and July 31, 1997, at Kilifi District Hospital, situated in an area of perennial, seasonal, and hyperendemic parasite transmission on the Kenyan coast.6 All samples were selected from a larger set from 910 pregnant women recruited for a separate investigation.6 Sample selection was done on the basis of HIV-1 status, gravidity (primigravid, 2–3 gravid, or >3 gravid), and placental histology (no malaria infection, acute infection, chronic infection, or past infection).13 We aimed to investigate 25 samples or more from HIV-1-negative women in all 12 categories of gravidity and histology, but could not always do so. We also included 75 HIV-1-positive women from the original investigation to allow assessment of the effect of this infection on VSA-PAM-specific IgG plasma values. Selection of samples from HIV-1-negative women was random within each group. Plasma samples from nine third-trimester pregnant women without malaria exposure were included as negative controls.9 Pooled plasma from malaria-exposed women in their third trimester with high
2. 3–4. 5–7.
8–15. 40
C
55
70
NC
85
100
Busua-CSA Busua-CSA
1. 2. 3–4.
Characteristic
Selected
All
Ethnic group Mijikenda Other Age group (years) Not known 14–19 20–34 35 and above Education and literacy Literate Primary, but illiterate Illiterate, no schooling Coral/stone-walled house House with latrine Ownership of radio Coral/stone house + latrine + radio Body-mass index <20 kg/m2 Pre-eclampsia Severe anemia (Hb <70g/L) Stillbirth Low birthweight (livebirths only)
477
910 349 (73·2%) 128 (26·8%)
p
5–7.
669 (73·5%) 241 (26·5%)
0·89
65 (13·6%) 90 (18·9%) 305 (63·9%) 17 (3·6%)
139 (15·3%) 177 (19·5%) 562 (61·8%) 32 (3·5%)
0·83
610 (67·0%) 71 (7·8%) 229 (25·2%) 320 (35·4%) 643 (70·7%) 562 (62·5%) 246 (27·2%)
0·40
473 476 472 474
910 335 (70·5%) 34 (7·2%) 106 (22·3%) 187 (39·5%)905 341 (71·6%)909 298 (63·1%)899 145 (30·6%)906
469 445 474 477 432
145 (31·1%)896 69 (15·5%) 862 74 (15·6%) 906 45 (9·4%) 910 84 (19·4%) 820
278 (31·0%) 0·97 122 (14·2%) 0·51 125 (13·8%) 0·36 90 (9·9%) 0·86 150 (18·3%) 0·62
477
475
910
50
D
70
MFI=59·6
0·13 0·73 0·82 0·18
Hb=haemoglobin.
Table 1: Principal characteristics of women included in study (selected) and those of the original study cohort6 (all) from which they were selected
284
8–15.
0
90 MFI=90·0
110
130 MFI=124·2
255 0 255 0 FITC fluorescence (channel number)
255
Figure 1: Relation between gravidity and delivery plasma concentrations of IgG (expressed in MFI) with specificity for parasite-encoded VSA expressed on the surface of P falciparum-infected erythrocytes MFI=mean fluorescence index. CSA=chondroitin sulfate A. NC=nonexposed controls. A–C: vertical lines in centre of box=median. Boxes=IQR. Whiskers=10% and 90% percentiles. Dots=outliers. A=IgG with specificity for VSA expressed by placental EJ24 isolate. B=IgG with specificity for VSA expressed by non-selected Busua. C=IgG with specificity for VSA expressed by Busua, after selection for adhesion to CSA (Busua-CSA). D=Histograms of fluorescent Busua-CSA infected erythrocytes after labelling by plasma samples with low (left), medium (centre), and high (right) VSA-specific IgG reactivity, respectively.
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
studied with (Busua-chondroitin sulfate A) or without (Busua) preceding adhesion selection by repeated panning on chondroitin sulfate A in vitro.9,10 The third isolate (FCR3, a long-term in-vitro culture isolate),9 was also investigated with (FCR3-chondroitin sulfate A) or without (FCR3) in-vitro selection for chondroitin sulfate A-adhesion. EJ24, Busua-chondroitin sulfate A, and FCR3-chondroitin sulfate A all showed the sex-specific and parity-dependent plasma IgG recognition characteristic of VSA expressed by parasite isolates related to pregnancyassociated malaria,9,10 whereas unselected Busua and FCR3 did not. The genotypic identities of all isolates were regularly confirmed by genotypic profiling at the polymorphic msp1, msp2, and glurp loci.9 Measurement of variant-specific IgG antibodies Plasma concentrations of VSA-specific IgG were measured by flow cytometry.15 Parasite cultures were enriched for late trophozoite and schizont erythrocytes by exposure to a strong magnetic field,15 labelled by ethidium bromide (to identify infected erythrocytes in flow cytometry data analysis), and sequentially exposed to plasma, secondary goat-anti-human IgG (Dako, Glostrup, Denmark), and tertiary fluorescein isothiocyanateconjugated rabbit-anti-goat Ig (Dako). Flow cytometry
data from 10 000 infected erythrocytes were acquired with a Coulter EPICS XL-MCL instrument (Coulter Electronics, Luton, UK). For all samples, the mean fluorescence index (MFI) was recorded as a measure of VSA-specific IgG reactivity.15 Samples in which forwardscatter and ethidium bromide analysis revealed technical problems were excluded from the analyses. Statistical analysis We assessed differences in proportions by 2 analysis, and group-wise differences by t test, Kruskal-Wallis, one-way ANOVA on ranks, and Mann-Whitney rank-sum test as appropriate. Parity-dependency was evaluated by Spearman’s rank-order correlation analysis with CIs calculated as described. We used a two-sample Kolmogorov-Smirnov test to compare distributions of VSA-specific IgG after correction for differences in data location and dispersion. The effects of potential confounding variables were assessed by multiple linear regression analysis with Stata (version 7.0).
A
(1 in 16)(1 in 4) (1 in 1) 0·20 EJ24
0·15 A NC
EJ24
0·10
None 0·05
Acute Chronic
0·00 50
Past
B 75
100
125
Placental infection
Busua
None Acute Chronic Past 40
55
70
85
100
C NC
Busua-CSA
None
140 (1 in 1)
0·20
150
B NC
110
(1 in 16) (1 in 4) Proportion of samples
50
80
Busua
0·15 0·10 0·05 0·00 39
C
57
75
93
(1 in 16) (1 in 4) (1 in 1) 0·20 Busua-CSA
0·15
Acute
0·10
Chronic 0·05
Past 50
70
90
110
130
0·00 44
68
92
116
FITC mean fluorescence index (channel number) FITC mean fluorescence index (channel number) Figure 2: Relation between placental histological findings and delivery plasma concentrations of IgG with specificity for parasite-encoded VSA expressed on surface of P falciparuminfected erythrocytes. CSA=chondroitin sulfate A. A–C: Lines in centre of box=median. Boxes=IQR. Whiskers=10% and 90% percentiles. Dots=outliers. A=IgG with specificity for VSA expressed by placental EJ24 isolate. B=IgG with specificity for VSA expressed by non-selected Busua. C=IgG with specificity for VSA expressed by Busua, after selection for adhesion to CSA (Busua-CSA).
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
Figure 3: Distribution of VSA-specific IgG plasma concentrations CSA=chondroitin sulfate A. Absolute values in nine non-exposed women shown as dots; proportions of values in 477 malaria-exposed study women shown as bars. Antibody concentrations in a four-fold titred pool of highly VSA-PAM-reactive plasma samples shown along the top of graphs. A=IgG with specificity for VSA expressed by placental EJ24 isolate. B=IgG with specificity for VSA expressed by non-selected Busua. C=IgG with specificity for VSA expressed by Busua, after selection for adhesion to CSA (Busua-CSA).
285
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
VSA-PAM IgG concentration (g/L)
EJ24 Busua-CSA FCR3-CSA ⭓2 VSA IgG high‡ Busua FCR3
Low*
High*
Difference†
p
77·5 (53) 75·4 (34) 76·4 (33) 75·7 (40) 80·6 (26) 81·5 (53)
95·4 (37) 91·1 (57) 90·1 (59) 92·4 (52) 86·9 (64) 90·2 (39)
–17·9 (–26·4 to –9·3) –15·7 (–24·5 to –6·8) –13·7 (–22·8 to –4·6) –16·6 (–25·2 to –8·1) –06·3 (–16·5 to 3·9) –08·6 (–17·7 to 0·4)
<0·0001 0·0007 0·004 0·0002 0·22 0·06
CSA=chondroitin sulfate A. *Mean Hb (g/L). Number of individuals in parentheses. †Mean (95% CIs). ‡Individuals with high IgG for at least two of the three isolates expressing VSA–PAM.
Table 2: Maternal haemoglobin concentrations (g/L) at time of delivery in women with chronic pregnancy-associated P falciparum malaria according to their plasma IgG concentration with specificity for type of VSA
Role of the funding source The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
Results Apart from inclusion and exclusion criteria (parity, placental pathology, and HIV-1 status), the characteristics of the 477 women included in the study were similar (p>0·13 in all cases) to those of the cohort (n=910)6 from which they were selected (table 1). This finding indicates that the selection procedure did not introduce any inadvertent bias. Concentrations of plasma IgG recognising intact erythrocytes infected by the placental EJ24 isolate correlated significantly with donor gravidity (rs=0·35 [95% CIs 0·26 to 0·42], p<0·0001, figure 1), but not age (rs=0·06 [–0·04 to 0·15], p=0·21), confirming the characteristic parity-dependency of plasma IgG recognition of VSAPAM-type antigens.9,10 By contrast, there was no significant correlation between gravidity and IgG specific for VSA expressed by the two isolates not related to pregnancyassociated malaria (Busua rs=0·06 [–0·03 to 0·15], p=0·20, figure 1) and FCR3 (rs=0·03 [–0·06 to 0·12], p=0·51). After repeated rounds of selection of Busua and FCR3 for the capacity of infected erythrocytes to adhere to chondroitin sulfate A, both selected sub-lines expressed VSA that caused significantly parity-dependent plasma IgG recognition (Busua-chondroitin sulfate A: rs=0·33 [0·25; 0·41], p<0·0001, figure 1, and FCR3-chondroitin sulfate A: rs=0·31 [0·23 to 0·39], p<0·0001). With respect to placental histology findings, VSA-specific IgG tended to be higher in women in whom pregnancyassociated malaria was either chronic or past than in those in whom it was or was not acute, indicating that such malaria leads to a boosting of immunity to parasites related and unrelated to pregnancy-associated malaria (figure 2). Primigravid women without histological evidence of pregnancy-associated malaria had uniformly low VSAPAM-specific plasma IgG concentrations at delivery, which were not significantly different from those in unexposed pregnant women, suggesting that they had never been exposed to VSA-PAM-expressing parasites despite lifelong
P falciparum exposure (not shown). Primigravid women with acute or chronic pregnancy-associated malaria had low-to-medium VSA-PAM IgG, probably indicating the slow development of a primary response to VSA-PAM in these women.10,11 However, a high proportion of multigravid women had moderate-to-high values, suggesting fast and boostable memory responses (not shown).9–12 Peripheral parasitaemia was not significantly associated with VSA-specific IgG values, maternal haemoglobin, or birthweight. However, peripheral parasitaemia was related to placental histology, being most common in women with chronically infected placentas and least common in those without, or with resolved, placental infection (p<0·0001). These findings confirm earlier results that peripheral parasitaemia in pregnant women living in highly endemic areas generally originates from a placental focus.16,17 HIV-1 infection is an important risk factor for pregnancy in malaria-endemic areas.18 We therefore compared concentrations of VSA-specific IgG in the 75 HIV-1-positive and the 402 HIV-1-negative women in the study cohort. Although values tended to be slightly lower among HIV-1-positive women, the difference was not significant for any of the five parasites tested (p between 0·07 and 0·94). Placental infection can persist despite high VSA-PAMspecific IgG, bringing into question the protectiveness of such antibodies,19 and indeed many women showed histological evidence of continuing pregnancy-associated malaria despite high VSA-PAM-specific IgG (figure 2). To investigate this issue further, we examined the distribution of VSA-specific antibodies in the study cohort. The distribution of IgG specific for the placental isolate EJ24 (figure 3) and both chondroitin sulfate A-selected sub-lines (Busua-chondroitin sulfate Aand FCR3-chondroitin sulfate A) (figure 3 and not shown) all appeared dichotomous, by contrast with the one-peaked distribution seen for plasma IgG reactivity to VSA expressed by parasites not related to pregnancy-associated malaria (Busua and FCR3) (figure 3 and not shown). Two-sample KolmogorovSmirnov tests confirmed that the shapes of the distributions of IgG concentrations with specificity for EJ24 (figure 3) and Busua-chondroitin sulfate A (figure 3) were similar (p=0·91), but different from that of IgG concentrations
Birthweight (kg)
EJ24 Busua-CSA FCR3-CSA ⭓2 VSA-PAM IgG high‡ Busua FCR3
Low*
High*
Difference†
p
2·69 (53) 2·56 (34) 2·62 (33) 2·68 (40) 2·78 (26) 2·70 (53)
2·92 (37) 2·95 (57) 2·92 (59) 2·94 (52) 2·73 (64) 2·91 (39)
–0·23 (–0·52 to –0·03) –0·39 (–0·74 to –0·23) –0·30 (–0·58 to –0·13) –0·26 (–0·55 to –0·10) 0·05 (–0·32 to 0·18) –0·21 (–0·39 to 0·07)
0·03 0·001 0·004 0·007 0·57 0·18
CSA=chondroitin sulfate A. *Median birthweight of offspring; number of individuals in parentheses. †Median (95% CIs). ‡Individuals with high IgG for at least two of the three isolates expressing VSA-PAM.
Table 3: Birthweights of children born to mothers with chronic pregnancy-associated P falciparum malaria according to mothers’ plasma IgG concentration with specificity for type of VSA
286
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
Constant Primigravidity* Sex of offspring† Mother’s weight (kg) Educational level‡ VSA-PAM (FCR3-CSA) (MFI) Non-PAM VSA (FCR3) (MFI)
Coefficient (slope) (95% CI)
p
Coefficient (slope) (95% CI)
p
3·46 (–0·29 to 7·22) 0·11 (–1·00 to 0·77) 0·10 (–0·73 to 0·94) –0·0069 (–0·057 to 0·043) 1·67 (0·56 to 2·78) 0·030 (0·009 to 0·013) 0·021 (–0·007 to 0·049)
0·074 0·81 0·81 0·79 0·004 0·001 0·15
0·61 (–0·53 to 1·75) 0·17 (–0·10 to 0·44) 0·35 (0·09 to 0·60) 0·019 (0·004 to 0·035) 0·164 (–0·17 to 0·50) 0·0088 (0·004 to 0·014) –0·00030 (–0·009 to 0·008)
0·30 0·22 0·0090 0·015 0·35 0·0010 0·95
CSA=chondroitin sulfate A. *Primigravidae (0) or multigravidae (1). †Female (0) or male (1). ‡Completed primary school (1) or not (0).
Table 4: Multiple linear regression analysis of pregnancy outcome in 89 women with chronic pregnancy-associated malaria
with specificity for VSA expressed by the unselected Busua isolate (figure 3) (EJ24: p=0·04; Busua-CSA: P=0·002). Based on this finding, we divided the women into those with low (reactivity less than that of a 1 in 16 dilution of the positive control plasma pool, figure 3) and high (⭓1 in 16) VSA-specific IgG, respectively, to determine the relation between VSA-specific IgG and clinically important pregnancy outcome measures. This procedure was done with respect to each of the five parasites used. Women with chronic pregnancy-associated malaria and low VSA-PAM-specific IgG were substantially more anaemic than were comparable women with adequate concentrations of VSA-PAM-specific IgG (table 2 and figure 4). The estimated effect of VSA-PAM-specific IgG on maternal haemoglobin concentrations varied between isolates, but was highly significant for all parasites
Maternal haemoglobin (g/dL)
A
None Low High
Acute Low High
Chronic Low High
Past Low High
p=0·07
p=0·94
p<0·001
p=0·51
(144) (43)
(40) (19)
(34) (57)
(56) (79)
14 11 8 5
B p=0·93
p=0·94
p<0·001
p=0·06
(144) (43)
(40) (20)
(34) (57)
(56) (80)
Birthweight (kg)
4·5 3·5 2·5 1·5 0·5 Placental infection and VSA-PAM-specific IgG concentrations Figure 4: Relation between pregnancy outcome, placental histology, and delivery plasma concentrations of IgG with specificity for parasite-encoded VSA expressed on surface of Busua-CSA-infected erythrocytes CSA=chondroitin sulfate A. Maternal haemoglobin (A) and birthweights (B) in women with low (<1 in 16 positive control pooled plasma from malariaexposed women in their third trimester with high VSA-PAM-specific IgG; figure 3) and high (>1 in 16) concentrations of VSA-PAM-specific IgG shown as medians (centre lines), IQR (boxes), 10% and 90% percentiles (whiskers), and outliers (dots). Donor grouping according to placental histology as in figure 2, and number of samples in each group indicated in parentheses along the bottom of graph. Overall intergroup differences were present for both haemoglobin (p<0·001) and birthweights (P=0·03), and p values for subsequent pair-wise comparisons of distributions of haemoglobin and birthweights in the histology groups shown.
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
expressing pregnancy-associated-malaria-type VSA, and non-significant for isolates expressing VSA unrelated to this malaria (table 2). Multiple linear regression analysis of the data for women with chronic pregnancy-associated malaria, and including primigravidity, sex of offspring, weight of mother, educational level, and concentrations of IgG with specificity for VSA expressed by chondroitin sulfate A-selected and unselected parasites as explanatory variables identified VSA-PAM-specific IgG as the strongest predictor of maternal haemoglobin, with educational level being the only other variable that was significant (table 4 and not shown). Importantly, neither VSA expressed by unselected parasites nor primigravidity significantly affected maternal haemoglobin concentrations (table 4 and not shown). This analysis indicated a difference in maternal haemoglobin concentrations of 15 g/L between women with low (MFI=65·9) and high (MFI=116·9) IgG with specificity for the VSA-PAM expressed by FCR3-chondroitin sulfate A, all other explanatory variables being equal (table 4). Analysis with respect to Busua-chondroitin sulfate A and EJ24 yielded similar results (17 g/L and 20 g/L, respectively). There was no significant relation between VSA-PAM-specific IgG concentrations and maternal haemoglobin in any of the other histology groups (figure 4). VSA-PAM-specific IgG concentrations also affected birthweight in women with chronic pregnancy-associated malaria, and this effect was independent of maternal haemoglobin levels. Women with chronic pregnancyassociated malaria and low VSA-PAM-specific IgG delivered babies who were much smaller than otherwise similar women (table 3 and figure 4). All nine grossly underweight babies (<2300 g) born to mothers with chronic placental P falciparum infection, including four of six stillbirths (not shown), were delivered by women without VSA-PAM IgG. The effect of VSA-PAM-specific IgG on birthweight varied between isolates, but was significant for all parasites expressing PAM-type VSA, and non-significant for isolates expressing VSA unrelated to pregnancy-associated malaria (table 3). Multiple linear regression analysis as described identified VSA-PAM-specific IgG as the strongest predictors of birthweight, with sex of offspring and mother’s weight being other significant variables (table 4 and not shown). This analysis indicated a difference in birthweight of 0·45 kg between women with low and high values of FCR3-chondroitin sulfate A VSA-PAM-specific IgG, all other explanatory variables being equal (table 4). Analysis with respect to Busua-chondroitin sulfate A and EJ24 yielded somewhat lower estimates (0·31 kg in both cases). In all the other groups of women, birthweights were closely similar in women with low and high VSAPAM-specific IgG, respectively (figure 4).
Discussion Our study of 477 pregnant Kenyan women showed that both maternal haemoglobin and birthweight were
287
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
highly dependent on concentrations of VSA-PAM-specific IgG antibodies in women of all parities and with histopathological evidence13 of chronic placental P falciparum infection. By contrast, we did not detect a relation between amounts of these antibodies and clinical outcome measures in women without pregnancyassociated malaria, or in those with acute disease (in which adverse clinical consequences would be unlikely to have developed yet) or previous disease (in which postcontrol compensation would probably have occurred). We have previously shown that the adverse clinical consequences of pregnancy-associated malaria are seen in women of all parities and that they are concentrated among those who harbour placental infections for long periods of time.6 Importantly, however, our results point to a highly probable mechanism of protection—ie, the possession of high values of IgG with specificity for the serologically characteristic VSA (VSA-PAM) that are expressed by P falciparum parasites causing pregnancyassociated malaria. Women with chronic P falciparum infection in the absence of VSA-PAM IgG were considerably more anaemic at delivery than women with similar characteristics and significant VSA-PAM-specific IgG concentrations (mean 76 g/L vs 92 g/L). In an independent study,20 similar to ours, but done in an area of much higher transmission intensity and restricted to secundigravidae, the researchers did not note a significant relation between antibodies inhibiting the adhesion of infected erythrocytes to chondroitin sulfate A, and amounts of maternal haemoglobin, whereas our data are in line with a previous meta-analysis, estimating the effect of P falciparum malaria on maternal anaemia to be 13–19 g/L.21 Similarly, we noted that only women with both chronic placental infection and low VSA-PAM IgG concentrations delivered much smaller babies than did corresponding women with higher values (mean 2·68 kg vs 2·94 kg). It is noteworthy that the distribution of birthweights was not normal since all grossly underweight babies (<2300 g) in women with chronic placental infection were delivered by women with low VSA-PAM IgG concentrations. Furthermore, the estimated difference in birthweight between offspring of chronically infected women with and without VSA-PAM-specific IgG increased when we controlled for potential confounding factors. Duffy and Fried20 estimated an even bigger difference in Kenyan secundigravidae women, whereas previous meta-analyses have resulted in somewhat5 to considerably22 lower estimates than ours. The reasons for these discrepancies are unclear. However, because the relation between gravidity and anti-pregnancy-associated malaria immunity is likely to be strongly dependent on transmission intensity, differences in endemicity between studies are probably implicated, in addition to striking differences in study design and analysis. In general, clinical disease caused by P falciparum occurs in individuals without appreciable amounts of antibody specific for VSA expressed by the infecting parasite.23,24 Similarly, the clinical consequences of pregnancy-associated malaria are borne by those without VSA-PAM-specific IgG, whereas placental parasitaemia seems to be able to persist without overt pathological findings in pregnant women with substantial VSA-PAMspecific immunity. Although we did not have access to autologous parasites, pregnancy-associated-malaria-type VSA expressed by placental and chondroitin sulfate Aselected isolates are antigenically much more similar to each other than are non-pregnancy-associated-malaria VSA12 (and unpublished).
288
RELEVANCE OF THIS PAPER TO PRACTICE BACKGROUND
Pregnancy-associated Plasmodium falciparum malaria is an important cause of morbidity leading to low birthweight and maternal anaemia. Infected erythrocytes found in pregnant women express a different set of antigens compared with those found in men and non pregnant women. Maternal antibodies to these pregnancy-associated infected erythrocytes have been identified, but the relation of these antibodies to pregnancy outcomes is unclear. This paper found higher antibody levels to the pregnancy associated antigens, but not the non pregnancy associated antigens in women with chronic compared with acute pregnancy-associated malaria, which increased with number of pregnancies. However, lower antibody levels to the pregnancy-associated antigens were associated with smaller babies and with more severe maternal anaemia. IMPLICATIONS
In pregnant women, antibodies against an antigen expressed on pregnancy-associated P falciparum infected erythrocytes protected against low birthweight and maternal anaemia. As well as allowing a better understanding of the disease process, the findings suggest an area of investigation for future therapeutic strategies.
Thus, for pregnancy-associated-malaria, as for P falciparum malaria in general, clinical protection seems to rely on the ability to control parasitaemia rather than to eradicate an infection, resolving the apparent paradox that placental infection often persists despite high concentrations of protective VSA-PAM-specific IgG.19 Indeed, previous data support an inverse relation between VSA-PAM-specific IgG and placental parasitemia,10 but more information—not least for VSA-PAM-specific immunity in relation to cumulative placental parasite load during pregnancy—is clearly needed, and represents a future challenge. Additionally, the type of assay being used dictated sample size in our study, which resulted in low statistical power with respect to the comparisons between HIV-1-positive and HIV-1-negative women. Results of several studies have suggested that pregnancy-associated-malaria-induced T-helper-1-type cellular responses are related to adverse pregnancy outcome, whereas T-helper-2-type responses are associated with protection from the disease.25 These and our findings may well represent different sides of the same immunological coin, and, if so, it will be important to determine whether inadequate VSA-specific antibodymediated control of parasitaemia results in excessive inflammation or vice versa. Our investigation provides evidence suggesting a causal relation between VSA-PAM IgG and the clinical consequences of PAM, thus raising hopes for the protective efficacy of VSA-PAM-based vaccination against the disease. Findings from an independent study in Kenya lend support to this inference.20 Until data from prospective studies become available, the apparent causality is strengthened by the absence of effect of IgG with specificity for VSA unrelated to pregancy-associated malaria, notably IgG to VSA expressed by the very isolates (Busua and FCR3) from which two of those isolates in which a strong effect could be detected (Busua-chondroitin sulfate A and FCR3-chondroitin sulfate A, respectively) were derived. The molecular identity of VSA-PAM remains unclear. However, we have recently described substantial upregulation of a highly conserved and distinctly structured var gene in VSA-PAM-expressing parasites, which makes the corresponding PfEMP1 molecule encoded by this gene an attractive candidate.14 Although other PfEMP1 molecules have been implicated,26,27 their importance in relation to pregnancy-associated malaria seems doubtful.28,29
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
The authors are indebted to all the individuals who donated parasite and plasma samples for the study, and to the field staff involved in the sample collection. Kirsten Pihl, Anne Corfitz, and Maiken Christensen are thanked for excellent technical assistance. The study received financial support from the Commission of the European Communities (QLK2CT-2001–01302 PAMVAC), the Danish Development Assistance (ENRECA-104.Dan.8.L.306), and the Danish Medical Research Council (SSVF-22–02–0571). The report was published with the permission of Norbert Peshu, Director of KEMRI.
References 1
2
3
4 5
6
7 8 9
10
11
12 13
Yamada M, Steketee R, Abramowsky C, et al. Plasmodium falciparum associated placental pathology: a light and electron microscopic and immunohistologic study. Am J Trop Med Hyg 1989; 41: 161–68. Fried M, Duffy PE. Adherence of Plasmodium falciparum to chondroitin sulphate A in the human placenta. Science 1996; 272: 1502–04. Achur RN, Valiyaveettil M, Alkhalil A, Ockenhouse CF, Gowda DC. Characterization of proteoglycans of human placenta and identification of unique chondroitin sulfate proteoglycans of the intervillous spaces that mediate the adherence of Plasmodium falciparum-infected erythrocytes to the placenta. J Biol Chem 2000; 275: 40344–56. Brabin BJ. An analysis of malaria in pregnancy in Africa. Bull World Health Organ 1983; 61: 1005–16. Guyatt HL, Snow RW. Malaria in pregnancy as an indirect cause of infant mortality in sub-Saharan Africa. Trans R Soc Trop Med Hyg 2001; 95: 569–76. Shulman CE, Marshall T, Dorman EK, et al. Malaria in pregnancy: adverse effects on haemoglobin levels and birthweight in primigravidae and multigravidae. Trop Med Int Health 2001; 6: 770–78. Gardner MJ, Hall N, Fung E, et al. Genome sequence of the human malaria parasite Plasmodium falciparum. Nature 2002; 419: 498–511. Kyes S, Horrocks P, Newbold C. Antigenic variation at the infected red cell surface in malaria. Ann Rev Microbiol 2001; 55: 673–707. Ricke CH, Staalsoe T, Koram K, et al. Plasma antibodies from malaria-exposed pregnant women recognize variant surface antigens on Plasmodium falciparum-infected erythrocytes in a parity-dependent manner and block parasite adhesion to chondroitin sulphate A. J Immunol 2000; 165: 3309–16. Staalsoe T, Megnekou R, Fievet N, et al. Acquisition and decay of antibodies to pregnancy-associated variant antigens on the surface of Plasmodium falciparum infected erythrocytes that are associated with protection against placental parasitemia. J Infect Dis 2001; 184: 618–26. O’Neill-Dunne I, Achur RN, Agbor-Enoh ST, et al. Graviditydependent production of antibodies that inhibit binding of Plasmodium falciparum-infected erythrocytes to placental chondroitin sulfate proteoglycan during pregnancy. Infect Immun 2001; 69: 7487–92. Fried M, Nosten F, Brockman A, Brabin BT, Duffy PE. Maternal antibodies block malaria. Nature 1998; 395: 851–52. Bulmer JN, Rasheed FN, Francis N, Morrison L, Greenwood BM. Placental malaria. I. Pathological classification. Histopathology 1993; 22: 211–18.
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
14 Salanti A, Staalsoe T, Lavstsen T, et al. Selective upregulation of a single distinctly structured var gene in CSA-adhering Plasmodium falciparum involved in pregnancy-associated malaria. Mol Microbiol 2003; 49: 179–91. 15 Staalsoe T, Giha HA, Dodoo D, Theander TG, Hviid L. Detection of antibodies to variant antigens on Plasmodium falciparum infected erythrocytes by flow cytometry. Cytometry 1999; 35: 329–36. 16 Nguyen-Dinh P, Steketee RW, Greenberg AE, Wirima JJ, Mulenda O, Williams SB. Rapid spontaneous postpartum clearance of Plasmodium falciparum parasitaemia in African women. Lancet 1988; 2: 751–52. 17 Ofori MF, Staalsoe T, Bam V, et al. Expression of variant surface antigens by Plasmodium falciparum parasites in the peripheral blood of clinically immune pregnant women indicates ongoing placental infection. Infect Immun 2003; 71: 1584–86. 18 Steketee RW, Wirima JJ, Bloland PB, et al. Impairment of a pregnant woman’s acquired ability to limit Plasmodium falciparum by infection with human immunodeficiency virus type-1. Am J Trop Med Hyg 1996; 55 (suppl 1): 42–49. 19 Beeson JG, Cooke BM, Rowe JA, Rogerson SJ. Expanding the paradigms of placental malaria. Trends Parasitol 2002; 18: 145–47. 20 Duffy PE, Fried M. Antibodies that inhibit Plasmodium falciparum adhesion to chondroitin sulfate A are associated with increased birth weight and the gestational age of newborns. Infect Immun 2003; 71: 6620–23. 21 Guyatt HL, Snow RW. The epidemiology and burden of Plasmodium falciparum-related anemia among pregnant women in sub-Saharan Africa. Am J Trop Med Hyg 2001; 64: 36–44. 22 Garner P, Gülmezoglu AM. Drugs for preventing malaria-related illness in pregnant women and death in the newborn (Cochrane Review). In: The Cochrane Library, Issue 2, 2003. Oxford: Update Software. 23 Bull PC, Lowe BS, Kortok M, Molyneux CS, Newbold CI, Marsh K. Parasite antigens on the infected red cell are targets for naturally acquired immunity to malaria. Nature Med 1998; 4: 358–60. 24 Ofori MF, Dodoo D, Staalsoe T, et al. Malaria-induced acquisition of antibodies to Plasmodium falciparum variant surface antigens. Infect Immunol 2002; 70: 2982–88. 25 Okoko BJ, Enwere G, Ota MO. The epidemiology and consequences of maternal malaria: a review of immunological basis. Acta Trop 2003; 87: 193–205. 26 Buffet PA, Gamain B, Scheidig C, et al. Plasmodium falciparum domain mediating adhesion to chondroitin sulfate A: a receptor for human placental infection. Proc Natl Acad Sci USA 1999; 96: 12743–48. 27 Reeder JC, Cowman AF, Davern KM, et al. The adhesion of Plasmodium falciparum-infected erythrocytes to chondroitin sulfate A is mediated by P falciparum erythrocyte membrane protein 1. Proc Natl Acad Sci USA 1999; 96: 5198–202. 28 Rowe JA, Kyes SA, Rogerson SJ, Babiker HA, Raza A. Identification of a conserved Plasmodium falciparum var gene implicated in malaria in pregnancy. J Infect Dis 2002; 185: 1207–11. 29 Salanti A, Jensen ATR, Zornig HD, et al. A sub-family of common and highly conserved var genes expressed by CSA-adhering Plasmodium falciparum. Mol Biochem Parasitol 2002; 122: 111–15.
289
For personal use. Only reproduce with permission from The Lancet publishing Group.
Mechanisms of disease
Variant surface antigen-specific IgG and protection against clinical consequences of pregnancy-associated Plasmodium falciparum malaria Trine Staalsoe, Caroline E Shulman, Judith N Bulmer, Ken Kawuondo, Kevin Marsh, Lars Hviid
Summary Background Pregnancy-associated malaria caused by Plasmodium falciparum adherence to chondroitin sulfate A in the placental intervillous space is a major cause of low birthweight and maternal anaemia in areas of endemic P falciparum transmission. Adhesion-blocking antibodies that specifically recognise parasite-encoded variant surface antigens (VSA) are associated with resistance to pregnancyassociated malaria. We looked for a possible relation between VSA-specific antibody concentrations, placental infection, and protection from low birthweight and maternal anaemia. Methods We used flow cytometry to measure VSA-specific IgG concentrations in plasma samples taken during child birth from 477 Kenyan women selected from a cohort of 910 women on the basis of HIV-1 status, gravidity, and placental histology. We measured VSA expressed by one placental P falciparum isolate and two isolates selected or not selected for chondroitin sulfate A adhesiveness in-vitro. Findings Concentrations of plasma IgG specific for VSA, expressed by chondroitin sulfate A-adhering parasites (VSA in pregnancy-associated malaria or vsa-pam), increased with gravidity and were associated with placental histological findings. Women with chronic pregnancy-associated malaria and low or absent VSA-PAM-specific IgG had lower haemoglobin values (reduced by 17 g/L; 95% CI 8·1 –25·2) and delivered smaller babies (birthweight reduced by 0·26 kg; 0·10–0·55) than did corresponding women with high VSA-PAM-specific IgG. No such relation was shown for concentrations of IgG with specificity for non-pregnancy-associated malaria VSA. Interpretation VSA-PAM-specific IgG protects against low birthweight and maternal anaemia. Our data indicate an important mechanism of clinical protection against malaria and raise hope for the clinical effectiveness of a potential VSA-based vaccine against pregnancy-associated malaria. Lancet 2004; 363: 283–89
Centre for Medical Parasitology at Department of Infectious Diseases and Department of Clinical Microbiology, Copenhagen University Hospital (Rigshospitalet) and Institute for Medical Microbiology and Immunology, University of Copenhagen, Copenhagen, Denmark (T Staalsoe PhD, L Hviid PhD); London School of Hygiene and Tropical Medicine, London, UK (C E Shulman MRCGP); Centre for Geographical Medicine Research Coast, Kenya Medical Research Institute, Kilifi, Kenya (C E Shulman MRCGP, K Kawuondo HND, Prof K Marsh FRCP); and Department of Pathology, University of Newcastle upon Tyne, Royal Victoria Infirmary, Newcastle upon Tyne, UK (J N Bulmer MRCPath) Correspondence to: Dr Lars Hviid, Department of Infectious Diseases M7641, Rigshospitalet, Blegdamsvej 9, 2100 Copenhagen, Denmark. (e-mail: [email protected])
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
Introduction In areas of intense parasite transmission, severe Plasmodium falciparum malaria is largely a childhood disease due to gradual acquisition of protective immunity, although pregnant women—and in particular primigravidae— constitute an important exception. PREGNANCY-ASSOCIATED MALARIA, which is characterised by accumulation of infected erythrocytes in the placental intervillous space,1–3 is a major cause of low birthweight and maternal anaemia in areas of endemic parasite transmission.4–6 A detailed understanding of the pathogenesis and immunology of pregnancy-associated malaria has emerged only after results of investigations showing that infected erythrocytes obtained from the placenta of women with malaria are unique in their ability to adhere to chondroitin sulfate A in vitro and in their absence of adhesiveness to other receptor molecules commonly exploited by non-placental infected erythrocytes.2 These findings suggest that parasites responsible for pregnancy-associated malaria express a distinct set of molecules mediating adhesion to receptors located only in placental tissue,3 explaining the sudden susceptibility to malaria in otherwise clinically immune women at the time of their first pregnancy.4 Parasite-encoded VARIANT SURFACE ANTIGENS (VSA) inserted into the membrane of the infected erythrocyte are the primary mediators of the characteristic adhesion of erythrocytes infected by P falciparum. The best characterised of these adhesins is P falciparum erythrocyte membrane protein 1 (PfEMP1); a family of polymorphic, high molecular weight proteins encoded by var genes present at about 60 loci per genome.7 VSA can mediate adhesion to a number of receptors in the host vasculature, including chondroitin sulfate A.8 Several results strengthen the hypothesis that parasites in pregnancy-associated malaria express unique VSA (VSA-PAM) that are antigenically distinct from VSA present on non-placental infected erythrocytes. Thus, men exposed to malaria never have VSA-PAM-specific IgG despite high values of other VSA-specific antibodies, a finding that we have termed sex-specificity.9,10 Also, women exposed to malaria do not acquire VSA-PAM-specific antibodies earlier than halfway through their first pregnancy.10,11 Finally, plasma concentrations of IgG with specificity for VSA-PAM, but not for other VSA, are highly correlated with gravidity in pregnant, malaria-exposed women (parity-dependency).9,10 The ability of VSA-PAM-specific IgG to interfere with chondroitin sulfate A-specific adhesion in vitro,9,10,12 and the association between VSA-PAM IgG values and protection against placental parasitemia,10 indicate that the parity-dependent increase in VSA-PAM-specific IgG concentrations and in resistance to pregnancy-associated malaria are causally related, but reliable evidence has not been forthcoming. We thus set out to investigate the hypothesis that VSA-PAM-specific IgG protects against maternal anaemia and low birthweight, which are the most important clinical consequences of pregnancy-associated malaria. 283
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
GLOSSARY PREGNANCY-ASSOCIATED MALARIA
A specific syndrome seen only in pregnant women and characterised by accumulation of parasitised erythrocytes in the placenta. Such malaria is associated with substantial maternal and perinatal morbidity and mortality. PfEMP1
Plasmodium falciparum erythrocyte membrane protein 1. A family of high molecular weight parasite proteins expressed on the surface of parasitised erythrocytes and mediating their adhesion to a range of host receptors. VARIANT SURFACE ANTIGEN
Parasite-encoded, clonally-variant antigens, inserted into the surface membrane of the infected erythrocyte and thought to be an important target of protective immunity to P falciparum malaria.
VSA-PAM-specific IgG reactivity was used neat, and diluted 1 in 4 and 1 in 16 as positive controls. Sample size was determined by the need to undertake all measurements on a specific parasite isolate in one experiment, imposing an upper limit of about 500 on total plasma sample size. Parasites and in-vitro selection for adhesion to chondroitin sulfate A We used three P falciparum isolates cultured in vitro by standard methodology.10 One isolate (EJ24) was obtained at delivery from the placenta of a woman with pregnancyassociated malaria. The isolation, adhesion properties, and other characteristics of EJ24 are described in detail elsewhere.14 The second isolate (Busua) was obtained from the peripheral blood of a non-immune man, and was
VSA-PAM
A
Variant surface antigen in pregnancy-associated malaria.
NC
EJ24 EJ24
1.
Methods
2. 3–4. 5–7. 8–15. 50
B
75
100
125
NC
150 Busua Busua
1. Gravidity
Plasma samples and study population We studied plasma samples taken during delivery from 477 pregnant women, giving birth between Jan 1, 1996, and July 31, 1997, at Kilifi District Hospital, situated in an area of perennial, seasonal, and hyperendemic parasite transmission on the Kenyan coast.6 All samples were selected from a larger set from 910 pregnant women recruited for a separate investigation.6 Sample selection was done on the basis of HIV-1 status, gravidity (primigravid, 2–3 gravid, or >3 gravid), and placental histology (no malaria infection, acute infection, chronic infection, or past infection).13 We aimed to investigate 25 samples or more from HIV-1-negative women in all 12 categories of gravidity and histology, but could not always do so. We also included 75 HIV-1-positive women from the original investigation to allow assessment of the effect of this infection on VSA-PAM-specific IgG plasma values. Selection of samples from HIV-1-negative women was random within each group. Plasma samples from nine third-trimester pregnant women without malaria exposure were included as negative controls.9 Pooled plasma from malaria-exposed women in their third trimester with high
2. 3–4. 5–7.
8–15. 40
C
55
70
NC
85
100
Busua-CSA Busua-CSA
1. 2. 3–4.
Characteristic
Selected
All
Ethnic group Mijikenda Other Age group (years) Not known 14–19 20–34 35 and above Education and literacy Literate Primary, but illiterate Illiterate, no schooling Coral/stone-walled house House with latrine Ownership of radio Coral/stone house + latrine + radio Body-mass index <20 kg/m2 Pre-eclampsia Severe anemia (Hb <70g/L) Stillbirth Low birthweight (livebirths only)
477
910 349 (73·2%) 128 (26·8%)
p
5–7.
669 (73·5%) 241 (26·5%)
0·89
65 (13·6%) 90 (18·9%) 305 (63·9%) 17 (3·6%)
139 (15·3%) 177 (19·5%) 562 (61·8%) 32 (3·5%)
0·83
610 (67·0%) 71 (7·8%) 229 (25·2%) 320 (35·4%) 643 (70·7%) 562 (62·5%) 246 (27·2%)
0·40
473 476 472 474
910 335 (70·5%) 34 (7·2%) 106 (22·3%) 187 (39·5%)905 341 (71·6%)909 298 (63·1%)899 145 (30·6%)906
469 445 474 477 432
145 (31·1%)896 69 (15·5%) 862 74 (15·6%) 906 45 (9·4%) 910 84 (19·4%) 820
278 (31·0%) 0·97 122 (14·2%) 0·51 125 (13·8%) 0·36 90 (9·9%) 0·86 150 (18·3%) 0·62
477
475
910
50
D
70
MFI=59·6
0·13 0·73 0·82 0·18
Hb=haemoglobin.
Table 1: Principal characteristics of women included in study (selected) and those of the original study cohort6 (all) from which they were selected
284
8–15.
0
90 MFI=90·0
110
130 MFI=124·2
255 0 255 0 FITC fluorescence (channel number)
255
Figure 1: Relation between gravidity and delivery plasma concentrations of IgG (expressed in MFI) with specificity for parasite-encoded VSA expressed on the surface of P falciparum-infected erythrocytes MFI=mean fluorescence index. CSA=chondroitin sulfate A. NC=nonexposed controls. A–C: vertical lines in centre of box=median. Boxes=IQR. Whiskers=10% and 90% percentiles. Dots=outliers. A=IgG with specificity for VSA expressed by placental EJ24 isolate. B=IgG with specificity for VSA expressed by non-selected Busua. C=IgG with specificity for VSA expressed by Busua, after selection for adhesion to CSA (Busua-CSA). D=Histograms of fluorescent Busua-CSA infected erythrocytes after labelling by plasma samples with low (left), medium (centre), and high (right) VSA-specific IgG reactivity, respectively.
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
studied with (Busua-chondroitin sulfate A) or without (Busua) preceding adhesion selection by repeated panning on chondroitin sulfate A in vitro.9,10 The third isolate (FCR3, a long-term in-vitro culture isolate),9 was also investigated with (FCR3-chondroitin sulfate A) or without (FCR3) in-vitro selection for chondroitin sulfate A-adhesion. EJ24, Busua-chondroitin sulfate A, and FCR3-chondroitin sulfate A all showed the sex-specific and parity-dependent plasma IgG recognition characteristic of VSA expressed by parasite isolates related to pregnancyassociated malaria,9,10 whereas unselected Busua and FCR3 did not. The genotypic identities of all isolates were regularly confirmed by genotypic profiling at the polymorphic msp1, msp2, and glurp loci.9 Measurement of variant-specific IgG antibodies Plasma concentrations of VSA-specific IgG were measured by flow cytometry.15 Parasite cultures were enriched for late trophozoite and schizont erythrocytes by exposure to a strong magnetic field,15 labelled by ethidium bromide (to identify infected erythrocytes in flow cytometry data analysis), and sequentially exposed to plasma, secondary goat-anti-human IgG (Dako, Glostrup, Denmark), and tertiary fluorescein isothiocyanateconjugated rabbit-anti-goat Ig (Dako). Flow cytometry
data from 10 000 infected erythrocytes were acquired with a Coulter EPICS XL-MCL instrument (Coulter Electronics, Luton, UK). For all samples, the mean fluorescence index (MFI) was recorded as a measure of VSA-specific IgG reactivity.15 Samples in which forwardscatter and ethidium bromide analysis revealed technical problems were excluded from the analyses. Statistical analysis We assessed differences in proportions by 2 analysis, and group-wise differences by t test, Kruskal-Wallis, one-way ANOVA on ranks, and Mann-Whitney rank-sum test as appropriate. Parity-dependency was evaluated by Spearman’s rank-order correlation analysis with CIs calculated as described. We used a two-sample Kolmogorov-Smirnov test to compare distributions of VSA-specific IgG after correction for differences in data location and dispersion. The effects of potential confounding variables were assessed by multiple linear regression analysis with Stata (version 7.0).
A
(1 in 16)(1 in 4) (1 in 1) 0·20 EJ24
0·15 A NC
EJ24
0·10
None 0·05
Acute Chronic
0·00 50
Past
B 75
100
125
Placental infection
Busua
None Acute Chronic Past 40
55
70
85
100
C NC
Busua-CSA
None
140 (1 in 1)
0·20
150
B NC
110
(1 in 16) (1 in 4) Proportion of samples
50
80
Busua
0·15 0·10 0·05 0·00 39
C
57
75
93
(1 in 16) (1 in 4) (1 in 1) 0·20 Busua-CSA
0·15
Acute
0·10
Chronic 0·05
Past 50
70
90
110
130
0·00 44
68
92
116
FITC mean fluorescence index (channel number) FITC mean fluorescence index (channel number) Figure 2: Relation between placental histological findings and delivery plasma concentrations of IgG with specificity for parasite-encoded VSA expressed on surface of P falciparuminfected erythrocytes. CSA=chondroitin sulfate A. A–C: Lines in centre of box=median. Boxes=IQR. Whiskers=10% and 90% percentiles. Dots=outliers. A=IgG with specificity for VSA expressed by placental EJ24 isolate. B=IgG with specificity for VSA expressed by non-selected Busua. C=IgG with specificity for VSA expressed by Busua, after selection for adhesion to CSA (Busua-CSA).
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
Figure 3: Distribution of VSA-specific IgG plasma concentrations CSA=chondroitin sulfate A. Absolute values in nine non-exposed women shown as dots; proportions of values in 477 malaria-exposed study women shown as bars. Antibody concentrations in a four-fold titred pool of highly VSA-PAM-reactive plasma samples shown along the top of graphs. A=IgG with specificity for VSA expressed by placental EJ24 isolate. B=IgG with specificity for VSA expressed by non-selected Busua. C=IgG with specificity for VSA expressed by Busua, after selection for adhesion to CSA (Busua-CSA).
285
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
VSA-PAM IgG concentration (g/L)
EJ24 Busua-CSA FCR3-CSA ⭓2 VSA IgG high‡ Busua FCR3
Low*
High*
Difference†
p
77·5 (53) 75·4 (34) 76·4 (33) 75·7 (40) 80·6 (26) 81·5 (53)
95·4 (37) 91·1 (57) 90·1 (59) 92·4 (52) 86·9 (64) 90·2 (39)
–17·9 (–26·4 to –9·3) –15·7 (–24·5 to –6·8) –13·7 (–22·8 to –4·6) –16·6 (–25·2 to –8·1) –06·3 (–16·5 to 3·9) –08·6 (–17·7 to 0·4)
<0·0001 0·0007 0·004 0·0002 0·22 0·06
CSA=chondroitin sulfate A. *Mean Hb (g/L). Number of individuals in parentheses. †Mean (95% CIs). ‡Individuals with high IgG for at least two of the three isolates expressing VSA–PAM.
Table 2: Maternal haemoglobin concentrations (g/L) at time of delivery in women with chronic pregnancy-associated P falciparum malaria according to their plasma IgG concentration with specificity for type of VSA
Role of the funding source The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
Results Apart from inclusion and exclusion criteria (parity, placental pathology, and HIV-1 status), the characteristics of the 477 women included in the study were similar (p>0·13 in all cases) to those of the cohort (n=910)6 from which they were selected (table 1). This finding indicates that the selection procedure did not introduce any inadvertent bias. Concentrations of plasma IgG recognising intact erythrocytes infected by the placental EJ24 isolate correlated significantly with donor gravidity (rs=0·35 [95% CIs 0·26 to 0·42], p<0·0001, figure 1), but not age (rs=0·06 [–0·04 to 0·15], p=0·21), confirming the characteristic parity-dependency of plasma IgG recognition of VSAPAM-type antigens.9,10 By contrast, there was no significant correlation between gravidity and IgG specific for VSA expressed by the two isolates not related to pregnancyassociated malaria (Busua rs=0·06 [–0·03 to 0·15], p=0·20, figure 1) and FCR3 (rs=0·03 [–0·06 to 0·12], p=0·51). After repeated rounds of selection of Busua and FCR3 for the capacity of infected erythrocytes to adhere to chondroitin sulfate A, both selected sub-lines expressed VSA that caused significantly parity-dependent plasma IgG recognition (Busua-chondroitin sulfate A: rs=0·33 [0·25; 0·41], p<0·0001, figure 1, and FCR3-chondroitin sulfate A: rs=0·31 [0·23 to 0·39], p<0·0001). With respect to placental histology findings, VSA-specific IgG tended to be higher in women in whom pregnancyassociated malaria was either chronic or past than in those in whom it was or was not acute, indicating that such malaria leads to a boosting of immunity to parasites related and unrelated to pregnancy-associated malaria (figure 2). Primigravid women without histological evidence of pregnancy-associated malaria had uniformly low VSAPAM-specific plasma IgG concentrations at delivery, which were not significantly different from those in unexposed pregnant women, suggesting that they had never been exposed to VSA-PAM-expressing parasites despite lifelong
P falciparum exposure (not shown). Primigravid women with acute or chronic pregnancy-associated malaria had low-to-medium VSA-PAM IgG, probably indicating the slow development of a primary response to VSA-PAM in these women.10,11 However, a high proportion of multigravid women had moderate-to-high values, suggesting fast and boostable memory responses (not shown).9–12 Peripheral parasitaemia was not significantly associated with VSA-specific IgG values, maternal haemoglobin, or birthweight. However, peripheral parasitaemia was related to placental histology, being most common in women with chronically infected placentas and least common in those without, or with resolved, placental infection (p<0·0001). These findings confirm earlier results that peripheral parasitaemia in pregnant women living in highly endemic areas generally originates from a placental focus.16,17 HIV-1 infection is an important risk factor for pregnancy in malaria-endemic areas.18 We therefore compared concentrations of VSA-specific IgG in the 75 HIV-1-positive and the 402 HIV-1-negative women in the study cohort. Although values tended to be slightly lower among HIV-1-positive women, the difference was not significant for any of the five parasites tested (p between 0·07 and 0·94). Placental infection can persist despite high VSA-PAMspecific IgG, bringing into question the protectiveness of such antibodies,19 and indeed many women showed histological evidence of continuing pregnancy-associated malaria despite high VSA-PAM-specific IgG (figure 2). To investigate this issue further, we examined the distribution of VSA-specific antibodies in the study cohort. The distribution of IgG specific for the placental isolate EJ24 (figure 3) and both chondroitin sulfate A-selected sub-lines (Busua-chondroitin sulfate Aand FCR3-chondroitin sulfate A) (figure 3 and not shown) all appeared dichotomous, by contrast with the one-peaked distribution seen for plasma IgG reactivity to VSA expressed by parasites not related to pregnancy-associated malaria (Busua and FCR3) (figure 3 and not shown). Two-sample KolmogorovSmirnov tests confirmed that the shapes of the distributions of IgG concentrations with specificity for EJ24 (figure 3) and Busua-chondroitin sulfate A (figure 3) were similar (p=0·91), but different from that of IgG concentrations
Birthweight (kg)
EJ24 Busua-CSA FCR3-CSA ⭓2 VSA-PAM IgG high‡ Busua FCR3
Low*
High*
Difference†
p
2·69 (53) 2·56 (34) 2·62 (33) 2·68 (40) 2·78 (26) 2·70 (53)
2·92 (37) 2·95 (57) 2·92 (59) 2·94 (52) 2·73 (64) 2·91 (39)
–0·23 (–0·52 to –0·03) –0·39 (–0·74 to –0·23) –0·30 (–0·58 to –0·13) –0·26 (–0·55 to –0·10) 0·05 (–0·32 to 0·18) –0·21 (–0·39 to 0·07)
0·03 0·001 0·004 0·007 0·57 0·18
CSA=chondroitin sulfate A. *Median birthweight of offspring; number of individuals in parentheses. †Median (95% CIs). ‡Individuals with high IgG for at least two of the three isolates expressing VSA-PAM.
Table 3: Birthweights of children born to mothers with chronic pregnancy-associated P falciparum malaria according to mothers’ plasma IgG concentration with specificity for type of VSA
286
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
Constant Primigravidity* Sex of offspring† Mother’s weight (kg) Educational level‡ VSA-PAM (FCR3-CSA) (MFI) Non-PAM VSA (FCR3) (MFI)
Coefficient (slope) (95% CI)
p
Coefficient (slope) (95% CI)
p
3·46 (–0·29 to 7·22) 0·11 (–1·00 to 0·77) 0·10 (–0·73 to 0·94) –0·0069 (–0·057 to 0·043) 1·67 (0·56 to 2·78) 0·030 (0·009 to 0·013) 0·021 (–0·007 to 0·049)
0·074 0·81 0·81 0·79 0·004 0·001 0·15
0·61 (–0·53 to 1·75) 0·17 (–0·10 to 0·44) 0·35 (0·09 to 0·60) 0·019 (0·004 to 0·035) 0·164 (–0·17 to 0·50) 0·0088 (0·004 to 0·014) –0·00030 (–0·009 to 0·008)
0·30 0·22 0·0090 0·015 0·35 0·0010 0·95
CSA=chondroitin sulfate A. *Primigravidae (0) or multigravidae (1). †Female (0) or male (1). ‡Completed primary school (1) or not (0).
Table 4: Multiple linear regression analysis of pregnancy outcome in 89 women with chronic pregnancy-associated malaria
with specificity for VSA expressed by the unselected Busua isolate (figure 3) (EJ24: p=0·04; Busua-CSA: P=0·002). Based on this finding, we divided the women into those with low (reactivity less than that of a 1 in 16 dilution of the positive control plasma pool, figure 3) and high (⭓1 in 16) VSA-specific IgG, respectively, to determine the relation between VSA-specific IgG and clinically important pregnancy outcome measures. This procedure was done with respect to each of the five parasites used. Women with chronic pregnancy-associated malaria and low VSA-PAM-specific IgG were substantially more anaemic than were comparable women with adequate concentrations of VSA-PAM-specific IgG (table 2 and figure 4). The estimated effect of VSA-PAM-specific IgG on maternal haemoglobin concentrations varied between isolates, but was highly significant for all parasites
Maternal haemoglobin (g/dL)
A
None Low High
Acute Low High
Chronic Low High
Past Low High
p=0·07
p=0·94
p<0·001
p=0·51
(144) (43)
(40) (19)
(34) (57)
(56) (79)
14 11 8 5
B p=0·93
p=0·94
p<0·001
p=0·06
(144) (43)
(40) (20)
(34) (57)
(56) (80)
Birthweight (kg)
4·5 3·5 2·5 1·5 0·5 Placental infection and VSA-PAM-specific IgG concentrations Figure 4: Relation between pregnancy outcome, placental histology, and delivery plasma concentrations of IgG with specificity for parasite-encoded VSA expressed on surface of Busua-CSA-infected erythrocytes CSA=chondroitin sulfate A. Maternal haemoglobin (A) and birthweights (B) in women with low (<1 in 16 positive control pooled plasma from malariaexposed women in their third trimester with high VSA-PAM-specific IgG; figure 3) and high (>1 in 16) concentrations of VSA-PAM-specific IgG shown as medians (centre lines), IQR (boxes), 10% and 90% percentiles (whiskers), and outliers (dots). Donor grouping according to placental histology as in figure 2, and number of samples in each group indicated in parentheses along the bottom of graph. Overall intergroup differences were present for both haemoglobin (p<0·001) and birthweights (P=0·03), and p values for subsequent pair-wise comparisons of distributions of haemoglobin and birthweights in the histology groups shown.
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
expressing pregnancy-associated-malaria-type VSA, and non-significant for isolates expressing VSA unrelated to this malaria (table 2). Multiple linear regression analysis of the data for women with chronic pregnancy-associated malaria, and including primigravidity, sex of offspring, weight of mother, educational level, and concentrations of IgG with specificity for VSA expressed by chondroitin sulfate A-selected and unselected parasites as explanatory variables identified VSA-PAM-specific IgG as the strongest predictor of maternal haemoglobin, with educational level being the only other variable that was significant (table 4 and not shown). Importantly, neither VSA expressed by unselected parasites nor primigravidity significantly affected maternal haemoglobin concentrations (table 4 and not shown). This analysis indicated a difference in maternal haemoglobin concentrations of 15 g/L between women with low (MFI=65·9) and high (MFI=116·9) IgG with specificity for the VSA-PAM expressed by FCR3-chondroitin sulfate A, all other explanatory variables being equal (table 4). Analysis with respect to Busua-chondroitin sulfate A and EJ24 yielded similar results (17 g/L and 20 g/L, respectively). There was no significant relation between VSA-PAM-specific IgG concentrations and maternal haemoglobin in any of the other histology groups (figure 4). VSA-PAM-specific IgG concentrations also affected birthweight in women with chronic pregnancy-associated malaria, and this effect was independent of maternal haemoglobin levels. Women with chronic pregnancyassociated malaria and low VSA-PAM-specific IgG delivered babies who were much smaller than otherwise similar women (table 3 and figure 4). All nine grossly underweight babies (<2300 g) born to mothers with chronic placental P falciparum infection, including four of six stillbirths (not shown), were delivered by women without VSA-PAM IgG. The effect of VSA-PAM-specific IgG on birthweight varied between isolates, but was significant for all parasites expressing PAM-type VSA, and non-significant for isolates expressing VSA unrelated to pregnancy-associated malaria (table 3). Multiple linear regression analysis as described identified VSA-PAM-specific IgG as the strongest predictors of birthweight, with sex of offspring and mother’s weight being other significant variables (table 4 and not shown). This analysis indicated a difference in birthweight of 0·45 kg between women with low and high values of FCR3-chondroitin sulfate A VSA-PAM-specific IgG, all other explanatory variables being equal (table 4). Analysis with respect to Busua-chondroitin sulfate A and EJ24 yielded somewhat lower estimates (0·31 kg in both cases). In all the other groups of women, birthweights were closely similar in women with low and high VSAPAM-specific IgG, respectively (figure 4).
Discussion Our study of 477 pregnant Kenyan women showed that both maternal haemoglobin and birthweight were
287
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
highly dependent on concentrations of VSA-PAM-specific IgG antibodies in women of all parities and with histopathological evidence13 of chronic placental P falciparum infection. By contrast, we did not detect a relation between amounts of these antibodies and clinical outcome measures in women without pregnancyassociated malaria, or in those with acute disease (in which adverse clinical consequences would be unlikely to have developed yet) or previous disease (in which postcontrol compensation would probably have occurred). We have previously shown that the adverse clinical consequences of pregnancy-associated malaria are seen in women of all parities and that they are concentrated among those who harbour placental infections for long periods of time.6 Importantly, however, our results point to a highly probable mechanism of protection—ie, the possession of high values of IgG with specificity for the serologically characteristic VSA (VSA-PAM) that are expressed by P falciparum parasites causing pregnancyassociated malaria. Women with chronic P falciparum infection in the absence of VSA-PAM IgG were considerably more anaemic at delivery than women with similar characteristics and significant VSA-PAM-specific IgG concentrations (mean 76 g/L vs 92 g/L). In an independent study,20 similar to ours, but done in an area of much higher transmission intensity and restricted to secundigravidae, the researchers did not note a significant relation between antibodies inhibiting the adhesion of infected erythrocytes to chondroitin sulfate A, and amounts of maternal haemoglobin, whereas our data are in line with a previous meta-analysis, estimating the effect of P falciparum malaria on maternal anaemia to be 13–19 g/L.21 Similarly, we noted that only women with both chronic placental infection and low VSA-PAM IgG concentrations delivered much smaller babies than did corresponding women with higher values (mean 2·68 kg vs 2·94 kg). It is noteworthy that the distribution of birthweights was not normal since all grossly underweight babies (<2300 g) in women with chronic placental infection were delivered by women with low VSA-PAM IgG concentrations. Furthermore, the estimated difference in birthweight between offspring of chronically infected women with and without VSA-PAM-specific IgG increased when we controlled for potential confounding factors. Duffy and Fried20 estimated an even bigger difference in Kenyan secundigravidae women, whereas previous meta-analyses have resulted in somewhat5 to considerably22 lower estimates than ours. The reasons for these discrepancies are unclear. However, because the relation between gravidity and anti-pregnancy-associated malaria immunity is likely to be strongly dependent on transmission intensity, differences in endemicity between studies are probably implicated, in addition to striking differences in study design and analysis. In general, clinical disease caused by P falciparum occurs in individuals without appreciable amounts of antibody specific for VSA expressed by the infecting parasite.23,24 Similarly, the clinical consequences of pregnancy-associated malaria are borne by those without VSA-PAM-specific IgG, whereas placental parasitaemia seems to be able to persist without overt pathological findings in pregnant women with substantial VSA-PAMspecific immunity. Although we did not have access to autologous parasites, pregnancy-associated-malaria-type VSA expressed by placental and chondroitin sulfate Aselected isolates are antigenically much more similar to each other than are non-pregnancy-associated-malaria VSA12 (and unpublished).
288
RELEVANCE OF THIS PAPER TO PRACTICE BACKGROUND
Pregnancy-associated Plasmodium falciparum malaria is an important cause of morbidity leading to low birthweight and maternal anaemia. Infected erythrocytes found in pregnant women express a different set of antigens compared with those found in men and non pregnant women. Maternal antibodies to these pregnancy-associated infected erythrocytes have been identified, but the relation of these antibodies to pregnancy outcomes is unclear. This paper found higher antibody levels to the pregnancy associated antigens, but not the non pregnancy associated antigens in women with chronic compared with acute pregnancy-associated malaria, which increased with number of pregnancies. However, lower antibody levels to the pregnancy-associated antigens were associated with smaller babies and with more severe maternal anaemia. IMPLICATIONS
In pregnant women, antibodies against an antigen expressed on pregnancy-associated P falciparum infected erythrocytes protected against low birthweight and maternal anaemia. As well as allowing a better understanding of the disease process, the findings suggest an area of investigation for future therapeutic strategies.
Thus, for pregnancy-associated-malaria, as for P falciparum malaria in general, clinical protection seems to rely on the ability to control parasitaemia rather than to eradicate an infection, resolving the apparent paradox that placental infection often persists despite high concentrations of protective VSA-PAM-specific IgG.19 Indeed, previous data support an inverse relation between VSA-PAM-specific IgG and placental parasitemia,10 but more information—not least for VSA-PAM-specific immunity in relation to cumulative placental parasite load during pregnancy—is clearly needed, and represents a future challenge. Additionally, the type of assay being used dictated sample size in our study, which resulted in low statistical power with respect to the comparisons between HIV-1-positive and HIV-1-negative women. Results of several studies have suggested that pregnancy-associated-malaria-induced T-helper-1-type cellular responses are related to adverse pregnancy outcome, whereas T-helper-2-type responses are associated with protection from the disease.25 These and our findings may well represent different sides of the same immunological coin, and, if so, it will be important to determine whether inadequate VSA-specific antibodymediated control of parasitaemia results in excessive inflammation or vice versa. Our investigation provides evidence suggesting a causal relation between VSA-PAM IgG and the clinical consequences of PAM, thus raising hopes for the protective efficacy of VSA-PAM-based vaccination against the disease. Findings from an independent study in Kenya lend support to this inference.20 Until data from prospective studies become available, the apparent causality is strengthened by the absence of effect of IgG with specificity for VSA unrelated to pregancy-associated malaria, notably IgG to VSA expressed by the very isolates (Busua and FCR3) from which two of those isolates in which a strong effect could be detected (Busua-chondroitin sulfate A and FCR3-chondroitin sulfate A, respectively) were derived. The molecular identity of VSA-PAM remains unclear. However, we have recently described substantial upregulation of a highly conserved and distinctly structured var gene in VSA-PAM-expressing parasites, which makes the corresponding PfEMP1 molecule encoded by this gene an attractive candidate.14 Although other PfEMP1 molecules have been implicated,26,27 their importance in relation to pregnancy-associated malaria seems doubtful.28,29
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet publishing Group.
MECHANISMS OF DISEASE
The authors are indebted to all the individuals who donated parasite and plasma samples for the study, and to the field staff involved in the sample collection. Kirsten Pihl, Anne Corfitz, and Maiken Christensen are thanked for excellent technical assistance. The study received financial support from the Commission of the European Communities (QLK2CT-2001–01302 PAMVAC), the Danish Development Assistance (ENRECA-104.Dan.8.L.306), and the Danish Medical Research Council (SSVF-22–02–0571). The report was published with the permission of Norbert Peshu, Director of KEMRI.
References 1
2
3
4 5
6
7 8 9
10
11
12 13
Yamada M, Steketee R, Abramowsky C, et al. Plasmodium falciparum associated placental pathology: a light and electron microscopic and immunohistologic study. Am J Trop Med Hyg 1989; 41: 161–68. Fried M, Duffy PE. Adherence of Plasmodium falciparum to chondroitin sulphate A in the human placenta. Science 1996; 272: 1502–04. Achur RN, Valiyaveettil M, Alkhalil A, Ockenhouse CF, Gowda DC. Characterization of proteoglycans of human placenta and identification of unique chondroitin sulfate proteoglycans of the intervillous spaces that mediate the adherence of Plasmodium falciparum-infected erythrocytes to the placenta. J Biol Chem 2000; 275: 40344–56. Brabin BJ. An analysis of malaria in pregnancy in Africa. Bull World Health Organ 1983; 61: 1005–16. Guyatt HL, Snow RW. Malaria in pregnancy as an indirect cause of infant mortality in sub-Saharan Africa. Trans R Soc Trop Med Hyg 2001; 95: 569–76. Shulman CE, Marshall T, Dorman EK, et al. Malaria in pregnancy: adverse effects on haemoglobin levels and birthweight in primigravidae and multigravidae. Trop Med Int Health 2001; 6: 770–78. Gardner MJ, Hall N, Fung E, et al. Genome sequence of the human malaria parasite Plasmodium falciparum. Nature 2002; 419: 498–511. Kyes S, Horrocks P, Newbold C. Antigenic variation at the infected red cell surface in malaria. Ann Rev Microbiol 2001; 55: 673–707. Ricke CH, Staalsoe T, Koram K, et al. Plasma antibodies from malaria-exposed pregnant women recognize variant surface antigens on Plasmodium falciparum-infected erythrocytes in a parity-dependent manner and block parasite adhesion to chondroitin sulphate A. J Immunol 2000; 165: 3309–16. Staalsoe T, Megnekou R, Fievet N, et al. Acquisition and decay of antibodies to pregnancy-associated variant antigens on the surface of Plasmodium falciparum infected erythrocytes that are associated with protection against placental parasitemia. J Infect Dis 2001; 184: 618–26. O’Neill-Dunne I, Achur RN, Agbor-Enoh ST, et al. Graviditydependent production of antibodies that inhibit binding of Plasmodium falciparum-infected erythrocytes to placental chondroitin sulfate proteoglycan during pregnancy. Infect Immun 2001; 69: 7487–92. Fried M, Nosten F, Brockman A, Brabin BT, Duffy PE. Maternal antibodies block malaria. Nature 1998; 395: 851–52. Bulmer JN, Rasheed FN, Francis N, Morrison L, Greenwood BM. Placental malaria. I. Pathological classification. Histopathology 1993; 22: 211–18.
THE LANCET • Vol 363 • January 24, 2004 • www.thelancet.com
14 Salanti A, Staalsoe T, Lavstsen T, et al. Selective upregulation of a single distinctly structured var gene in CSA-adhering Plasmodium falciparum involved in pregnancy-associated malaria. Mol Microbiol 2003; 49: 179–91. 15 Staalsoe T, Giha HA, Dodoo D, Theander TG, Hviid L. Detection of antibodies to variant antigens on Plasmodium falciparum infected erythrocytes by flow cytometry. Cytometry 1999; 35: 329–36. 16 Nguyen-Dinh P, Steketee RW, Greenberg AE, Wirima JJ, Mulenda O, Williams SB. Rapid spontaneous postpartum clearance of Plasmodium falciparum parasitaemia in African women. Lancet 1988; 2: 751–52. 17 Ofori MF, Staalsoe T, Bam V, et al. Expression of variant surface antigens by Plasmodium falciparum parasites in the peripheral blood of clinically immune pregnant women indicates ongoing placental infection. Infect Immun 2003; 71: 1584–86. 18 Steketee RW, Wirima JJ, Bloland PB, et al. Impairment of a pregnant woman’s acquired ability to limit Plasmodium falciparum by infection with human immunodeficiency virus type-1. Am J Trop Med Hyg 1996; 55 (suppl 1): 42–49. 19 Beeson JG, Cooke BM, Rowe JA, Rogerson SJ. Expanding the paradigms of placental malaria. Trends Parasitol 2002; 18: 145–47. 20 Duffy PE, Fried M. Antibodies that inhibit Plasmodium falciparum adhesion to chondroitin sulfate A are associated with increased birth weight and the gestational age of newborns. Infect Immun 2003; 71: 6620–23. 21 Guyatt HL, Snow RW. The epidemiology and burden of Plasmodium falciparum-related anemia among pregnant women in sub-Saharan Africa. Am J Trop Med Hyg 2001; 64: 36–44. 22 Garner P, Gülmezoglu AM. Drugs for preventing malaria-related illness in pregnant women and death in the newborn (Cochrane Review). In: The Cochrane Library, Issue 2, 2003. Oxford: Update Software. 23 Bull PC, Lowe BS, Kortok M, Molyneux CS, Newbold CI, Marsh K. Parasite antigens on the infected red cell are targets for naturally acquired immunity to malaria. Nature Med 1998; 4: 358–60. 24 Ofori MF, Dodoo D, Staalsoe T, et al. Malaria-induced acquisition of antibodies to Plasmodium falciparum variant surface antigens. Infect Immunol 2002; 70: 2982–88. 25 Okoko BJ, Enwere G, Ota MO. The epidemiology and consequences of maternal malaria: a review of immunological basis. Acta Trop 2003; 87: 193–205. 26 Buffet PA, Gamain B, Scheidig C, et al. Plasmodium falciparum domain mediating adhesion to chondroitin sulfate A: a receptor for human placental infection. Proc Natl Acad Sci USA 1999; 96: 12743–48. 27 Reeder JC, Cowman AF, Davern KM, et al. The adhesion of Plasmodium falciparum-infected erythrocytes to chondroitin sulfate A is mediated by P falciparum erythrocyte membrane protein 1. Proc Natl Acad Sci USA 1999; 96: 5198–202. 28 Rowe JA, Kyes SA, Rogerson SJ, Babiker HA, Raza A. Identification of a conserved Plasmodium falciparum var gene implicated in malaria in pregnancy. J Infect Dis 2002; 185: 1207–11. 29 Salanti A, Jensen ATR, Zornig HD, et al. A sub-family of common and highly conserved var genes expressed by CSA-adhering Plasmodium falciparum. Mol Biochem Parasitol 2002; 122: 111–15.
289
For personal use. Only reproduce with permission from The Lancet publishing Group.