PLOS ONE ® Cheek for ‘updates G ovenaccess Caton: Dambrun M, DechavaneC, Gigue Brin V, Cand, vet, ta. (2022) Retospectve study of toxoplasmosis prevalence inpregtant women in Benin ands aon wih ml, PLoS ONE 17()e0262018, ipso, ng/10.137tfourre pone naaz018 tor: Gordon Langsky, stu natona ea santé et dela echerene médicale nit Coin, FROOCE Received: May 7, 2021, ‘Accept: Deceiver 15,2021 Published: nua 72002 Peer Review Mistry: PLOS recognizes the benefits of transparency in he pee review process therefore we enable the publication ot aloe conan of per evew and author responses alongside mal published aries. The dori history of his ace is vata here: hips:do:or/1.1371joumal pone 0262018 Copyright© 2022 Dambrun eal Tiss an open acoess aril red under to terms ofthe Creative Commons Atbution Lions, which permis unrest use, stron, and reproduction in any maar, pove he ginal author and source are ceded. Data Aaiailiy Statement: Te data used were either erated tom te quate and ‘unittv easels ro te STOPPAM sty on RESEARCH ARTICLE Retrospective study of toxoplasmosis prevalence in pregnant women in Benin and its relation with malaria Magalie Dambrun"*, Célia Dechavanne'’, Nicolas Guigue®, Valérie Briand™, Tristan Candau’, Nadine Fievet', Murielle Lohezic', Saraniya Manoharan', Nawal Sare*, Firmine Viwami*, Francois Simon*, Sandrine Houzé", Florence Migot-Nabias' 1 Université de Pars, MERIT, IRD, Patis, France, 2 Service de Microbiologe, Assistance Pubique— Heptaux de Paris, Groupe Hospitalier Saint-Louis Larooisicre-Femand: Wid, Paris, France, 3 Service do Parastoioge, Assistance Pubique—Hoplaux oe Pans, Groupe Hospital Bichat Clause-Bernard, Pars, France, 4 Insitut de Recherche Cinique du Bénin, Cotonou, Bénin 1 Current address: Malados Infctiouses dans les Pays ressources Linitées (IDLIC tear), INSERM U2! Institut de Recherche pour le Développement, Université de Bordeaux, Bordeaux, France * magale dambrun@u parsir Abstract Background Globally distributed with variable prevalence depending on geography, toxoplasmosis is @ zoonosis caused by an obligate intracellular protozoan parasite, Toxoplasma gondi. This disease is usually benign but poses a risk for immunocompromised people and for new- bons of mothers with a primary infection during pregnancy because of the risk of congenital toxoplasmosis (CT). CT can cause severe damage to fetuses-newboms. To our knowledge, no study has been conducted in sub-Saharan Arica on toxoplasmosis seroprevalence, seroconversion and CT in a large longitudinal cohort and furthermore, no observation has been made of potential relationships with malaria Methods We performed a retrospective toxoplasmosis serological study using available samples from a large cohort of 1,037 pregnant women who were enrolled in a malaria follow-up dur- ing the 2008-2010 period in a rural area in Benin. We also used some existing data to inves- tigate potential relationships between the maternal toxoplasmosis serological status and recorded malaria infections Results Toxoplasmosis seroprevalence, seroconversion and CT rates were 52.6%, 9.4% and 0.2%, respectively, reflecting the population situation of toxoplasmosis, without targeted medical intervention. The education level influences the toxoplasmosis serological status of women, with women with litle or no formal education have greater immunity than others. Surpris- ingly, toxoplasmosis seropositive pregnant women tended to present lower malaria infection during pregnancy (number) or at delivery (presence) and to have lower IgG levels to PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 1720PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin ala o generated by this study on toxoplasmosis. The newly acquied dat on ‘piss, inked infomation on inctsion ches in th STOPPAD st. are arate rom the Open Se Framework dtaase (hps/os louey,.D0110.176080SFIOLTSEY. Regarding the mata data tom the STOPPAM Prost hey are not cure pub avaiable Hower, claboratns ae encouraged and ary resarcher intrested in explring data may coat ‘recy the researcher responsible fr he cal SSTOPPAM comile:NcaeTukue Wéam (STOPPAM dats, ema: nase namics) Funding: This study was suppor by nit do Maecin et Eidémolopi Ataule (MEA) (tpn nthe form ofa grant io FAN, [osa201Amba}, Competing interests: Te authors have erred thatno competing irs exist Plasmodium falciparum Apical Membrane Antigen 1, compared to toxoplasmosis seronega- tive women. Conclusions The high toxoplasmosis seroprevalence indicates that prevention against this parasite remains important to deploy and must be accessible and understandable to and for allindi- Viduals (educated and non-educated). A potential protective role against malaria conferred by a preexisting toxoplasmosis infection needs to be explored more precisely to examine the environmental, parasitic and/or immune aspects. Introduction Toxoplasmosis i one ofthe most common parasitic diseases caused by Toxoplasma gondii an intracellular protozoan parasite belonging to the phylum Apicomplexa. This parasite can infect all warm-blooded animals. Approximately 30% ofthe human population is infected world wide, bt the prevalence of infection is variable between both arcas and communities, accord- ing to climate, lifestyle and diet [1]. Toxoplasmosis isthe third burden of foodborne illness in Earope for WHO [2], while the American Centers for Disease Control and Prevention (CDC) considers ita neglected disease due to the scant attention paid to it in terms of surveillance, prevention and/or teatment (3,1. Toxoplasmosis i usually benign. However, it has adverse consequences in immunocom- promised people and fetuses-newborns from women who have contracted toxoplasmosis dur ing their pregnancy. Congenital toxoplasmosis (CT) is the result of mother-to-child vertical transmission ofthe parasite, and the risk increases with gestational age, The severity of the encountered disorders is inversely related to the pregoancy period at maternal infection. The disorders may range from severe abnormalities (mainly neurological and ophthalmic) or abor- tion during the frst trimester of gestation, to manifestations of variable severity during the sec ond trimester and asymptomatic traces at the third trimester [1, 5 6]. Other risk factors, including immune and genetic host factors as well as characteristics ofthe T, gondii strain, are involved inthe severity ofthe disease [6]. In T gordi non-immune pregnant women, teat- ‘ment is recommended as soon as the toxoplasmosis contamination i proven in order to reduce the risk of parasite transmission tothe fetus and of sequelae in neonates. The contami nation can be detected indirectly by the seroconversion ofthe woman, which is defined by the appearance of specific IgG directed tT. gondii, Treatments are given depending on the stage of pregnancy. For example, in France, spiramycine 9 million units/day is recommended dur- ing the fist twimester [6,7] whereas pyrimethamine S0mg/day associated with sulfadiazine Sxlg/day and folini acid SOmg/week are recommended from the second trimester tothe end. of pregnancy [6 The most effective way to detect seroconversion during pregnancy, and then to apply an antiparasitic treatment to prevent CT is to implement a national serological screening pro- gram of pregnant women. Although this practice would be effective and less financially bur- densome than monitoring pregnancy [8-10], itis rarely instituted globally (5,6, 11] and therefore hygiene and diet guidelines remain the most used means of ist line prevention [6, 12] The lack of systematic serological follow-up of pregnant women limits the informative data on actual seroprevalence and the incidence of seroconversion, especially as the infection is asymptomatic, potentially hiding a considerable number of CT cases. Although many PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 2/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin countries lack adequate data, some epidemiological meta-analyses have been performed, lead. ing to global estimates of 33.8% for acquired toxoplasmosis among pregnant women [13], (0.69%-1.1% for the seroconversion rate [4] and 190,100 cases/year (1.5 cases per 1,000 births) for the overall CT incidence [15] Although data on human toxoplasmosis are less numerous in Africa than in some other countries [13, 16] they show that the seroprevalence rae, recently estimated at a global vale of 48.7%, undergoes a great variability between countries {13). Indeed, the seroprevalence rates recorded during the last decade in West Africa ranged from 20% to 70%, depending on countries [17-20], despite different methodologies used. A downward trend is recorded com pared to previous periods, thus the toxoplasmosis seroprevalence decreased from 74% in 2006 {21 to 40% in 2016 [19] in Ghana, and from 54% in 1993 [22] to 49% in 2012 [23] in Cotonou, the economic capital of Benin. A recent meta-analysis in Benin indicated a toxoplasmosis ‘mean seroprevalence of 47% among pregnant women over the 1990-2018 period [24 Inter regional variability is also observed in Benin with seroprevalence varying, from 30% in 2011 in the Atacora department in the north [25] to 49% in 2012 in the Littoral department in the south including Cotonou [23], and 36% in 2016 inthe Atlantic department inthe west [26] As with many other sub-Saharan countries, Benin is also impacted by malaria. Ths disease is caused by Plasmodium spp., another Apicomplexa parasite. The main species involved is Plasmon falciparum, known to impair the fetal development in case of pregnancy-associ- ated malaria, The latest WHO report indicated that 11 million pregnant women living in sub- Saharan malaria-endemic areas were exposed in 2018 to malaria infection. This resulted in 1696 ofthe eases of low-birth-weight children recorded in these areas [27]. WHO recommends an intermittent preventive treatment against malaria during pregnancy (IPTp) using.a sulfa- doxine-pyrimethamine (SP) combination (IPTp-SP), the same molecules as for toxoplasmosis treatment but dosed differently. Since October 2012, IPTp-SP has been recommended as soon as possible from the second trimester of pregnancy, with at least three doses spaced one month, until delivery [27 0 date, there is no description of toxoplasmosis seroprevalence, seroconversion and CT rates on large cohorts of pregnant women and their infants in West Africa. To our knowledge, only one study performed in Ghana addressed the diagnosis of Plasmodium and Toxoplasma co-infections at delivery among fewer than 100 pregnant women recruited during the third tr imester of pregnancy [28] In tis study, we provide additional information on toxoplasmosis during pregnancy in Benin by performing a retrospective serological study using plasmas of pregnant women from a previous study on gestational malaria. This offered a unique opportunity to explore the potential relationships between maternal toxoplasmosis serological status and Plasmodium malaria and IPT-SP in pregnancy. Material and methods Population group The women under study participated in the “Strategy ‘To Prevent Pregnancy-Associated Malaria” (STOPPAM) project, implemented fom 2008 to 2010 in southern Benin, where the climate is subtropical [29, 30). The STOPPAM project was designed to explore the relation- ships between the timing of malaria infection during pregnancy and clinical and immunologi cal consequences for both mothers and infants Women were enrolled in three dispensaries from one semi-rural and two rural sites, in the district of Comé in the Mono province, located 70km west of Cotonou. The main occupations of the population are farming, fishing and trading, The inclusion criteria were: living for more PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 3/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin than 6 months within 15 km from the dispensary, having a gestational age under 24 weeks and having planned to deliver at the hospital. A total of 1,037 women were enrolled, and 982 were followed-up through pregnancy, including 891 until delivery [23] At each visit (eg, monthly antenatal care (ANC) visit incase of an emergency and at deliv: cry) clinical and obstetrical data aswell as biological samples (inchiding blood samples which were used for this study) were collected. The STOPPAM project followed WHO recommendations in effect at that time, so women received IPTp, iron and folic acid as per national guidelines. When diagnosed malaria infected, women were treated with quinine o SP following the recommended WHO process 9} Design of the serological study on toxoplasmosis Determination of the toxoplasmosis seroprevalence rate. Serology against T. gondi was carried out on the available ethylenediaminetetraacetic acid (EDA) frozen plasma samples obtained at each woman's enrollment into the STOPPAM project (Pig, 1), either atthe frst ANC visit or atthe consultation following it, usually after one month. ToRC (Toxoplasmosis, Rubella, Cytomegalovirus) IgG on the BioPlex” 2200 System (Bio-Rad) was assessed accord ing to the manufacturer's recommendations. It isa multiplex immunoassay for quantitative detection of anti-T. gondii and anti-rubella IgG and for qualitative detection of anti-CMV (cytomegalovirus) I ina single reaction from a single serum or plasma sample. Anti-rubella and anti-CMV IgG results are not presented here. A specific software associated withthe Bio Plex 2200 automated analyzer (Bio-Rad) was used, and T. gor results were expressed according tothe following thresholds esults <9 IU. were negative, those between 10-11 TU/mL were doubtfal, and results > 12 1U/mL were positive. Incase of positive results atthe inclusion, the woman was considered as immunized and no further IgG assay was performed. Determination of the toxoplasmosis seroconversion rate during pregnancy. For oxo plasmoss seronegative women at inclusion, a measurement of anti-T gondii IgG was per formed on available samples corresponding to the end ofthe study, usually at delivery, or the previous consultation not exceeding one month before delivery (Fg 1) Platelia” TOXO IgG (Gio-Rad), an indirect ELISA immunoassay for quantitative determination of anti-T. gondii 1gG in human serum or plasma, was used according to the manufacturer’ recommendations. The absorbance was read at 450/620nm on Asys UVM 340 spectrophotometer (Biochrom). Results corresponding to Ig titers <6 IU/ml. were negative; they were equivocal for IgG titers comprised between 6 and 9 (excluded) IU/ml. and positive for IgG titers > 9 1U/ml. When the test was positive at the end of pregnancy for a woman previously diagnosed as seronegative, Platelia” TOXO IgG was carried simultaneously on both inclusion and exit study plasma samples ofeach individual to check the transition from negativity to positivity using the same immunoassay. For the raze women with discrepant results on the sample ofthe end ofthe study (ie, the plasma of the end of the study was tested positive for the first assay and then negative during the second run, both runs being assessed with Platelia” TOXO IgG), LDBIO Toxo I! 1gG confirmation (LDBIO Diagnostics) was used according to the manufac turers recommendations. This immunoblot qualitative assay i proposed to confirm positive ‘or equivocal results obtained by classical serological tests, Presence of specific anti-T. gondit IgG was confirmed ifat least three bands were observable at either 30, 31, 33, 0 or 45 kDa, inluding the band at 30 kDa in all cases. Pregnant women who evolved from negative to positive anti-T. gondii IgG between inclu sion and exit ofthe study were analyzed more specifically. Namely, all available plasma samples that had been sequentially collected at each visit (an average of six visits) during pregnancy PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 4/20PLOS ONE Toxoplasmosis and malaria during pregnancy in Benin Inclusion 1,037 pregnantwomen Toxoplasmosis seroprevalence 58 PSs' depleted determination 979 PSs available* 512 PSs? 462 PSs” 5 technical failures® Toxo POS Toxo NEG® End ofthe study 112 PSs depleted Toxoplasmosis seroconversion and CT 350 PSs available® determination 12 PSs 338 PSs Toxo POS? Toxo NEG? K— 3 mother-and-cord PS pairsdepleted 9 mother-and-cord PS pairs available® 2cT 7 withoutCT Fig. low chart of the seroical study om toxoplseni. P'S = plasma sample. * asa lama samples at ncn, oF dig the fst anenaal care ofthe neat vit pot exceeding one moat ale the at ne. died and wnt pas samples Gee Materia od Methods" Methodalgial constraints" poor quality pasa samples wih to mac cll debris scroll topless resuls postive (Toao POS) or negative (Toxo NEG) avaiable plans samples corresponding to the tend ofthe sty sally a delivery or dring «previ st not exceeding one month ble diver." CT ~ congenital toxoplasmosis. ip:6ior/10.137 pura pore 0262018 goot were tested in a single experiment with Platelia” TOXO IgG, together with the initial analyzed samples (inclusion and exit) to evaluate the seroconversion period. The determination was refined by detecting anti-T. gondii IgM on the plasma samples corresponding to the period. PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 5/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin that circumscribed the IgG seroconversion. For this purpose, the qualitative Platelia” TOXO IgM (Bio-Rad) detection assay was used, according to the manufacturer's recommendations. The absorbance was read at 450/620nm under the same conditions as Platelia” TOXO IgG. Results were expressed asa ratio, which may be negative (<0.8) equivocal (from 0.8 to <1) or positive (> 1) ‘The IgG avidity test described by Robert-Gangneux etal 31] was applied using Platelia™ TOXO IgG (Bio-Rad) for one woman (pregnant woman #1, or PW1) with equivocal IgG results throughout the pregnancy. This method allows for the determination of an avidity index expressed in percentage depending on the age of infection: a high avidity index indicates, a formerly acquired immunity (> 10 months) and a lower avidity index indicates a possible recently contracted toxoplasmosis infection. Then itis even possible to estimate the period of Infection tobe les than 3 months, 3 to 5 months and 5 to 10 months Detection of cases of Congenital Toxoplasmosis (CT). ‘The detection of CT in newborns was made for the group of women who had seroconverted. Qualitative Platelia” TOXO IgM (Bio-Rad) detection assay was used on cord blood plasmas, according to the manufacturer's ree- ‘ommendations. Also, an immunoblot toxoplasmosis assay (Toxoplasma Western Blot IgG IgM, LDBIO Diagnostics) was performed for the comparison of mother-and-cord immunological 1gG profiles. Maternal plasma samples (from circulating blood at delivery) and fetal plasma samples (from cord blood) vere used. According to the manufacturer's recommendations, ‘maternal and cord plasma samples were incubated on two separate and contiguous strips from the same nitrocellulose transfer membrane on which T. gondii antigens separated by electro- phoresis were bound. A polyclonal goat anti-human IgG coupled with alkaline phosphatase was, used as conjugate, CI’is suggested by the presence of a band in the cord but not the maternal plasma ata molecular weight less than 120 kDa, Because of insufficient plasma quantities, asim ilar immunoblot toxoplasmosis assay using an IgM conjugate was not performed. Methodological constraints This sero-epidemiological study was constrained by sample availability, quantity and quality. Firs, some plasma samples were flly utilized following the initial malaria research program (Gig 1). Then, to avoid damaging the analysis automat for the ToRC assay, due to insufficient volume or too high viscosity, a number of enrollment plasma samples were diluted 1/3 in washing butfer, as recommended by the manufacturer. More precisely, this adaptation con- cerned 348 out ofthe 979 processed samples (35.5%). Five of them enced in technical failure, Without applying any dilution corrective factor, these samples yielded 189 seropositive and 154 seronegative results. The interpretation ofthe seronegative results could have been tricky, due tothe 0-9 scale of values corresponding to the negativity threshold of the test (<9 1U/ mL), but matching with the serological results at study ext (realized with Platelia” TOXO IgG on non-diluted samples) allowed for confirmation of seronegativity for 105/154 of them. The remaining 49 plasma samples at exit were not available because having been depleted: it was decided to consider these 49 plasma samples as seronegative, insofar as their inclusion in or excision from the analysis had little influence on the seronegativity rate of the entre cohort (1 = 462/974, 47.496 vs, n= 413/925, 44.6%). Statistical analysis Firstly, we calculated the prevalence of maternal toxoplasmosis based on IgG levels to T. gondii atthe start ofthe study: IgG titers (UI/mL) were transformed into binary variables (seronega. tive/scropositive) according to the thresholds described above. Toxoplasmosis seroconversion was defined as the transition from seronegativity at inclusion to seropositivity at the end of the PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 6/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin follow-up. The timing of seroconversion, as well as maternal and newborn characteristics in case of seroconversion were described. Finaly, the prevalence of CT (95% CD), defined as mother-and-cord immunological IgG different profiles, was calculated Maternal and newborn characteristics asociated with T. gondii serological maternal status at inclusion were determined using chi-square test or t-tests For these analyses, women who had seroconverted during pregnancy (n = 12) were excluded from these analyses and were considered asa separate group. ‘The following characteristics were tested for their possible association with serological toxo- plasmosis maternal status: 1) scio-epidemiological data: village of residence, season at enrol: ‘ment, mother’s education level, bednet possession and number of medical visits (ANC and. emergency) during follow-up; 2) maternal clinical data: age, gravity, gestational age at enrol- ‘ment and at delivery determined by ultrasound [32]; and 3) neonatal data: prematurity (<37 ‘weeks gestation), stillbirth, birth weight and low birth weight (<2,500 grams) Secondly, we assessed the association between T. gondii serological maternal status at inclu sion and maternal malaria infection (at inclusion, during pregnancy and at delivery). Malaria infection during pregnancy was defined as atleast one positive thick blood smear (1'BS) during [ANC and emergency visits. Malaria infection at delivery was considered positive when para- sites were detected either in peripheral or placental blood. A logistic univariate analysis was, performed to assess this association, and following multivariate logistic model adjusted on ‘maternal socio-demographic and clinical characteristics was conducted to ensure an associa tion existed, At least, the mean of maternal plasma antibody levels to the Apical Membrane Antigen | of P. falciparum (anti-P{AMAL IgG), both at inclusion and delivery, was log-trans- formed and compared according tothe T. gondii serological maternal status at inclusion. IgG directed to PYAMAL, which isa merozoite antigen from the erythrocytic asexual stages of P. {falciparum and also a conserved apicomplexan protein, were measured by Enzyme-Linked, ImmunoSorbent Assay (ELISA) as previously described [33]. Associations were then assessed using a linear univariate analysis followed by a multivariate linear regression model to account for potential confounding factors such as maternal socio- demographic and clinical character- istics. The differences of anti-P{AMAI IgG levels between inclusion and delivery were estab lished using a multivariate mixed model where antibody levels at inclusion and at delivery were paired for each woman. Data were analyzed with Stata™ Software, Version 13 (StatCorp LP, College Station, TX, USA) and the graphics were done using Graph Pad Prism (Version 8.1.2) Ethics The STOPPAM project was approved by the ethics committees ofthe Research Institute for Development (IRD) in France and of the Science and Health Faculty (University of Abomey. Calavi) in Benin, Informed written consent with the possibility of withdrawing from the study at any time was obtained from all women before enrollment, The retrospective sero-epidemio- logical investigation of toxoplasmosis during pregnancy performed on the STOPPAM cohort \was approved by the STOPPAM project committee. All the methods were performed in accor- dance with the institutional guidelines and regulations pertaining to research involving humans. Results ‘Toxoplasmosis seroprevalence, seroconversion and CT ‘The different steps of the serological survey are presented in Fig 1. Among plasma samples col lected at the inclusion of 1,037 pregnant women in the STOPPAM study, 979 samples from the first visit were available forthe determination of toxoplasmosis seroprevalence, Excluding PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 7120PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin S samples due to technical failure, 512 and 462 women were found serologically positive and negative to toxoplasmosis, respectively. Therefore, the toxoplasmosis IgG seroprevalence was 52.6% (512/974). Among the 462 toxoplasmosis seronegative women at enrolment, 350 sam- ples at delivery (or the preceding period when plasma samples at delivery were depleted) corre sponding to 75.7% of them, were available to establish the toxoplasmosis seroconversion rate Twelve women were found to have contracted a primo infection by T. gondii between these two measurement points, their serological IgG toxoplasmosis status having changed from neg ative to positive. Therefore a toxoplasmosis seroconversion rate of 3.4% (12/350, 95% Cl {0.0178:0.059)) was observed when considering seronegative pregnant women with available plasma samples at delivery. Among the 12 cases identified with seroconversion during pre nancy, 9 mother-and-cord plasma pairs were available for highlighting potential cases of CT. Since 2 CT cases could be observed only based on IgG immunoblot results (qualitative Plate- lia TOXO IgM cord blood results being negative and IgM immunoblotting could not be per formed), the congenital toxoplasmosis rate was 0.2%, corresponding to 2/835 lve births (95% 10.0003; 0.086)), ic, a ratio of 2.39 per 1,000 live births, among the initial group of 974 women with known toxoplasmosis serological status; knowing that 19 births were twins, 163 deliveries remained unrecorded, for 816 women (i, 835-19) among the whole study group 0979 women. These 163 situations were distributed into 80 women lost follow-up or hav- ing stopped the follow-up (49.1%), 35 stillbirths (21.5%), 32 miscarriages (19.6%) and 16 women whose pregnancy has been invalidated (9.8%). Women’s and newborns’ characteristics according to their toxoplasmosis serological profile Environmental and clinical characteristics recorded during the STOPPAM study forthe moth ers and their newborns were used for the present analysis. It appears that . gondii seropositive women were older (27.6 + 6.2 years vs. 25.1 46:0 years: P<0,001) and had higher graviity (87 £2295 3.1 + 1.8; P= (1003) than those who remained seronegative all along their preg- nancy (Table 1). Gestational age at delivery did not differ between the two geoups. Regarding living conditions, no difference was observed between groups of women regarding cither their rural or semi-rural location or their bednet possession. Also, no difference regarding presence of anemia was evidenced betwreen groups. Finally, T. gondif seropositive women had com- pleted a lower education level than T. gondii seronegative ones (P = 0.006). On their own, newborns did not differ in terms of prematurity, silbieth, mean beth weight or low birth weight prevalence according to the toxoplasmosis srological status of their moth- cers during pregnancy (Table 2). Seroconversion maternal group and infants ‘able 1 also presents the clinical data of women who contracted toxoplasmosis during preg nancy (n = 12) Their mean age at inclusion was 286 + 5.5 yeas, their mean number of previ ous pregnancies was 38 + 19; they were enrolled at 15.9+5.9 gestation weeks and gave bieth at 39.9 1.6 gestation weeks. Por 7 out ofthese 12 women, plasma samples at inclusion were diluted 1/3 fr the ToRC assay, and their negativity confirmed later using the Pltelia” TOXO 1gG assay on non-luted plasma samples. Among infants born to this seroconversion mater- nal group, one case of prematurity and two cases of low birth weight were recorded (Table 2). Ultrasonography performed during the follow-up did not reveal any abnormality in fetuses suspected of CT Since an average of six blood simples were collected for this group during pregnancy, this allowed overtime ant-T. gondii IgG measurements for an individual estimate of the PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 8/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin ‘Table 1, Clinical and environmental characteristics of pregnant women according to their T. ond serological status Serological toxoplasmosis status inlsion Materalcharactesisin P= | T. gondit seroconverting a | Mean(esD)or% | a | Mean (=8D)or% a Mean(2SD)0r% Sos* | 7o6(02) | 3° 25460) | <000 | 1228655) sa 37629) 50 sei) | ems | ast) Primigest 7 15.0% ‘ot 24% sooo 1 43% Seconds 0. 215% 07 29% 2 167% Muli 35 035% 2a s3a% 2 750% station age a enrolment (weeks)* s2 161 (247) 450 lesa) aos | 2 159(e 59) station age at diver (wecks) 440% | a9 29) west | eae ra ost | nt 91) Living te Ruralstes(Akodchs and Ounleme Pedsh) 289 565% 261 sae os 8 ere Semi-rurl ste (Come) 23 35% 169 so0% 4 3338 Matera education None 307 0% 28 507% 0.006 8 m6 Paral primary sa 193% 07 29% 4 saa (Complete primary ay 66% 2 93% Beyond primary 6 123% 2 16.2% Number af visits (ANC + urgency) sa sa1(222) 50 sne2y | aw | sais) Season at colment Dey ses 22 oe 26 59% ov 2 1678 Not ry season 300 skh a4 si a a8 Bainet possession Ye. 6 324% 19 sine os 4 a8 No 6 70% st cous a 67% tp SP dine Nodose 36 70% 26 a8 os | os Onedose 38 74% 8 7% ° o Two does os 50% set 53% 2 100" Taree doses 3 068 1 02% o om ‘Atindsion s1509" ou amas or on | 6 sa ‘Adelivery vse 9% 295° 468% on | ar 28 * Pralveofthe test * orch-squared et P< 005inbol available valves out F512 women, “avaiable values out of 50 women, “gestational age determined by ultrasound, “avaiable values out of 12 women ips: on/10.137 Yura pone 02620181001 seroconversion period, as illustrated in Fig 2. Anti-T. gondii IgM levels were measured on sam- ples circumscribing the seroconversion time-point, identifiable by the change of anti-T. gondii IgG results from negative to positive, in order to strengthen the diagnostic argument, Results were negative in all cases except one for which IgM levels changed from borderline to positive (PWS). Three women may have seroconverted towards the end of the first trimester of preg nancy (PW1, PW2 and PW3). For one of them (PW), ant-T. gondii IgG doubtful results PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 9/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin “Table 2. Clinical characteristics of newborns acording to the T: gol serological stats oftheir mother during pregnancy. [Newborn linia characteristics T. gon seropositive mothers T.gondifscroncgative mothers | PT. gondii seroconvertng mothers (a= 40"1512) (a= 9541450) (anv) = Mean(sSD)or% | m|-Mean(= SD) or =| Mean(=SD) or Promaturiy (= 37 wees) a“ 100% a 104% ot oa sails 21 9% ut 35% ox oo % Birth weight (| 4355 2953 (536) sas! 2961 (£903) oss 3040 (8631) ow bith weight (= 25009) £ a 42% # 122% oe 2 142% ‘available newbora vues *. Pale ofthe univariate logistic regesion, the values mentioned here are indicative 1 eiing value Sing vale, "ro missing vals * calculated onthe numberof infants with documented bith weight psotor/10.17 tourna pone 02620181002 needed an assessment using an IgG avidity test. Low and crescent results from the first to the third measure led to a broad estimate for the seroconversion period covering the end of the fist trimester and the whole second trimester (Fig 2). Two women were infected during the 6" month (PW4 and PWS), and their newborns had low birth weight, including one (PWS) \who was born prematurely with no apparent ink with toxoplasmosis, Seven women (from PW6 to PW12) seroconverted during the third trimester of pregnancy, including two (PW9 and PW10) during the end of the eighth month and two (PW11 and PW12) during the ninth month, The two congenital toxoplasmosis cases that were identified were the result of a Frimeser stimester : Bedtwimester, Sed imester Monthotprepancy 2 3 boa 5 ‘ 7 ® ° Revmotoremwney, > go a0 1112 13 4 as fis 17 18 a9 20.21 22 23 2425 26 27 28|29 HO HH BT oH pw om Oi om on Om wet PN , ° ° , ’ lows em ’ ’ . P wee " " " wn « Nw laws wo love Pw " NN PN " " " nm Ne " Noon " nw PN " " nw PN " " " " wm PIN 8 in PIN Fig2. Schematic representation ofthe timing of toxoplasmosis seroconversion inthe aubgroupof 12 pregaaal women. The representation ofthe gestational period ‘pins for the scond month pregnancy Ech pregnant woman (PW) coresponds to aie For cach ine he antenatal ete vil (ANC) ae repeesnte ie by ‘oe orto serological esas by ans" when the sample was not avaiable emergency vss (vst excluding ANC) are represented by an asterisk (”)- Each follow up ‘ut bya inclaton and ie by livery except or the PW woman wha was lot a llo-up aller her emergency vs. On exc ie, quale ani T. got IgG and gM results ae mentioned as negative (N) positive (0) oe doubtful (Dai. god IgGs mentioned inal cases ands allowed) by alt gong esl ‘when thas been achieved: [Gor Iga/lM. Core areas correspond o confirmed grey) o esate (hatched grey) seroconversion period Regarding possible CT ‘aes, mother and cord immsnolgia Ig profiles were kencl except for PW mienomedin bo where distin pofies were avr of CT. Mother and cond ‘mmc profiles were not pesormed for underined PWs hese of ensayo ether mother r cord Hood ample. mate seroconversion sociated witha tong CT spicion, "maternal seroconversion associat wth newborn low th weight maternal seroconversion essed wth prematire bith "ane of maternal seroconversion lato follow vp before delivery ips:607/ 10.137 Vural pone 0262018 gone PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 10/20PLOS ONE Toxoplasmosis and malaria during pregnancy in Benin ‘maternal infection occurring between the third and fourth months of pregnancy for one case (PW2) and atthe end ofthe pregnancy for another (PW9). Association between T. gondii serological status and P. falciparum infection during pregnancy Fig 3 presents the percentages of malaria-infected women at three time points (inclusion, fol low-up and delivery) Presence ofa Plasmodium infection at enrollment did not differ between T. gondii seronegative (Migcaion = 84/450, 18.7%) and seropositive (Dinan = 78/512, 15.2%, 16) pregnant women. Nevertheless, during follow-up and delivery, T.gondif seronegative women had more malaria infections than T. gondii seropositive ones (Nistow yp = 186/450, 41.3% 5. Moto up = 179/512, 35.0%, P= 0.04 and Macvery = 48/450, 10.7% VS. aatvery = 32/ 512, 6.3%, P=001). However, although this association did not reach significance in a multi- ‘arate analysis during the follow-up, this observation was supported by the negative ssocia- tion between the toxoplasma serological status and presence of a malaria infection at delivery (OR= 0531 and P= 0,012; Table 3) This result shows that women seropositive for T. gondii develop less malaria infections at delivery than women seronegative for T. gondi. This analysis has been conducted in a multivariate model meaning thatthe negative association between T. gondii serology and malaria infections at delivery is independent from the effect of adjustment variables on malaria infections at delivery. This was namely the case for IPTp which did not impact on the presence of a malaria infection at delivery, whatever the number of Pp doses administered during pregnancy (SI Table). a s 1 T. gondii seronegative women (n=450) IT. gondii seropositive women (n=512) Roe & Ss 8 6 Percentage of pregnant women (%) Ss Fig. Maternal malaria event in elation tothe T. gd serological tatu Maternal malaria infecton determina by TBS during the STOPPAM sty was used o compare pregnant women asording other toxoplasmosis serological stat T godt erongatve women aren ‘he grey bar (n= 50) and Tg seroposive women nthe Blak arn - 512). ipso or/10.137 Your pone 0262018 9on8 PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 31720PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin ‘Table3. Multivariate analysis exploring the asocation between lara infection and. gon maternal serological status ‘Malaria inection* during the follow-up ‘Malaria infcton® at delivery (n= 948) (n=) Independent variable Catageics [9st] | Pala’ | Odds Rati! | (05NCI | Pvaluek T. gondii postive serological satus? 0835 fostetite) 0228 0531 (os28087) oo Adjustment varisles Maternal age! 5-23) (23-29), 0717 | foass:1.053) | o.000 (0768 | (0387; 1526) (30-35) 04st (0285:0651] <0 | Liss fo.se2: 2421) Prmigest mules! women Laso (0972257) | 0099 9 | fos: 3.629) ving site ‘Aka Quedeme Ped 1736 [1195250] 0.008 07s fone ra7) ous 0430 (030805) | shows postive association “silicantP vale <0.05 in bold 4 Toxoplasmosis serological status defined at inclusion (a ~ 948) “age as boon divide into 3 periods, "maternal education was sequenced int 4 categories. ‘umber of vss has been dvd into 3 categories "number of PTp SP doses has been casi into 2 categories. ‘Thelogitic analyse was adjusted or maternal ae, gravida, living site maternal education, number of vss, dry season, bednet possesion and numberof IPTp SP doves pst on/10.137 ural pone 02620181003, Lastly, T. gondii seropositive women tended to have lower anticPFAMA\ IgG levels both at inclusion (P= 0.05) and at delivery (P= 0.002) compared to seronegative ones (Fig 4). In order to ensure the robustness of this association, a multivariate analysis adjusted for possible confounders was performed for anti-PfAMA! IgG levels at inclusion and at delivery (Table 4). Asexpected, malaria infections during the follow-up were associated with increased levels of anti-PfAMAL IgG in mothers at delivery. Also, and independently of malaria infection, anti- PJAMAL IgG levels were lower at delivery for T. gondii seropositive women compared to sero negative ones (P = 0.008; Table 4) A second analysis was performed to clarify when the levels, of anti-P/AMAI IgG differ between T. gondii seropositive and seronegative pregnant women. Again, as expected, between inctusion and delivery, the levels of anti-P{AMAI IgG increased if women were infected by P. falciparum during the follow-up. Independently of malaria infec- tions the levels of anti-PfAMAI IgG decreased more sharply in T. gondii seropositive women than in seronegative women (S2 Table) PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 12/20PLOS ONE Toxoplasmosis and malaria during pregnancy in Benin 6 0.05 0.002 — — _ & T gondii seronegative women (n=285) 8 ™™ T. gondii seropositive women (n=317)* Bo4 5S 59 S22 z < o Inclusion Delivery Fig 4 Maternal IgG levels to PPAMAA in elation to T. gon serological status. Ant DYAMAL Ig levels measured in STOPPAM stay were ‘compared ccording to materal noplamosis serological ttt Only women with complete res an-PYAMAL and ant-T gon IgG at incinion or delivery) were ker into account or the elation The bars represent the meat fan: YAMAT IgGlevey and the whiskers represent the standard deviations. 13 mining vale at delivery. gsc or/10.1371 ual pone.0262018 004 Discussion ‘To provide additional data on toxoplasmosis during pregnancy in Benin, we carried outa ret rospective sero-epicemiological study on a longitudinal cohort of pregnant women, Performed between 2008 and 2010 on 974 women living in a rural area of southern Benin, this study showed a toxoplasmosis seroprevalence rate of 52.6% with a seroconversion rate of 3.1% among seronegative women, aso interpretable as 4% among all women screened (value used subsequently in order to have the same comparison basis as that used in bibliographic reer- ences cited below), and a congenital toxoplasmosis rate of 0.24 among live birth. Although the follow-up ofthe cohort started early in pregnancy, with a mean gestational age of 17 weeks at inclusion [29], seroconversion and congenital toxoplasmosis rates were potentially underestimated for three main reasons. Firstly, the followup design began at the first ANC around the fourth month of pregnancy, soit cannot be excluded that seroconversion cases having occurred during the periconceptional period or atthe beginning of pregnancy went unnoticed, especially since we did not measure the anti-T. gondif IgM. However, the risk of not having detected these eases is low since during this period, toxoplasmosis infections are rarely involved in maternal-fetal transmission and, iftransmitted, frequently reslt in sponta neous miscarriage or malformations that would have been detected during ultrasound moni toring performed throughout the study [34]- Secondly, the depletion in some samples at delivery, as well asthe lack of some samples due to miscarriages or silliths, contributed surely oan underestimation of the number of T: gondié seronegative women having serocon verted. Lastly, the non-monitoring of serological and ophthalmological follow-up of infants luring their first year of ie may have hampered the detection of tardive C1 Despite these potential biases the 14% seroconversion rate observed in the study aligns with the 1.68% value found overall in sub-Saharan AVrica [14]. Ona smaller scale in Benin, this rate of 14% in 2008-2010 is significantly higher than the values resulting from successive stud ies carried out later on smaller numbers of women, namely in 2011 (30% seroprevalence and 0.4% seroconversion rates among 283 women screened in the northern rural area) [25] in 2012 (48% seroprevalence and 0.7% seroconversion rates among 266 women screened in the PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 33/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin ‘Table. Multivariate analysis expforing the astocation Between ant-P/AMAL IgG levels at inclusion and delivery and. gordi maternal serological sats. {n-F/AMAL IgG alincaion (9 =650)* __an-P/AMA Hg at diver (n= 627)* Independent arabe Cage Coc [95%CN]—PralaeCoeP | osner) | Prat” got postive serological sats cust [os7008) 0250 0295 fast&.asrs)_| aos Maria efecto daring allow sp cay (axmos6s) | aot ‘Adjustnent variables. Matera age Us] 3.29) ost | Fassse2s0] —ao10 amas fosioa7s)_| —ossi (30-35) ay | poarn.017] 003-0250 | (ost-o0sa| | 02st Priiget multe women 092 [446.0262) 0s 0030 (03540235) | ase Lisig ate ‘Aba vedere Pea ass) (OMGOSe] cea. asi [ass o0ns) | oon coms oy form 0217] om 156 (1412.03) _| 0 shows a positive assockton “sgnficant vale <005 in bol, “lari infction was define bya est one post “sage hos been divided nto 5 periods ve TBS during pregnancy (follow-up or delivery) ‘maternal edcaton was sequenced int 4 categories * numberof visits ashecn divided into 3 categories umber of IP Tp SP doseshas ben csi into 2 categories, “The linear analysis was adjusted for quantitative TBS, maternal ge, gravidae living site, maternal education, number of visits dry season, bednet possession and numberof IFT SP doses. hips on/10.137 urna pone 02520181004 southern urban area) [23] and in 2016 (36% seroprevalence and 0.5% seroconversion rates among, 399 women screened in the southern rural area) [26]. Some factors can explain these differences: 1) two ofthe cited studies focused on toxoplasmosis, and therefore seronegative women were made aware of the risk of infection through the adoption of preventive measures {23 26]; and 2) spatial and temporal heterogeneity in the distribution ofthe parasite may have ‘occurred, as described in the introduction for Benin, and in agreement with observations reported in the annual reports of the French National Reference Center for Toxoplasmosis [35]. In view of these factors, and knowing that different serological techniques were used, the higher seroconversion rate found in our study can also be explained by its retrospective aspect: the original women’s follow-up focused on malaria without any attention being paid to tox0- plasmosis s0 this infection followed its natural course without any preventive measures. The absence (except in one case) of specific anti-T. gondii IgM has not been explained. It may be due to their lability, i. transient or short time presence, during the specific IgG positivation PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 14/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin period [36]. However, an absence of specific IgM to toxoplasmosis is not uncommon, with no precise explanation provided [37] In terms of public health, tis] ing to consider the situation with respect to other countries that manage toxoplasmosis in pregnancy differently. During the same 2008-2010 period as considered here, the United States, where no routine toxoplasmosis gestational screening was conducted, reported a low toxoplasmosis seroprevalence rate of 9% [38] associated with high CT rates ranging from 0.01% to 0.1% depending on the states [39,40]. In comparison, France reported a higher 37% seroprevalence rate associated with a lower 0.02% CT rate, as an out: come ofits toxoplasmosis gestational screening [41]. In Benin, where the screening is limited to hospitals or large private clinics and is costly [23, 26], one would expect the high prevalence rate of 52.6% to be associated with a much higher rate of CT than in the United States where the circulation ofthe parasite is low. But the fairly close value of 0.2% is compelling and sug- gests possible reducing effect on in utero 7. gondii transmission, which could be attributable to the IPp-SP administered during pregnancy, a treatment that is inexistent in non-tropical countries such asin the United States. Investigating the impact of IPTp-SP on T. gondii infection is not new [37]. Despite ethical impossibilities precluding experiments on it, we ean hypothesize that IPTp-SP administered against malaria could reduce the impact of pregnancy-associated toxoplasmosis on fetuses. The doses would not prevent infection but would decrease the serious parasitic effects. Cur- rently, PTp-SP is administered as often as possible from the second trimester of pregnancy in ‘malaria endemic areas [27], and WHO currently recommends at least three SP doses during, pregnancy. When the STOPPAM project was implemented in 2008-2010, the current national guidelines were followed and, under the supervision of midwives, 86% of the women received two SP doses (each dose correspondling to 1,500mg sulfadoxine and 75mg pyrimethamine) spaced atleast one month apart from the second trimester of pregnancy [42~44]. In parallel fetal ultrasonography was performed three times at gestational weeks 26, 30 and 36 [32] and did not reveal any abnormality ofthe eephalic structure among the fetuses from the serocon- version maternal group, Similarly, no birth defects were observed. Despite this low number of observations, it is tempting to hypothesize thatthe IPTp-SP may have contributed to limiting, the severe adverse consequences iue to toxoplasmosis asin the case of PW2, To illustrate this hypothesis, even ifthe aim of the study was different, a multicenter randomized tral was con- ducted to compare the efficacy and tolerance of pyrimethamine (SOmg/day) associated with sulfadiazine (3x1g/day) and folinic acid versus spiramycin (3x1y/day), in order to reduce T. {gondii placental transmission [45]. Although few individuals were in the compared groups due to the interruption of the trial, this study found a trend toward lower T. gondii placental trans- mission in the SP group, in which no fetal cerebral toxoplasmosis lesions were observed. Thus, although the sulfamide-pyrimethamine regimens were very different between this latter study and ours (since the parasite targets were different), ited to a reduced severity ofthe fetal sequelae, which isin agreement with the absence of ultrasonographic signs in our PW2 case despite a maternal infection having occurred at around 4 months of gestation. Strikingly, this study observed a negative relationship between T: gondii seropositivity and both the number of malaria infections during pregnancy and the presence of placental malaria at delivery. Paradoxically this finding was accompanied by lower anti-PfAMAI IgG levels in T. gondii seropositive women, compared to others, especially at delivery. Would T. gondii pres ence provide protection to the host against malaria? Several elements may be proposed to explain this: Seropositivity to T. gondii reflects not only a past infection with this parasite but also the maintenance of a control immunity against the encysted forms ofthe parasite, mainly in the brain and muscle tissues 46]. On the other hand, the AMAL protein isa conserved element of PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 45/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin Apicomplexa parasites [47~19] (with about thirty percent homology between T. gondii and Plasmodium) for which it would play an essential role inthe parasite positioning on the host cell's membrane for favoring cell invasion [47]. Thus, toxoplasmosis infection could generate anti-AMAL IgG potentially cross-reacting with PFAMAL antigen sites. Consequently, this could either down-regulate the production of specific anti-F{AMAI IgG necessary to fight against malaria or slow down/prevent the formation of the tight attachment zone between Plasmon and the host cel, created by a parasitic protein complex [17]. Indeed, the produc tion of anti-AMAI Ig would block the interaction between RON (thoptry neck protein) and AMA\ localized in the moving junction necessary for parasite invasion [50,51]. To further investigate this hypothesis, it would be necessary to measure anti-TgAMAI IgG levels and to introduce these new data in the global analysis. Another explanation could be that the persistent installation of T. gondii in its host involves a dynamic regulation of combined celular and molecular immunoregulatory networks [52 55], which can impact the newly infecting Plasmodium parasite using similar pathways. This, could result in less production of anti-P{AMAI IgG to effectively fight plasmodial parasites. Tis hypothesis seems paradoxical due tothe clearly demonstrated association between anti- PJAMAL IgG levels and the clinical and parastological protection against malaria [56]. Never~ theless, it may be consistent in that these IgG levels, measured at delivery reflected a lower need for T gondii seropositive women to react against P. falciparum infection during their pregnancy. Finally, and without excluding the previous explanations, another suggestion i based on experimental observations where mice primo-infected with P. berghei, treated with antimalar- ial drugs and then infected with T. gondi, developed immune benefits such as faster prodc- tion of anti-T. gondii IgG in comparison to mice infected with T, gondif alone [53] Altogether, these possible explanations may support the observation in this study of better immune control towards malaria for pregnant women chronically infected by T: gondit Except for three previous studies—two of which used PCR methods on Ghanaian pregnant women only at delivery 28] or at the recruitment om the third trimester of pregnancy [37] and the third study investigating hematological parameters in young children in Cameroon (57]— to our knowledge, this study isthe frst to simultaneously investigate T. gondii and P. faleipa- ‘rum ina longitudinal human cohort. This retrospective sero-epidemiological study on toxo- plasmosisin Benin was primarily dedicated to drawing the medical staff’ attention to this pathology, which is underestimated during pregnancy in African countries where other infec- tious pressures such as that exerted by malaria is extremely concerning. For the first time in Benin, it offered the possibility of investigating a cohort of nearly 1,000 pregnant women, Although tis observation goes back 10 years, the high CT rate conctuled in the findings rein forces the importance of implementing preventive measures as much as possible against toxo plasmosis to pregnant women in the absence of knowledge oftheir serological status and when no screening program exists. Indeed, ths study aso put forward that women with alow educa- tion level were more likely to be seropositive to T. gondii. This aspect, also highlighted in a study in the United States [58] is undoubtedly the cause and the consequence of less access to and less awareness of prevention messages among these women. This study also concluded that a preexisting infection by T. gondit plays a potential protec tive role against malaria. It follows that it is essential to understand the interactions between pathogens infecting a single human host, in order to properly measure their effects, which are ncither strietly protective nor strictly deleterious but imbalanced. In this sense, further investi- gations on the cross-reactivity between IgG specific for T. gond and P. falciparum would be necessary in groups representative ofthe general population in malaria endemic areas, in order to remove any bias linked to the LP Ip delivered to pregnant women. PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 16/20PLOS ONE Toxoplasmosis and malaria during pregnancy in Benin Supporting information SI Table, Multivariate analysis of effect of toxoplasmosis serological status and IPTp on aalaria infection at delivery. ocx) 2 Table, Multivariate analysis of differences in ant-PfAMAL IgG levels between inclusion and delivery. (ocx Acknowledgments ‘We thank the teams of the microbiology and virology units from the Saint Louis Hospital as wellas the virology team from the Bichat-Claude Bernard Hospital in Pais for welcoming us. ‘We are particularly grateful to the team of the parasitology unit from the Bichat-Claude Ber- ard Hospital fr their warm welcame and assistance. This study was part ofthe STOPPAM ‘Strategy To Prevent Pregnancy-Associated Malaria” collaborative project. We are grateful to all the women who participated in the study. Finally, we thank Sharon Calandra for her proofreading Author Contributions Conceptualization: Magalie Dambrun, Frangois Simon, Sandrine Houzé, Hlorence Migot- Nabias Data curation: Magalie Dambrun, Célia Dechavanne, Valérie Briand, Formal analysis: Magalie Dambrun, Céia Dechavanne, Valérie Briand, Florence Migot Nabias Funding acquisition: Florence Migot- Nabias. stan Candau, Murielle Lohezic, Saraniya Investigation: Magalie Dambrun, Nicolas Guigus Manoharan, Nawal Sare, Firmine Viwami. Methodology: Magalie Dambrun, Nicolas Guigue, Francois Simon, Sandrine Houzé, Florence Migot-Nabias. Project administration: Magalie Dambrun, Florence Migot-Nabias. Resources: Magalie Dambrun, Nadine Fievet, Florence Migot-Nabias. Supervision: Florence Migot Nabias. Validation: Magalie Dambrun, Célia Dechavanne, Nicolas Guigue, Valérie Briand, Sandrine Houzé, Florence Migot-Nabias, Writing ~ original draft: Magalie Dambrun, Célia Dechavanne, Sandrine Houzé, Florence Migot-Nabias. References 1. Robort-Gangnoux F, Dard ML. Epidemiology of and Diagnostic Swatogios or Toxoplasmosis. Clin Microbiol lev. 2012; 25: 264-296. hips /ido.org0.1128/0MF.05013-11 PMID: 22491772 2. WHO. The burden o foodborne diseases inthe WHO European Region. WHO publications (2017). ‘Available rom: ips:/fw. our who nVenfeat topics disease preventirviood-saety/publcabons! 2017the-burden-0 foodbome-cisoases-n-the-who-european region 2017. PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 17/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin 10. 1 2 a 4. 18, 168. 1” 8 1. Parise ME, Hotez Py, ShtskerL Neglected Parasitic Infection the Unted States: Needs and Oppor- lunes. Arm J Trop Mad Hyg. 2014; 90: 783-788, hitps/dok.or/10. 42609 13-0727 PMID: ‘24808243 ‘WoochalD, Jones JL, Cantsy PT, Wikins PP, Montgomery SP. Neglected parasitic infections: what ‘every family physician needs to know. Am Fam Physician. 2014; 89: 603-811. PMID: 24266216 ‘Wallon M, Peyron F. Congenital Toxoplasmosis: Plea fora Neglected Disease. Pathogens. 2018; 7: 28, ipso. or/10.3200'pathogens7010025 PMID: 29473606 Peyton F, Volver, Mandelbrot L, Wallon M, Parroux R, Kiefer F, eta. Maternal and Congenital “Toxoplasmosis: Dsagnoss and Treatment Recommendatons ofa French ktuescpinary Working Group. Pathogons. 2019; 8:24. iips:/60.o7g/10:S300/parogonss010024 PMID: 30781652, Giraud C. Congenital toxoplasmosis in Laboratory agnosis of acquired toxoplasmosis in immunocom patent subjects including pregnant women), congenial oxoplasmosis (pre-and postnatal diagnosis) and ocular toxoplasmosis. Argumeriary. HAS (French National Author for Heal~Haute Autotede Sante) 181SBN rumber'978-2-11-151461-4; 2017, Sillaggon E, Cartier CS, Sauter M, McLeod R, Materal Serologic Screening to Prevent Congenital “Toxoplasmosis: A Decison- Analytic Economic Model PLoS Neg! Trop Ds. 2011; 5: 8133, hips). ‘o7g/10.137 Houma pr.0001383 PMID: 21880546 Psa AR, Kasper DC, Sawers L, Waiter, Hayde M, Stilwaggan E. Congenital toxoplasmosis in Aus- ‘ra: Prenatal screening for prevention is cost-saving. PLOS Neg) Trop Dis. 2017; 11: €0005648,hps/ ‘doLorg/10.137 journal pat 0005648 PMID: 2865640 Binquet C, Lejeune C, Serer V, PeyronF,Bertaux AC, Scemama O, eal. The cost-electveness of neaonatal versus prenatal screening for congenital toxoplasmosis. PLos ONE. 2019, 1460221708 htps:7doiorg/10.1871/joumeal pone 0221709 PMID: 31532766 Poyron F, MeLeod R, Aizenberg D, Contopoulos4oannidisD, Kiefer F, Mandebbrot, et al Congenital “Toxoplasmosis in France and he United States: One Parasite, Two Diverging Approaches. PLoS Negl “Trop Ds. 2017; 11: 00008222. hps/dot on 0.187 journal pnic.0005222 PMID: 28207736 Pangui LJ, Ghat OB, Kamga Waldo AR, Bakou SN. Point surlatoxoplasmose en Afrique de ouest et ‘aucoatte, RASPA 201%, 11: 29-40. Rostam A, Ria SM, Gamble HR, Fak Y, Nourolahpour Shiadoh M, Danesh M, tal. Glial prova- lence of latent cxoplasmosisin pregnant women: a systematic review and meta-analysis. Cin Microbiol Infect. 2020, 26: 673-683, hips. or/10. 1016) mi 2020.01.008 PMID: 31972316 Rostami A, RiahiSM, Contopoules‘oannids DG, Gamble HR, Fak Y, Shiadeh MN, etal Acute toxo- plasma infection in pregnant women worldwide: systomatic review and meta-analysis, PLoS Negl “Trop Ds. 2018; 18: 60007807. hps:/olorg10. 137 journal pric. 0007807 PMID: 31609965 “Torgerson PR, Mastoiacavo P. The global burden of congentaloxoplasmess:a systematic review. Bull World Heath Organ. 2013; 9 501-508. nips:/ido orp/10.247/BLT 12.1117 PMID: 29825877, Paul, Kiwolu|, Mmbaga 8, Nazarotn A, SabuniE, Maro A, ot a. Toxoplasma gondi seroprevalence ‘among pregnan women atlending artenalalelinicn Northern Tanzania Trop Med Healy, 2018; 46: 539, hips/op.o1g/10 1 186/541 182-018-0122-9 PMID: 30479556 LingussiLS, Nagalo BM, Biseye C, Kagoné TS, SanouM, Tao etal. Seroprevalence of toxoplasmo- sis and rubois in pregnant women attending antenatal private cnc at Ouagadougou, Burkina Faso, Asian Pac J Trop Med, 2012; 5; 810-819, hips/doi.or/10.1016'S 1885-7646; 12)601 48-5 PMID: 23043021 Ndiaye D, Séne PD, Nolaye M, Faye B, Nee JL, Nair. [Update on toxoplasmosis prevalence based ‘9n serological tests in pregnant women n Dakar, Senegal rom 2002 to 2006), Med Trop. 2017; 71 1or-102: PMID: 21585110, wotlo KD, Ghansah A, Oso JH, Frempong KK, Obed S, Frimpong EH, ta. Indication of Risk of Motherte-Child Toxoplasma gona Transmission in the Greater Accra legion of Ghana. Matern Chic Heath J. 2016; 20: 2581-2588, hips: org/10.1007/s10995-016-20842 PMID: 27465060 ‘NayiGO, Popodia AT, Dina, Okatie N.Pre-pregnancy counseling in Lagos: report on the fis 1000, ‘cases, Cin Exp Obstet Gynecol 2013; 40; 358-360, PMID: 24283165 Ail, Edu SA, Apea-Kubl KA, Boamah D, Bosompem KM, Edoh D. Sere-epidemiotogy of toxoplasmosis ‘amongst pregnant women inthe greater Accra region of Ghana. Ghana Med J. 2008; 43: 107=114 p70 or/10.431 aig) 44s 55:25 PMID: 20126322 Rodier MH, Berthonneau J, Bourgon A, Giraudeau G, Agus G, Burucoa C eta, Seroprevalonces of “Toxoplasma, malaria, ubela, cytomegalovius, HIV and toponemal infections among pregnant women in Cotonou, Fspublic of Benin Acta Trop. 1995; 58: 271-277. hiips/éalog/10.1010001- 706x(95)00087-u PMID: 8533662 PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 18/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin 2. 2. at. 37. a 2. ‘Ogouyém-Hounto A, Agayahoun-Chokk F, Sisinto Savi de Tove ¥, Biokou Bankole B, Adin de Souza V, Assogba M, etal. [Evaluation ofa rap dagnestic etn tne agnosis of toxoplasmosis in pregnant women in Cotendu (Benin), Bull Soe Pathol Exot 2014; 107:85-€9.hips/so\. og 10.1007/ 1813149-014-0955-8 PMID: 24630135 “Tonouhewa ABN, Amaghégnon , Aichadé SP, Hamidovié A, Mercier A, ambeun Metal. [Seropreva- lence of Toxoplasmosis among Preqnant Women in Benin: Meta-analysis and Meta-tegression|. Bull ‘Soc Pathol Exot 2019, 112: 79-88, htps:/do\or/10.S165fospe-2019-0078 PMID: 31478522 De Paschaie M, CerianiC, CerllT, Cagnn D, Cavalli S,Cialone A, etal. Antenatal screening for Texoplasia gona, Cylomogalovinis, rubella and Treponema palidum infections in noahoen Bonin. ‘Trop Med int Heath. 2014; 19: 743-746, hips /do.org/10-111 im. 12296 PMID. 24612218 Seinto Savde Tove Y, Ogouyor Hounto A, Vinadou Vedoune M, vai A, Aol D, Georgia Barklesou , eal. [Seroprevalonce and factors associated wih toxoplasmoai n pregnant women fv- ing in rural areas in Benin] Pan Atncan Med J. 2018; 29: 112 ‘Wor malaria report in Geneva: World Heath Organization. Licence: CC BY-NC-SA 3 01GO (2018). ‘Avalable rom: hips: /uwew who itfpubications/tten/ 9786241565721 Blay EA, Ghansah A, Otchere J, Koku R, Kwofle KD, Bi L, et al. Congenital toxoplasmosis and preg- nancy malaria detection post-partum: Effective diagnosis and ts impication for efcent management of ‘Congonital infection. ParasitlInt.2015; 64: 603-8. hiis./\doory'10. 016) parn.2015.08.004 PMID: 26264261 Huyan BT, FievotN, GbaguidG, Dechavanne S, Borgella S, Guéz0-Mévo 8, eal. Influence ofthe Tim ing of Malai Infection dung Pregnancy on Birth Weight anon Matomal Anemia in Benin. Am J Trop Mea Hyg, 2011; 85: 214-220. ip: org/104260/9fmn 20111-0103 PMID: 21813857 Huyah BT, Fieve N, Gbaguid G, Borgella S, Mévo BG, Massougbea A, et a. Malaria associated symptoms in pregnant women followed-up in Benin, Malar J 2011; 1072. itso. rg/10 11867 1475-2875-10-72 PMID: 2145408 ober Gangneux ., ViefouC, Toure Schaefer C, Depouy:Camet [Property of antibody avy in tne detection of toxoplasmosis seroconversion. Ann Biol Cin. 1998, 56; 586-9, PMID: 9769483 Briand V, Saal J, GhafariC, Huynh BT, FevetN, Schmiegelow C, el Fetal Grow Peston Ie ‘Associaiod Wi Malaria in Prognancy: A Prospective Longitudinal Study n Benn. J Infect Die. 2016; 244: 417-425, htps:/do\.or/10.108:ntojwi58 PMID: 27385340 'Ndam NT, Denoeuc-Nam L, Dotichamou J, Viwamn F, Salant, Nelsen MA, etal. Protective antibod ies against placental malaria and poor outcomes during pregnancy, Benin. Emerg infect Dis. 2015; 24 {813°823.htps:/do.1g/10.3201/oid2105.141626 PMID: 25868123 Dunn D, Wallon M, Peyron F, Petersen E, Peckham C, Gilbert R, Mother-to-child transmission of txo- plasmosis: sk estimates for cinical counseling. Laneat. 1999, 352: 1829-1893 hips:/doi.org0, ‘ot6/S0140.6736(08}08220-8 PMID: 10359407 French national reference center for toxoplasmosis. Available rom: p/cnrtoxoplasmase chu rims. te Ficker-Hidalg H, Cimon B, Chemla C, Darde ML, Delhaes L, olivier Oe al. Toxoplasma Serocen- version with Negative or Transient Immunoglobulin Min Pregnant Women: Myth o ey? A French Mulicenter Ratrospactve Study. J Cin Microbiol.2018; 51: 2103-2111, hips/o.g/10.11281CM. ‘00160-13 PMID: 23616461 Arthur-Mensan JneR, Blay A, Ayil LarbiJ, SuZUK'T, Ohta N. Efeceness of SP-PTpfor Malaria and Evidence for tne Need oT gonalnfeion Preventive Policy during Prognancy in Ghana. J Infect Dis Epidomiol 2016, 2: 018 ‘Jones J. Kruszon-Moran D, vera HN, Price C, Wikns PP. Toxoplasma gondi Seroprevalence inthe United States 2000-2010 and Comparison wit the Past Two Decades. Am.J Trop Med Hyg, 2014 0: 1135-1130. htps/dot org 0.426aVam. 14-0013 PMID: 24710615 Jones JL Parse ME, Fiore AE. Neglecied Parasitic Infections inthe United States: Toxoplasmosis. Am ‘Trap Mad Hyg, 2014; 90; 794-790, hips:s/o.0rg/10426a/afmnh 180722 PMID: 24808246 Hampton MM. Congenital Toxoplasmosis: A Review. Neonatal Netw, 2018; 34: 274-278 psd. ‘01g/10.189170730-0892.34 5.274 PMID: 26802827 Nogareda F, Le Sat, Viena De Valk, Goulet V. Incidence and prevalence ot Toxoplasma gona Intecton in women n France, 1980-2020; mode-based estimation. Epdemvol Infect. 2074; 142: 1661— 1670. pico org10-1017/S0350268813002756 PMID: 24225712 Huyah BT, FievetN, Briand V, BorgolaS, Massouabod A, Deloron Po a. Consequences of gesta- ‘ral malaria on bth weight: nding the best imelrame or interme prevenve realmentadminis- ‘raion. PLoS One. 2012; 7: £85342. hitps:/do:.og/10.1371/joumal pone 0085942 PMID: 22514730 PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 39/20PLOS ONE “Toxophasmosis and malaria during pregnancy in Benin a7. ‘0, st. 2. Briand V, CotellG, Massoughod) A, CotM. Intermittent preventive treatment forthe prevention of malaria durig pregnancy n Nigh tranmiesion areas Malat J 2007; 62 160. peso on 10 17050 11475-2875 6-160 PMID: 18053200 WHO policy bre forthe implementation of intermitient preventive treatment of malarain pregnancy using Sufadoxine-pyrimotnamine(IPTp-SP). World Heath Organization (2013 ev 2014), Available ‘rom: hips ww who inlmalarafpublcations/atoz/polcy brie! iptp_sp_polcy_ecommendationer MandolbrotL. Kieffer, Sita R, Laurichesse-Dolmas H, Winer N, Mesnard, et. Prenatal therapy with Dyrimethamine + sulfadiazine vs spiramycin to reduce placental tansmission of toxoplasmosis: a mult ‘conta randomized tial. Am J Obstl Gynecol. 2018; 219: 386.01~385 9. hps/do\.or/10. 1016). '2)0g.2018.05.031 PMID: 20870736 Dubey JP, Lintsay DS, Speer CA. Structures of Toxoplasma gonditachyzoles,bradyzotes, and po: rozotes and biology and development o issue cysts. Cn Microbiol ev. 1998; 11: 287-209 nips:// \doLorg/10.1128CMA. 11.2267 PMID: 9554564 BargieriDY, Andermatton N, Lagal , ThibergeS, Whitslaw JA, Tarciouxl, etal. Apical membrane ant- ‘gent mediates apicomplexan paraste attachment but dispensable for host cal ivasion Nat Com- mun. 20%; 4: 2582. htps:/ido o1g/10 1038/ncomms552 PMID: 24108241, He AB, LekutsC, Grigg ME, Bradley PJ, Dubremetz JF, Ontaga-BariaE etal. Toxoplasma gondt homologue of Plasmodium apical membrane tigen 1 i invovod in invasion of host cls. Infoct Immun, 2000; 68: 7078-7086. iips:/'di.ora/10.1 128/168, 12.7078-7086.2000 PMID: 11083833 Crawford, Tonkin ML, Gruje O, Boulanger MJ. Sctura characterzation of apical membrane antigon 1 (AMAt) tom Toxoplasma gordi Biol Chem. 2010; 285: 15644-15652, hps/do.or/10-1074%be ‘Mi09.002619 PMID: 20204017 LLamarque Mi, Besteire 8, Papoin J, Roques M, Vuliez-e Normand B, Morion-Guyet J etal The RON2-AMA\ interactions cea step in moving junction-dependen invasion by Apicomplexan para- Sites. PLos Pathog, 2011; 7:2.81001276. hips cot org/ 10.187 Vjoumal pat 1001276 PMID: 2iss7a43 CCotins CR, Withers: Martin C, HackotF, Blackman MJ. An inhibitory antibody blocks interactions betwoen components ofthe malarial invasion machinery. PLos Pathog. 2009; 5: 1.01000273. ilps. ‘do.org/10.1371/joural ppat 1000273 PMID: 19165323 Butler NS, Hares TH, lader LI. Regulation of Immunopathogenesis During Pasmodtum and Toxo: Piasma Infections: More Parallel than Distinctions? Trends Parasitol 2013; 29 593-602. ipso, 01g/10-1016} 92013 10.002 PMID: 24784186 Seties EW, Moser LA, Hats TH, KnollLJ. Toxoplasma gandi Upregulates interleukn-12 to Prevent Plamodium berghe-Induced Experimental Cerebral Malana, Inf Immun, 2014; 82: 1949-53, hips ‘dol oxg/10.1 128181 0125913 PMID: 28596082 Loe DH, Chu KB, Kang HJ, Loe SH, Quan FS. Previous Infection with Plasmodium berghoi Contors Resistance to Taxoplasma gondilifecton in Mico. Korean J Parasitol 2019; 57: 88-98. hips. ora! 10.8347. 2019.57 2.93 PMID: 31104401 |WongZS, Sokol-Borel SL Olas P, Dubey JP, Boye JP. Head-o-nead comparsons of Toxoplasma ‘ondiand ts near eiatve Hammondia hammenal teveal dramate differences in tos response and ‘tlectore wih species-spectietunctons. PloS Pathog, 2020; 18: @1008528,ips/do\o/10.1371/ Jura ppt. 1008528 PMID: 32574210 ‘Adamou R, Dechavanne C, Sassou |, 'Aime'da T, Bouraima A, Sonon P, etal. Plasmodium faepa ‘ummerozote surface antigen-specitc eyophic IgG and control of malaria infection in aBeninese birth cohort. Malar J. 2018; 18 194. htps/do\ on/10.1186'812896-019-2831-x PMID: 31185938 CCumber SN, KameniNguekam YE, Jala S. The Effects of Toxoplasmosis and Malaria Confection on Malaria Parasite Density and Hematological Paramtrs in Children (0-6 Years) inthe Nkobisson Heath Distt, Cameroon, Journal of Family Medicine and Health Caro. 2016; 2: 81-88. ‘Jones J, Kruszon Moran D, Elder S, Rivera HN, Press C, Montoya JG, et al. Toxoplasma gonciinic- ton nthe United States, 2011-2014 Am J Trop Med Hyg. 2018; 9 551-557 ip: org/10.42697 aj. 17-0677 PMID: 28260660 PLOS ONE | ips:idoiorg/10.1371fournal pone. 0262018 January 7, 2022 20/20