301309

International Epidemiological Association 2001 International Journal of Epidemiology 2001;30:1309–1313
Effect of prenatal treatment on the risk of
intracranial and ocular lesions in children
with congenital toxoplasmosis
Background Hydrocephalus, intracranial calcification and retinochoroiditis are the most common manifestations of tissue damage due to congenital toxoplasmosis, but the effectof prenatal treatment on these outcomes is unclear. We aimed to determine theeffect of prenatal treatment for toxoplasmosis on the risk of intracranial andocular lesions in congenitally infected children at 3 years of age.
A cohort of mothers identified during pregnancy with toxoplasma infection andtheir 181 liveborn children with confirmed congenital toxoplasmosis was retro-spectively analysed to determine the presence of intracranial and ocular lesions.
As few women are not treated, we compared the effects of the treatment potency(pyrimethamine-sulfadiazine versus spiramycin or no treatment), and the timingof treatment, on the risks of intracranial lesions, time to detection of ocular lesions,and detection of any lesions (intracranial or ocular) by 3 years of age. Analysestook account of the gestation at maternal seroconversion.
There was no evidence for an effect of pyrimethamine-sulfadiazine on intracranial,ocular or any lesions by 3 years: odds ratio (OR) for any lesions 0.89 (95% CI : 0.41,1.88). There was no evidence of an effect of delayed treatment on ocular lesions(hazard ratio = 0.69, 95% CI : 0.28, 1.68) or any lesions by 3 years of age (OR = 0.44,95% CI : 0.16, 1.19).
Conclusions Our study failed to detect a beneficial effect of early or more potent anti toxo- plasma treatment on the risks of intracranial or ocular lesions in children withcongenital toxoplasmosis. However, larger, prospective studies, which determinethe effect of prenatal treatment on long-term developmental outcomes are requiredto justify changes in clinical practice.
Congenital toxoplasmosis, prenatal treatment, intracranial calcification, retino-choroiditis The aim of prenatal treatment for infection with Toxoplasma age in 33% (28/85) of congenitally infected children followed in gondii is to prevent neurological or visual impairment in infected three population-based cohort studies.3–5 Of the 28 with lesions, children either by preventing mother to child transmission 3 (4%) had severe neurological impairment. Retinochoroidalof infection, or, once fetal infection has occurred, by limiting lesions have been reported in 16% (14/85) children at 1–6 years cell damage caused by the parasite.1,2,pp.173–75 In this report, old,3–5 of whom up to half may have permanent visual impair- we present results on the effect of prenatal treatment on lesions ment, due to the size of the lesion and involvement of the macular attributable to parasite-induced cell damage in children with area.3–6 Information is lacking on the risk of developmental confirmed congenital toxoplasmosis. The lesions most com- impairment in children with intracranial or ocular lesions. How- monly associated with congenital toxoplasmosis are intracranial ever, many appear to develop normally6,7 and a small study calcification or hydrocephalus, which are usually detected during comparing 17 congenitally infected children and their siblings infancy, and retinochoroiditis, which may appear at any age.2 raises the possibility that impairment may be worse in those Intracranial and/or ocular lesions were detected before school with ocular than intracranial lesions (R Eaton, New EnglandNeonatal Screening Program, personal communication).
The effect of prenatal treatment for toxoplasmosis on intra- Centre for Paediatric Epidemiology and Biostatistics, Institute of Child Health,London, UK.
cranial or ocular lesions is uncertain. In a study of 144 consecutive women referred to five fetal medicine centres, including 64 was recorded for analysis. We did not analyse the effect of treat- with infected children, Foulon et al.8 found a reduction in the ment on recurrence of lesions or on the size or site of lesions.
risk of lesions at 1 year of age in children born to women given Congenital toxoplasmosis was diagnosed if there was persist- any prenatal treatment compared with those not treated (odds ence of specific IgG antibodies beyond 12 months of age.
ratio [OR] = 0.3; 95% CI : 0.10, 0.86). However, this study and Absence of congenital toxoplasmosis was based on a decline in a case-control study of 103 infected children,9 found no evidence IgG specific antibody beyond detectable levels after discontinua- for an effect of the more potent regimen of pyrimethamine- tion of treatment. All mother-child pairs in which infection sulfadiazine compared with spiramycin alone. Both studies status was unknown were excluded from the analysis.
endeavoured to account for the effect of gestation at maternalinfection10 but failure to exclude women referred due to fetal Postnatal treatment
abnormalities may have overestimated the risk of lesions in Neonates with suspected infection were prescribed pyrimethamine (3 mg/kg/3 days) and sulfadiazine (75 mg/kg/day) for 3 weeks, We investigated the effect of prenatal treatment on the risk of followed by spiramycin (0.375 × 106 units/kg/day) for 2–5 weeks, signs in a cohort of children with congenital toxoplasmosis born followed by pyrimethamine (6 mg/kg/10 days) and sulphadoxine to infected women identified in Lyon, France. As few women (125 mg/kg/10 days) with folinic acid for at least 12 months. received no treatment, we examined whether the risk of lesions All other infants received spiramycin alone pending further was related to the potency of treatment used or to the interval evidence of congenital infection status.
between maternal seroconversion and the start of treatment.
Analysis
We investigated the effect of prenatal treatment in two ways.
First, we compared the risk of lesions in children born to Patients
women treated with spiramycin with those treated with The study is based on liveborn children with congenital toxo- pyrimethamine-sulfadiazine or not treated. We hypothesized plasmosis of women who were prospectively identified with that pyrimethamine-sulfadiazine would be more effective in toxoplasma infection by the toxoplasma reference laboratory in reducing the risk of lesions because it reaches higher levels in Lyon, France, between 1987 and 1995. The study methods have fetal blood and, in contrast to spiramycin, is able to penetrate been described elsewhere.10,11 Diagnosis of maternal infection the blood brain barrier.12 We conducted separate analyses of the was based on: (a) detection of seroconversion (change from neg- effect of treatment on the risks of intracranial and ocular lesions ative to positive specific IgG antibodies); or (b) detection of IgM and on any lesions (intracranial or ocular) detected by 3 years specific antibodies and, low IgG avidity (Ͻ35%), and/or rising of age (follow-up was complete for 95% of the person years of specific IgG titre in women who were IgG positive at the first follow-up to 3 years of age). Second, we compared the risk of prenatal test. We excluded women referred for suspected fetal intracranial, ocular or any lesions in children born to women infection or abnormality based on scrutiny of referral letters or prescribed any treatment within 4 weeks of estimated serocon- non-sequential dates for detection of maternal infection, fetal version with those treated 4 or more weeks after seroconversion infection or fetal abnormalities. Spontaneous fetal losses or terminations were excluded from the analysis as fetal inves- We used a statistical model, previously described10,11 to take tigations for lesions (ultrasound or autopsy) were not routinely account of the effect of gestation at seroconversion and the interval between seroconversion and treatment (defined as treat-ment delay) on the risk of lesions. For women who were IgG Prenatal treatment
positive at the first prenatal test, we assumed that seroconversion After confirmation of infection, women were prescribed occurred between conception and the first positive test.
spiramycin (9 × 106 units/day) until delivery. If fetal diagnosis In the analyses of the risk of intracranial lesions and of any was positive, or maternal infection was acquired after 32 weeks, lesion by age 3, we assumed that the effects of treatment and treatment consisted of pyrimethamine (50 mg/day) and sulfa- gestational age at seroconversion were additive on a logistic scale.
diazine (3 g/day) alternating 3-weekly with spiramycin until We used a Weibull model to analyse the effect of treatment on the delivery (subsequently referred to as pyrimethamine-sulfadiazine).
time to first detection of the first ocular lesion and assumed that Women with confirmed fetal infection and evidence of intra- the hazard was proportional according to treatment categories cranial calcification or hydrocephalus on fetal ultrasound were and gestation at seroconversion. Kaplan-Meier estimates were used to determine the proportion of children without ocularsigns according to year of age. The association between the Follow-up
presence of ocular and intracranial lesions was assessed using Paediatric examinations were scheduled during the neonatal period, at 2, 5, 8, and 12 months of age, and annually thereafter,or until congenital infection had been excluded. At each visit, the child was examined for retinochoroidal lesions by directophthalmoscopy, usually after dilation of the iris, and a blood Mother to child transmission of infection occurred in 194 of the sample was taken for serology testing. Infants underwent cranial 704 infected pregnancies with complete data (Figure 1, preced- ultrasonography or radiography or both. The dates at which ing paper11), of which 181 resulted in a live birth. Of the intracranial calcification, hydrocephalus or the first ocular lesion 13 non-live births, 8 fetuses were terminated after 22 weeks, was first detected or the date of the last negative examination 5 due to hydrocephalus or ventricular dilatation on ultrasound PRENATAL TREATMENT OF CONGENITAL TOXOPLASMOSIS not confirmed until the next radiological examination at 3 yearsold. Two children had normal skull X-rays and ultrasound ex-aminations after birth but intracranial calcification was detectedon CT scan when they were re-examined, due to epilepsy, at 2 and 3 years of age.
Retinochoroidal lesions were detected in 37 children after a median follow-up of 6 years 5 months. Figure 1 shows that7/181 (4%) children had lesions detected during the first monthof life, 16 (9%) by 6 months, 19 (11%) by 12 months, 29 (16%)by 3 years, 32 (19%) by 5 years and 36 (23%) by 7 years. Over-all, 21/181 children were not completely followed up to 3 years.
Follow-up was complete for 95% (516/543) of the total personyears up to age 3.
Hydrocephalus, intracranial calcification, and/or ocular lesions were detected by 3 years of age in 38/181 (21%) live-born children. Ocular lesions were more common in children Figure 1 Time till detection of first retinochoroidal lesion in
181 children with congenital toxoplasmosis
with intracranial calcification (9/17, 53%) than in childrenwithout intracranial calcification (21/164, 13%; P = 0.0015).
Vertical lines represent 95% confidence intervals.
Effect of type of treatment
There was no evidence that the risk of intracranial lesions (hydro- scan: one also had retinochoroidal lesions detected at autopsy. Five cephalus or intracranial calcification) was reduced in children fetuses spontaneously aborted or were stillborn, one immediately born to mothers prescribed pyrimethamine-sulfadiazine after cordocentesis at 30 weeks gestation. All analyses presented compared with those prescribed spiramycin alone (adjusted are based on the 181 congenitally infected liveborn children.
OR = 0.90, 95% CI : 0.29, 2.64) and similarly for untreated Prenatal treatment consisted of pyrimethamine-sulfadiazine mothers compared with those prescribed spiramycin (adjusted in 70/181 women (39%), spiramycin alone in 89/181 women OR = 1.04, 95% CI : 0.05, 8.13) (Table 1). Second, there was no (49%), and 22/181 (12%) were not treated. The estimated median evidence that women prescribed pyrimethamine-sulfadiazine interval between seroconversion and the start of treatment with were less likely to have children who developed ocular lesions pyrimethamine-sulfadiazine was 7 weeks (interquartile range at any age compared with women prescribed spiramycin alone [IQR] 4–10 weeks), and for spiramycin, 4 weeks (IQR 2–7 weeks).
(adjusted hazard ratio = 1.13, 95% CI : 0.56, 2.47). We estimated All 181 infected children were treated postnatally. In 177 that if the mother seroconverted at 24 weeks gestation and gave children, treatment was started within 2 months of birth. All birth to an infected child, the risk of ocular lesions by age 3 but four children were prescribed pyrimethamine-sulfadiazine years was 15.9% given spiramycin alone, and 17.7% given and/or pyrimethamine-sulphadoxine. The median age when pyrimethamine-sulfadiazine treatment prenatally: estimated treatment was stopped was 22 months (IQR 17 months– difference—1.8%, (95% CI : –12.7, 8.7%). Finally, we found no evidence that the type of prenatal treatment had an effect onthe risk of any lesions by age 3 years (adjusted OR associated Intracranial and ocular lesions
with pyrimethamine-sulfadiazine compared with spiramycin = Hydrocephalus was detected after birth in two children: one also had intracranial calcification; the other died at 8 days of agewith hepatitis and disseminated intravascular coagulation after Effect of treatment delay on clinical signs
delivery at 34 weeks gestation, following maternal serocon- The distribution of mother-child pairs, according to type of lesions, version between 15 and 22 weeks and prenatal treatment with estimated treatment delay and gestation at seroconversion is spiramycin. Intracranial calcification was reported in a total of shown in Figure 2. An estimated 40% (72/181) of women were 17/181 (9%) children. None of the lesions in liveborn children treated within 4 weeks after seroconversion, 34% (58/181) were detected prenatally. In 14 children, the first skull X-ray or were treated between 4 and 8 weeks, and 16% (29/181) after cranial ultrasound after birth revealed intracranial calcifications.
8 weeks or more. Of the 22 women who were not treated, In one child, intracranial calcification was suspected at birth but postnatal treatment was started within the first three weeks of Table 1 Effect of prenatal treatment on intracranial and ocular lesions in children with congenital toxoplasmosis
Intracranial calcification or hydrocephalus
Retinochorioditis
Any lesion by age 3 years
Adjusted odds ratio
Adjusted hazard ratio
Adjusted odds ratio
Treatment regimen
Note: all analyses are adjusted for gestation at seroconversion.
evidence that spiramycin compared with no treatment reducedthe risk of intracranial or ocular lesions but due to the smallsample size of the untreated group, the power to detect a sig-nificant difference between spiramycin and untreated women islow. We also found no evidence that delayed treatment (after 4or more weeks after seroconversion) or no treatment, reducedthe risk of clinical signs by 3 years of age, compared with earlytreatment (within 4 weeks of seroconversion).
Unexpectedly, the risk of intracranial lesions was significantly reduced in children born to women treated after a delay or nottreated at all. Children born to women treated after 4 or moreweeks or not treated had a significantly lower risk of intracraniallesions, than for women treated within 4 weeks of seroconversion(adjusted OR = 0.08, 95% CI : 0.01, 0.45). This result may haveoccurred by chance and is not confirmed by preliminary find-ings from a larger, prospective European Multicentre Study onCongenital Toxoplasmosis (EMSCOT, unpublished data, R Gilbert).
Figure 2 Estimated gestational age at seroconversion and treatment
We also considered alternative explanations. First, abnormal fetal ultrasound findings may have led to prompt treatment. After Dark grey dots = both ocular and intracranial lesions, black dots = re-checking all patient files, we found no cases of suspected ocular lesions by age 3, light grey dots = intracranial lesions and fetal ultrasound abnormalities prior to diagnosis of maternal transparent dots = no lesions detected. Treatment delay is plotted infection. Second, we investigated whether maternal symptoms against gestation at seroconversion. We assumed that for untreatedwomen, treatment of the child started at delivery. Both measures are of infection could be associated with early treatment and an based on the estimated date for seroconversion, which was a weighted increased risk of clinical signs but found no evidence of an asso- mean of all possible dates for each woman (weighted by the ciation: 5% (11/181) of women reported symptoms, of whom one gave birth to a child with signs. A third possible explanationfor the reduced risk of intracranial lesions with delayed or no life in 19 children, after 11⁄2 months in two and after 3 months treatment is that early treatment with pyrimethamine-sulfadiazine suppresses the bone marrow2,p.227 and compromises the maternal Mothers who gave birth to children with intracranial lesions and/or fetal immune responses that limit parasite-induced cell were significantly more likely to have seroconverted earlier in damage.13–15 Our results need to be investigated in more power- pregnancy (OR per week of gestation = 0.90, 95% CI : 0.84, ful studies that can analyse the effect of early versus delayed 0.97), whereas we did not detect a significant effect of gestation treatment according to treatment type.
at seroconversion on the risk of ocular lesions (OR = 0.97, 95% The lack of evidence for an effect of the more potent regimen of pyrimethamine-sulfadiazine compared with spiramycin alone There was no evidence that delayed treatment had a signifi- or no treatment, may be due to treatment after encystment of cant effect on the time to detection of ocular lesions (adjusted the parasite has occurred. Once T. gondii tachyzoites have hazard ratio for treatment after 4 weeks/no treatment compared crossed the placenta, transformation to the bradyzoite form can with treatment within 4 weeks of seroconversion = 0.69, 95% occur within days,16 probably due to stresses caused by the CI : 0.28, 1.68) or on the risk of any lesions by 3 years of age humoral and cell mediated immune responses.17 Neither (adjusted OR = 0.44, 95% CI : 0.16, 1.19).
spiramycin nor pyrimethamine-sulfadiazine is effective against As treatment may have an effect on mother to child transmis- the encysted, bradyzoite form of the parasite, although sion, we determined the overall effect of early (within 4 weeks experimental studies show that both are effective against the of seroconversion) compared with delayed or no treatment on free tachyzoite form.12 As treatment is always given some time, the risk of lesions by 3 years of age in infected and uninfected often weeks, after maternal antibodies develop, it may be children born to seroconverting women (see preceding paper11).
given after encystment of the parasite. Due to limitations The adjusted OR for any lesion at 3 years associated with delayed of sample size and the fact that few women receive or no treatment compared with early treatment was 0.63 (95% pyrimethamine/sulfadiazine immediately after detection of maternal seroconversion, we could not investigate the effects oftiming of different types or dosages of treatment.
We found a significant effect of gestation at maternal sero- conversion on the risk of intracranial lesions, but a less marked We have analysed the largest reported cohort of children with effect on the risk of ocular lesions. These findings may reflect congenital toxoplasmosis, taking care to minimize selection bias differing immune responses at the two sites resulting in differ- due to women with affected fetuses, and taking account of the effect of gestation at maternal seroconversion on the risk oflesions.10 We found no evidence that prenatal treatment with pyrimethamine-sulfadiazine compared with spiramycin reducedthe risk of intracranial or ocular lesions although the 95% CI Our study failed to detect a beneficial effect of prenatal treat- included both beneficial and harmful effects. There was also no ment on the risk of clinical signs in infected children. Although PRENATAL TREATMENT OF CONGENITAL TOXOPLASMOSIS ours is the largest study to date to address this question, the and Environmental Protection (Netherlands), ISBN 90–9004179–6.
confidence intervals for the effects of different types of treat- ment are wide. Consequently, we cannot exclude potentially 6 Koppe JG, Loewer Sieger DH, de Roever Bonnet H. Results of 20-year beneficial or harmful effects. The significant finding of a beneficial follow-up of congenital toxoplasmosis. Lancet 1986;i:254–56.
effect of delayed treatment on intracranial lesions may be 7 Berrebi A, Kobuch WE, Bessieres MH et al. Termination of pregnancy due to chance and requires confirmation by further studies. A for maternal toxoplasmosis. Lancet 1994;344:36–39.
further concern is the lack of information on the association 8 Foulon W, Villena I, Stray-Pedersen, B et al. Treatment of toxo- between intracranial lesions and subsequent developmental plasmosis during pregnancy: a multicentre study of impact on fetal impairment. Larger, prospective studies, which determine the transmission and children’s sequelae at age 1 year. Am J Obstet Gynecol
1999;180:410–15.
effect of prenatal treatment on long-term developmental outcomes are therefore required to justify changes in clinical Couvreur J, Thulliez P, Daffos F et al. In utero treatment of toxoplasmicfetopathy with the combination pyrimethamine-sulfadiazine. Fetal Diagn Ther 1993;8:45–50.
10 Dunn D, Wallon M, Peyron F, Petersen E, Peckham CS, Gilbert RE.
Mother to child transmission of toxoplasmosis: risk estimates for
clinical counselling. Lancet 1999;353:1829–33.
The study was funded by The Wellcome Trust and by the Euro- 11 Gilbert RE, Gras L, Wallon M, Peyron F, Ades AE, Dunn, D. Effect of pean Commission BIOMED programme (BMH4-CT98–3927).
prenatal treatment on mother to child transmission of Toxoplasma We thank the staff of Laboratoire de Parasitologie et de Pathologie gondii: a cohort study of 554 mother-child pairs in Lyon, France. Int J Exotique, Hôpital de la Croix Rousse, Lyon, for help with data Epidemiol 2001;30:1303–08.
collection and Eskild Petersen for helpful comments on drafts of 12 Derouin F. Drugs effective against Toxoplasma gondii. Present status the report and facilitating the collaboration. We thank Martine and future perspective. In: Amboise-Thomas P, Petersen E (eds).
Wallon and Francois Peyron for providing the data and for Congenital Toxoplasmosis. Scientific Background, Clinical Management and contributing to the design, execution and interpretation of the Control. Paris: Springer Verlag, 2000, pp.95–110.
study. The conclusions in this paper do not reflect the opinions McLeod R, Dowel M. Basic immunology: fetus and the newborn. In: of the clinicians in the department which provided the data.
Amboise-Thomas P, Petersen E (eds). Congenital Toxoplasmosis. ScientificBackground, Clinical Management and Control. Paris: Springer Verlag,2000, pp.37–68.
Alexander J, Roberts CW, Walker W, Reichmann G, Hunter CA. Theimmunology of Toxoplasma gondii infection in the immune-competent 1 Jeannel D, Costagliola D, Niel G, Hubert B, Danis M. What is known host. In: Amboise-Thomas P, Petersen E (eds). Congenital Toxoplasmosis. about the prevention of congenital toxoplasmosis? Lancet 1990;336:
Scientific Background, Clinical Management and Control. Paris: Springer 2 Remington JS, McLeod R, Desmonts G. Toxoplasmosis. In: Remington 15 Dammann O, Leviton A. Infection remote from the brain, neonatal JS, Klein J (eds). Infectious Diseases of the Fetus and Newborn. 4th Edn.
white matter damage, and cerebral palsy in the preterm infant. Semin Pennsylvania: WB Saunders, 1995, pp.140–267.
Pediatr Neurol 1998;5:190–201.
3 Guerina NG, Hsu HW, Meissner HC et al. Neonatal serologic screening 16 Luder CGK, Giraldo Velasquez MA, Sendtner M, Gross U. Toxoplasma and early treatment for congenital Toxoplasma gondii infection. The gondii in primary rat CNS cells: differential contribution of neurons, New England Regional Toxoplasma Working Group. N Engl J Med astrocytes, and microglial cells for the intracerebral development and 1994;330:1858–63.
stage differentiation. Exp Parasitol 1999;93:23–32.
4 Lebech M, Andersen O, Christensen NC et al. Feasibility of neonatal 17 Roberts F, McLeod R. Pathogenesis of toxoplasmic retinochoroiditis.
screening for toxoplasma infection in the absence of prenatal Parasitol Today 1999;15:51–57.
treatment. Lancet 1999;353:1834–37.
18 Brezin AP, Kasner L, Thulliez P et al. Ocular toxoplasmosis in the fetus.
5 Conyn-van-Spaedonck MAE. Prevention of Congenital Toxoplasmosis immunohistochemistry analysis and DNA amplification. Retina 1994; in the Netherlands (Thesis). National Institute of Public Health 14:19–26.
International Epidemiological Association 2001 International Journal of Epidemiology 2001;30:1314–1315
Commentary: Little evidence of effective
prenatal treatment against congenital
toxoplasmosis—the implications for
testing in pregnancy
Testing for antibodies against Toxoplasma gondii in pregnancy is includes children of all seroconverting women in the analysis routinely offered in some countries. However, no randomized (the outcome is a combination of risk of transmission and risk controlled trials of the effect of treatment have been performed, of sequellae given transmission). If the latter study3 had limited and the question of whether testing in pregnancy should be the analysis to children under direct risk, it appears from inspec- encouraged rests on evidence from observational studies. Two tion of the reported data that the protective effect would have such studies are presented in this issue. The first investigated been smaller and insignificant, although this cannot be directly the effect of timing and type of treatment on the risk of vertical judged since the distribution of gestational age at infection (the transmission.1 The authors hypothesize that the lack of effect main confounder) is not presented by treatment group. A major in their study is explained by rapid transmission to the fetus problem with all the observational studies is the selection of after maternal infection, implying that the initiation of treatment patients to treatment, and given treatment, the selection to short or long time intervals before treatment starts.
The other study estimates the impact of timing and type of treatment on the risk of intracranial and ocular lesions in con- genitally infected children at 3 years of age.2 Even after controllingfor the length of gestation at maternal infection, they found a The observational studies discussed above do not provide con- paradoxical effect of treatment. A long interval from serocon- vincing evidence for beneficial effects of prenatal treatment of version to treatment lowered the risk of intracranial lesions, toxoplasmosis. Treatment in pregnancy should, in our opinion, whereas the opposite would be expected. The interval seemed be regarded as experimental, and only be performed as part of to have no impact on the risk of ocular lesions and the type of carefully conducted randomized trials.
treatment did not seem to influence the risk of congenital Even if there were a beneficial effect of treatment in preg- damage. The authors state that their results provide no evidence nancy, there may not be an overall positive effect of screening.
that prenatal screening is beneficial.
The effect of screening depends on the magnitude of the healthproblem, the estimated treatment effect, but also on the com-pliance to the screening programme and the treatment.5 Also, possible side effects of treatment must also be considered, as dis- The finding of no treatment effect on the transmission rate is in cussed by Gras et al. on the basis of their findings.2 Pyrimethamine agreement with a recent European multicentre study including is a folic acid antagonist. The use of folic acid antagonists in 144 infected women,3 and with systematic literature reviews.4,5 pregnancy has been associated with increased risk of neural However, estimates of transmission rates vary widely. In the tube defects.7 Termination of pregnancy with a healthy fetus present study,1 the proportion of infected children was 28%, and complications to invasive prenatal diagnosis are other im- while it was 19% in a large population-based Danish study.6 portant potential side effects. In addition to the economic costs In the multicentre study,3 the transmission rate varied from and the use of limited health resources, which traditionally 30% to 73% between centres, with an overall value of 44%.
have been in focus when debating screening initiation, the This variability raises issues of inter-study differences in psychological aspects have come increasingly into focus.8 Initial selection of cases or differences in laboratory methods.
false positive diagnoses are common when screening for disorders The finding in the second study2 that treatment has no of low prevalence. Both false and verified positive diagnoses beneficial effect on sequellae is in contrast to the results from may cause anxiety in the mother and her family throughout the the multicentre study,3 where a significant, protective effect of pregnancy and reduce the positive expectations for the new treatment is reported. The studies differ in their analytical child, even though the risk of a severely diseased child is low.
approach. The study by Gras et al.2 includes only outcomes in On this background, should one commence health district- children where transmission has occurred (i.e. the children who randomized controlled trials on the effect of introducing a screen- are at direct risk of sequellae), while the multicentre study3 ing programme of testing in pregnancy? In today’s situation,with little evidence of any beneficial effect of treatment, ouropinion is no. First, one should document that treatment is Section of Epidemiology, National Institute of Public Health, PO Box 4404Nydalen, 0403 Oslo, Norway.
PRENATAL TREATMENT OF CONGENITAL TOXOPLASMOSIS 2 Gras L, Gilbert RE, Ades AE, Dunn DT. Effect of prenatal treatment on the risk of intracranial and ocular lesions in children with congenital We suggest that the effect of prenatal treatment should be tested toxoplasmosis. Int J Epidemiol 2001;30:1309–13.
out using an ordinary double-blind placebo-controlled clinical 3 Foulon W, Villena I, Stray-Pedersen B et al. Treatment of toxoplasmosis trial with randomization on an individual basis.
during pregnancy: a multicenter study of impact on fetal transmission Additionally, better estimates of the burden of disease should and children’s sequelae at one year of age. Am J Obstet Gynecol 1999; be made, including population statistics. How many children 180:410–15.
in a certain population will suffer from the consequences of 4 Wallon M, Liou C, Garner P, Peyron F. Congenital toxoplasmosis: congenital toxoplasmosis? The observational studies give little systematic review of evidence of efficacy of treatment in pregnancy.
detail as to the degree of disability and illness experienced by Br Med J 1999;318:1511–14.
these children. Such data are also needed for evaluation of public 5 Eskild A, Oxman A, Magnus P, Bjørndal A, Bakketeig LS. Screening for toxoplasmosis in pregnancy: what is the evidence of reducing a We also suggest that controlled community trials could be health problem? J Med Screening 1996;3:188–94.
performed to estimate the effect of primary prevention directed Lebech M, Andersen O, Christensen NC et al. Feasibility of neonatal against established risk factors.9 Thus, one could randomize health screening for toxoplasma infection in the absence of prenatal treat-
ment. Lancet 1999;353:1834–37.
care districts to have intervention or no intervention, where the intervention might consist of detailed advice to women in early Hernandez-Diaz S, Werler MM, Walker AM, Mitchell AA. Neural tubedefects in relation to use of folic acid antagonists during pregnancy.
pregnancy to modify behaviour with respect to consumption Am J Epidemiol 2001;153:961–68.
of raw or undercooked meat and unwashed vegetables, and 8 Stewart-Brown S, Farmer A. Screening could seriously damage your behaviour with respect to contact with cat faeces.
health. Br Med J 1997;314:533–34.
9 Kapperud G, Jenum PA, Stray-Pedersen B, Melby K, Eskild A, Eng J.
Risk factors for toxoplasma infection in pregnancy. Results from a
prospective study in Norway. Am J Epidemiol 1996;144:405–12.
1 Gilbert RE, Gras L, Wallon M, Peyron F, Ades AE, Dunn DT. Effect of prenatal treatment on mother to child transmission of Toxoplasma
gondii: retrospective cohort study of 544 mother-child pairs in Lyon,
France. Int J Epidemiol 2001;30:1303–08.
International Epidemiological Association 2001 International Journal of Epidemiology 2001;30:1315–1316
Commentary: Efficacy of prenatal treatment
for toxoplasmosis: a possibility that cannot
be ruled out
In their retrospective cohort study of 554 mother-child pairs, A further problem is that most of the untreated women were Gilbert et al. did not detect a significant effect of prenatal infected during the third trimester of pregnancy. Figure 4 shows treatment on the risk of vertical transmission of toxoplasmosis.1 that only three women infected before 28 weeks of gestation This result is not surprising as there were very few untreated were not treated. The remaining 28 untreated women were women and the analysis of no treatment versus pyrimethamine- infected after 28 weeks. The effect of treatment in the third sulphadiazine was restricted to half of the cohort who did not trimester cannot be generalized to the whole of pregnancy.
undergo amniocentesis. The confidence interval (0.37–3.03) Finally, the authors explain their findings by suggesting thatfor the odds ratio (1.06) for no treatment compared with vertical transmission occurs soon after infection, during para- pyrimethamine-sulphadiazine was therefore very wide and sitaemia. This hypothesis is not supported by any scientific could include a doubling in the risk of transmission in untreated studies in humans. On the contrary, one study found that the women. Thus an absence of evidence of prenatal treatment sensitivity of prenatal diagnosis was lower in early than mid effect does not exclude a clinically important beneficial effect.
pregnancy, suggesting that vertical transmission may be delayedfor some women infected in early pregnancy.2 In the second report by Gras et al.,3 the authors unexpectedly found no evidence that prenatal treatment with pyrimethamine- Laboratoire de la Toxoplasmose, Institut de Puériculture, Paris, France.
sulphadiazine was more effective than spiramycin in reducing the risks of intracranial or ocular lesions in congenitally infected spiramycin. The periods of spiramycin treatment may have led infants by 3 years of age. A potential explanation for this result to parasitic relapses in fetal tissues, as shown in experimental is that mothers who transmitted the infection to their fetus models.6 The current treatment policy for women with a positive soon after infection were more likely to be treated with prenatal diagnosis is to prescribe continuous treatment with pyrimethamine-sulphadiazine than mothers infected at the same pyrimethamine-sulphadiazine until delivery. The data reported gestation but in whom transmission was delayed until later in by Gilbert et al.1 and Gras et al.3 provide no convincing evidence pregnancy. These two groups may not be comparable as fetuses infected earlier in pregnancy have a higher risk of clinical signs.
This explanation is suggested by the fact that mothers infectedbefore 32 weeks were only given pyrimethamine-sulphadiazine if the diagnosis of fetal infection was positive (i.e vertical trans- 1 Gilbert RE, Gras L, Wallon M, Peyron F, Ades AE, Dunn DT. Effect of mission occurred between maternal infection and the date of prenatal treatment on mother to child transmission of Toxoplasma fetal sampling). Other mothers infected before 32 weeks were gondii: retrospective study of 554 mother-child pairs in Lyon, France.
treated with spiramycin until delivery, either because the pre- Int J Epidemiol 2001;30:1303–08.
natal diagnosis was negative or not attempted. In this latter 2 Romand S, Wallon M, Franck J, Thulliez P, Peyron F, Dumon H.
group, transmission occurred either after amniocentesis or at Prenatal diagnosis using polymerase chain reaction on amniotic fluid some unknown time between the date of maternal infection for congenital toxoplasmosis. Obstet Gynecol 2001;97:296–300.
and delivery, that is later during gestation than in the group 3 Gras L, Gilbert RE, Ades AE, Dunn DT. Effect of prenatal treatment on receiving pyrimethamine-sulphadiazine.
the risk of intracranial and ocular lesions in children with congenital There are two further explanations for the lack of effect of toxoplasmosis. Int J Epidemiol 2001; 30:1309–13.
pyrimethamine-sulphadiazine. Firstly, there was a long delay Dunn D, Wallon M, Peyron F, Petersen E, Peckham C, Gilbert R.
before pyrimethamine-sulphadiazine was started. This was Mother-to-child transmission of toxoplasmosis: risk estimates for clinical
counselling. Lancet 1999;353:1829–33.
because the study was carried out more than 6 years ago, when mouse inoculation was the standard fetal diagnostic test4 Hohlfeld P, Daffos F, Costa JM, Thulliez P, Forestier F, Vidaud M.
Prenatal diagnosis of congenital toxoplasmosis with a polymerase- and pyrimethamine-sulphadiazine treatment would have been chain-reaction test on amniotic fluid. N Engl J Med 1994;331:
delayed for 3–6 weeks until results were known. Today, PCR analysis of amniotic fluid is widespread. Results are available 6 Piketty C, Derouin F, Rouveix B, Pocidalo JJ. In vivo assessment in one day and women with infected fetuses are treated much of antimicrobial agents against Toxoplasma gondii by quantification of earlier.5 Secondly, women in the study given pyrimethamine- parasites in the blood, lungs, and brain of infected mice. Antimicrob sulphadiazine actually received an alternating regimen with Agents Chemother 1990;34:1467–72.

Source: http://www.migato.com/conocele/docs/Gras2001.pdf