Presented in part: VIII International Congress on Toxoplasmosis, Porticcio, Corsica, France, 27–31 May 2005 (abstract 26).
We present a case of disseminated congenital toxoplasmosis in a newborn born to a mother who had been immunized against toxoplasmosis before conception. The mother was reinfected, likely by ingestion of imported raw horse meat during pregnancy. This clinical presentation is exceptional in France and raised the possibility of infection by a highly virulent Toxoplasma strain. The strain responsible was isolated from the peripheral blood of the newborn, and when genotyped with microsatellite markers, it exhibited an atypical genotype, one which is very uncommon in Europe but had been described in South America. We tested the hypothesis of a reinfection with a different genotype by using an experimental mouse model, which confirmed that acquired immunity against European Toxoplasma strains may not protect against reinfection by atypical strains acquired during travel outside Europe or by eating imported meat.
After primary maternal infection by Toxoplasma gondii during gestation, the parasite may enter the fetal circulation by infection of the placenta. Placental transmission is less frequent when infection is acquired before the tenth week of pregnancy and is very rare when infection is acquired before conception. Without treatment, the incidence of fetal infection is 10%–15% for acquisition during the first trimester, 30% for the second trimester, and 60% for the third trimester [1]. Early maternal infection (during the first and second trimester) may result in severe congenital toxoplasmosis, including fetal death and spontaneous abortion.By contrast, late maternal infection (during the third trimester) usually results in subclinical toxoplasmosis in newborns. In these cases, infection initially goes unnoticed, but these babies can develop chorioretinitis during later life [2]. Acute infection is followed by the formation of cysts in chronic infection and is associated with an immune response that usually confers protection against reinfection. This chronic infection is characterized by stable titers of specific IgG. In immunocompetent mothers who have been immunized against toxoplasmosis before conception, immune mechanisms prevent transmission of the infection to their fetuses. We report a case of life-threatening, disseminated congenital toxoplasmosis in an infant born to an immunocompetent mother who had been immunized against toxoplasmosis before conception. We also isolated and characterized the strain that caused the infection.
A 31-year-old woman, native to France, gave birth at term to a 3050 g baby by cesarean delivery (this method of delievery was decided on as a result of dystocia). It was her second child, and the results of serological testing during her previous pregnancy were consistent with a past Toxoplasma infection. Several hours after birth, the child presented hypotonia, liver and splenic enlargement, petechial purpura, and thrombocytopenia (platelet count, 17,000 platelets/µL). Sepsis investigations ruled out a bacterial infection (blood culture results and C-reactive protein remained negative) and viral infection due to cytomegalovirus, herpes simplex viruses 1 and 2, and echovirus. The following day, the baby became polypneic and hypoxic, requiring nasal oxygen therapy, despite receipt of wide-spectrum antibiotic therapy (ampicillin, gentamycin, and cefotaxime). A pulmonary radiograph showed an alveolar syndrome. The results of a pulmonary computed tomography (CT) scan and cardiac ultrasound were normal.
The diagnosis of congenital toxoplasmosis was suggested by the ophthalmologist, who found multiple foci of chorioretinitis on both eyes. The inflammation was moderate but involved the macula in the right eye. The diagnosis was confirmed by analysis of the newborns blood 8 days after birth: a polymerase chain reaction (PCR) assay with a direct amplification of the Toxoplasma B1 gene was positive, and T. gondii was isolated by inoculation into mice. PCR of cerebrospinal fluid samples was negative. Specific IgM antibody synthesis in serum from the newborn was demonstrated with an immunosorbent agglutination assay (ISAGA) (table 1). The results of an electroencephalogram, cranial radiograph, and ultrasound were normal, but a cerebral CT scan showed diffuse microcalcifications. Treatment with a combination of pyrimethamine, sulfadiazine, and folinic acid began 6 days after birth and was continued throughout the first year of life. After 1 week of treatment, hypotonia and thrombopenia disappeared. Pulmonary signs lasted for 12 days, and chorioretinitis stabilized without the use of steroids. The last cranial ultrasound showed no ventricular enlargement and only 1 calcification.
Fourteen months of follow-up serological testing showed a decrease in specific IgG titer followed by a rebound after discontinuation of therapy at 1 year of age (table 1). After 5 years of clinical follow-up, the childs neurological examination results were normal, and the child was performing satisfactorily in school. Ophthalmologic examination shows a large scar in the right macula. The visual acuity is 2/10 in the right eye and 10/10 in the left eye; she wears corrective glasses and needs an orthoptic aid.
Maternal clinical and laboratory data. The results of tests for Toxoplasma antibodies performed in May 2001, before the beginning of the pregnancy (October 2001), were consistent with past infection (i.e., the results showed the presence of specific IgG antibodies in the dye test without detectable IgM). Serological results remained unchanged until 4 June 2002, at 32 weeks of pregnancy, when an increase in IgG titer was observed and confirmed 3 weeks later by further testing(table 1).
The mother recalled severe asthenia, general dizziness, headache, and abdominal pain occurring on June 11. She had eaten raw horse meat several times in May and June. Tests for underlying immunosuppression in the mother had negative results, including complete phenotypes of circulating T lymphocytes, functional testing of T lymphocytes, and serological testing for HIV. The mother had had autoimmune thyroiditis (Hashimoto disease) for years, and the level of antithyroglobulin antibodies was high at the time of birth, signaling active thyroiditis.
Strain characterization. T. gondii was isolated from the newborn's peripheral blood 8 days after birth by mouse inoculation. The strain, called IPP-2002-URB, was virulent for mice at the time of isolation: all inoculated mice died 9 days after inoculation. The results of genotyping this strain with 5 microsatel-lite markers have been described elsewhere[3]; it exhibited an atypical genotype, one different from the 3 archetypal lineages of T. gondii designated as types I, II, and III.
Experimental model of reinfection. To test the hypothesis of reinfection, we created an experimental model of reinfection by IPP-2002-URB in mice chronically infected by a type II strain of T. gondii, PRU. We used a transfected PRU strain (PRU β galactosidase [PRU-β-Gal]) that expresses Escherichia coli β-galactosidase [4]. When submitted to X-gal staining, cysts of the transfected strain (PRU-β-Gal) appear blue, whereas those of the nontransfected strain (IPP-2002-URB) are not colored, which permits these 2 strains to be microscopically differentiated. Furthermore, these 2 strains can be distinguished by PCR assay by use of 1 microsatellite marker, B17, because the (TC)n dinucleotide tandem repeat of this marker is repeated 16 times in IPP-2002-URB and 7 times in archetypal type II strains such as PRU or ME49 [3]. The length polymorphism of these 2 different alleles (334 bp for PRU and 352 bp for IPP-2002-URB) can be detected by automated capillary electrophoresis by use of an ABI310 genetic analyzer and GeneScan software (Applied Biosystems), as described elsewhere [5]. Twelve mice were chronically infected by oral ingestion of PRU-β-Gal (20 cysts). Oral reinfection with 20 cysts of IPP-2002-URB was performed on 2 groups of 6 mice, at 1 and 4 months, respectively, after primary infection. Six weeks after reinfection, the brains of the mice were submitted to X-gal staining and microsatellite analysis with the B17 marker. By microscopic examination of each brain suspension, we identified reinfection in 4 of 5 mice in the 1-month group and 6 of 6 mice in the 4-month group (table 2). By microsatellite analysis, the 2 genotypes were detected in 4 of 5 reinfected mice in the 1-month group and 5 of 6 reinfected mice in the 4-month group (table 2).
In relatively few cases, congenital toxoplasmosis results from reactivation of a latent infection in pregnant women with altered immune status, such as that resulting from systemic lupus erythematosus or hematological malignancies [6] and HIV infection [7–9]. Some case reports involved congenital toxoplasmosis transmitted from immunocompetent mothers infected before conception, but clinical and serological data permitted researchers to date the occurrence of the primary infection to a few months before the onset of pregnancy. Lymphadenopathy was the most frequent clinical presentation, and it occurred 1 [10], 2 [6, 11, 12], or 3 months [13] before conception. In one case [14], a flu-like syndrome was reported by the mother 2 months before conception, whereas in another case [15], the recent infection was asymptomatic and could be identified only by serological data. In all these immunocompetent patients, the delayed transplacental transmission of Toxoplasma was explained by a persistent or recurrent parasitemia over a prolonged period.
To our knowledge, the case described in the present article is the sixth case of congenital toxoplasmosis as a consequence of maternal parasitemia following reinfection in an immunocompetent woman who had previous test results consistent with past toxoplasmic infection (table 3) that has been reported in the literature during the past 3 decades [16–20] . Our case is the first in which the Toxoplasma strain likely to be the origin of reinfection was isolated. In all 6 cases, the serological data collected before or at the beginning of the pregnancy ruled out a recent infection before conception (i.e., the presence of IgG and absence of IgM). Reinfection with another strain during pregnancy was a plausible explanation, with serological reactivation observed in the mother during gestation or at birth: increased titers of IgG were observed in all cases, the appearance of IgA was documented for 5 cases (IgA data was not reported for patient 5), and the absence of IgM was documented for all cases except patient 2 (for whom data showed the appearance of IgM at week 28 of amenorrhea). For 2 cases (patients 2 and 6 [the present case]), a flu-like syndrome consistent with an acute toxoplasmosis had been reported by the mother before the observed serological reactivation during pregnancy. To support the hypothesis of maternal reinfection during pregnancy, the authors of all 6 case reports documented the absence of any underlying immunodeficiency in the mothers. In the case we describe, the mother had active Hashimoto thyroiditis. In this autoimmune disease, a defect in a subpopulation of NK immune cells has been described [21]. As NK cells are critical for the initiation of an immunological intestinal response against T. gondii [22], it cannot be ruled out that this underlying disease could have favored oral reinfection.
Reported cases of congenital toxoplasmosis transmitted from immunized mothers who experienced reinfection during pregnancy, from literature of the past 3 decades.
The immune response of the mother might be overwhelmed when reinfection occurs as the result of a massive inoculum [16], a different infectious stage (oocyst vs. cysts) [17], or a strain with a different genetic background [20]. In the 6 cases of reinfection during pregnancy presented in table 3, epidemiological risk factors were present, such as contact with kittens (patients 1, 2, and 4), consumption of imported raw meat (horse meat was consumed by patient 6, described here), birthplace and residence, or travel abroad (patient 4 could have been exposed in Brazil and patients 3 and 5 in France after primary infection in Haiti and/or French Guiana and Angola, respectively).
All 6 cases of congenital toxoplasmosis after maternal reinfection were diagnosed because the newborns were symptomatic. In 3 cases (patients 2, 3, and 4), the clinical diagnosis was revealed by typical toxoplasmic chorioretinitis. In the 3 other cases (patients 1, 5 and 6), the clinical presentation was more severe, with a disseminated toxoplasmosis leading to spontaneous abortion in patient 1, cardiac abnormalities in the offspring of patient 5, and a bacterial sepsis-like infection in the infant described in this article, which is very uncommon in congenital toxoplasmosis [23]. These severe and unusual clinical presentations might be explained by an infection during pregnancy due to a highly virulent strain. In our case, the Toxoplasma strain (IPP-2002- URB) was isolated from the peripheral blood of the newborn, which traduces the high capacity for dissemination of this strain. Furthermore, the genetic analysis of the strain showed an atypical genotype that is very uncommon in France, where 96% of congenital toxoplasmosis is due to a single genotype, type II [24]. Conversely, type II strains seem rare outside Europe and North America. In South America, Toxoplasma strains are genetically more diverse, with many different genotypes described mainly in Brazil and the Guianas [3, 25–27]. These atypical South American strains, initially called “exotic” strains, belong to several haplogroups that are endemic to South America [28, 29]. Because a large amount of the horse meat eaten in France is imported from South America (Argentina, Uruguay, and Brazil), it is likely that the infecting strain in the case presented here originated from these countries, where similar atypical genotypes had already been described [3]. Even if there is still no evidence for a correlation between atypical strains and severe toxoplasmosis, disseminated cases in immunocompetent patients in French Guiana or Suriname [25, 30] and congenital toxoplasmosis [24] are more often associated with atypical genotypes of T. gondii.
The results of our experimental model confirmed reinfection by the atypical strain IPP-2002-URB in mice chronically infected by a type II strain of T. gondii. This experimental model had already been used to demonstrate the possibility of reinfection in mice by different T. gondii strains (type II challenged by type I or III) [4, 31, 32]. The definitive proof of reinfection in our patient would require the isolation and identification of the initial infecting strain, which would be nearly impossible to obtain. The only alternative would be serotyping [33, 34], with results showing different serotypes in the mother before and after serological reactivation, and in the newborn. Serotyping was attempted in this case, but was unsuccessful for the mother before reactivation because titers of anti-Toxoplasma antibodies were too low.
In France and probably elsewhere in Europe, congenital toxoplasmosis following maternal reinfection during pregnancy is exceptional. In these countries, past immunity usually protects against reinfection because generally only 1 genotype (type II) circulates in humans and in the environment [35, 36], and as a consequence, there is a very low probability of reinfection with different genotypes. However, in other areas, such as South America, where the genetic diversity of Toxoplasma is higher, reinfection with different genotypes is likely to occur more frequently. For instance, in Brazil, where cases of reinfection have been described for acquired toxoplasmosis [37], congenital toxoplasmosis due to reinfection during pregnancy might not be so exceptional. Thus, we assume that pregnant women from Europe who have been immunized against type II strains but are traveling in tropical areas or eat imported meat during pregnancy are at risk of reinfection by atypical strains. Conversely, as reported in this review (see patients 3 and 5 in table 3), women born in Africa or South America, and likely immunized against atypical Toxoplasma strains before conception, could also be at risk of reinfection by a type II strain when they migrate to European countries.
In conclusion, cases of reinfection with T. gondii involving an immunocompetent pregnant woman are exceptional but show that the presence of residual IgG-specific antibodies is not always synonymous with protection against a new Toxoplasma infection. To our knowledge, the case described here is the first in which the T. gondii strain involved in reinfection was isolated, characterized, and studied in an experimental mouse model that suggested immunity against a type II strain may not be protective against a reinfection with a different genotype, especially if it is atypical. Because there is increasing evidence that Toxoplasma strains are genetically more diverse outside Europe, health-care providers in Europe should be aware of this possibility when pregnant women travel abroad or eat imported meat.
We would like to thank our colleagues M. Robin and H. Vrillon from the maternity ward of Saint Maurice Hospital for providing serological results for the mother and Anne Dao and Jean-François Dubremetz for providing us the transfected strain PRU β-Gal.
Financial support: Centre National de Référence (CNR) Toxoplasmose (06-SMIP- 40–23); Toxoplasma Biological Resource Center.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy