Maternal and congenital toxoplasmosis in Mayotte: Prevalence, incidence and management

Background Toxoplasmosis is an infection caused by an intracellular protozoan, Toxoplasma gondii. It is usually asymptomatic, but toxoplasmosis acquired during pregnancy can cause congenital toxoplasmosis, potentially resulting in fetal damage. Epidemiological information is lacking for toxoplasmosis in Mayotte (a French overseas territory). We evaluated (1) the prevalence of maternal toxoplasmosis, (2) the incidence of maternal and congenital toxoplasmosis, and (3) the management of congenital toxoplasmosis in Mayotte. Methodology / Principal Findings We collected all the available data for toxoplasmosis serological screening during pregnancy and maternal and congenital cases of toxoplasmosis obtained between January 2017 and August 2019 at the central public laboratory of Mayotte (Mamoudzou). Using toxoplasmosis serological data from samples collected from 16,952 pregnant women we estimated the prevalence of toxoplasmosis in Mayotte at 67.19%. Minimum maternal toxoplasmosis incidence was estimated at 0.29% (49/16,952, 95% CI (0.0022–0.0038)), based on confirmed cases of maternal primary infection only. The estimated incidence of congenital toxoplasmosis was 0.09% (16/16,952, 95% CI (0.0005–0.0015). Missing data made it difficult to evaluate management, but follow-up was better for mothers with confirmed primary infection and their infants. Conclusions / Significance The seroprevalence of toxoplasmosis among pregnant women and the incidence of toxoplasmosis are higher in Mayotte than in mainland France. There is a need to improve the antenatal toxoplasmosis screening and prevention programme, providing better information to physicians and the population, to improve management and epidemiological monitoring.


Background
Toxoplasmosis is an infection caused by an intracellular protozoan, Toxoplasma gondii . It is usually asymptomatic, but toxoplasmosis acquired during pregnancy can cause congenital toxoplasmosis, potentially resulting in fetal damage. Epidemiological information is lacking for toxoplasmosis in Mayotte (a French overseas territory). We evaluated (1) the prevalence of maternal toxoplasmosis, (2) the incidence of maternal and congenital toxoplasmosis, and (3) the management of congenital toxoplasmosis in Mayotte.

Methodology / Principal Findings
We collected all the available data for toxoplasmosis serological screening during pregnancy and maternal and congenital cases of toxoplasmosis obtained between January 2017 and August 2019 at the central public laboratory of Mayotte (Mamoudzou). Using toxoplasmosis serological data from samples collected from 16,952 pregnant women we estimated the prevalence of toxoplasmosis in Mayotte at 67.19%. Minimum maternal toxoplasmosis incidence was estimated at 0.29% (49/16,952, 95% CI (0.0022-0.0038)), based on confirmed cases of maternal primary infection only. The estimated incidence of congenital toxoplasmosis was 0.09% (16/16,952, 95% CI (0.0005-0.0015). Missing data made it difficult to evaluate management, but follow-up was better for mothers with confirmed primary infection and their infants.

Conclusions / Significance
The seroprevalence of toxoplasmosis among pregnant women and the incidence of toxoplasmosis are higher in Mayotte than in mainland France. There is a need to improve the antenatal toxoplasmosis screening and prevention programme, providing better information to physicians and the population, to improve management and epidemiological monitoring. Toxoplasma gondii. It is usually asymptomatic, but toxoplasmosis acquired during pregnancy 28 can cause congenital toxoplasmosis, potentially resulting in fetal damage. Epidemiological 29 information is lacking for toxoplasmosis in Mayotte (a French overseas territory). We evaluated 30 (1) the prevalence of maternal toxoplasmosis, (2) the incidence of maternal and congenital In 2019, Mayotte had 270,000 inhabitants, with about 9,600 births recorded by the 91 regional health agency each year. Seroprevalence in mainland France has been decreasing since 92 the 1960s, and is currently estimated at about 30% [11,13]. Little information about 93 seroprevalence is available for Mayotte, and the real figures on the ground may be very different 94 from those for mainland France, because of the tropical weather conditions, which may 95 facilitate the survival of oocysts in the environment (hot and humid climate, long periods of 96 rainfall) and high rates of reservoir infection (favoured by the large number of young feral cats). 97 Moreover, this island has a high population density, with many living in precarious conditions, 98 high rates of poverty and illiteracy, and a diet very different from that in mainland France [14]. 99 The prevalence of TX could therefore be higher in Mayotte than in mainland France, and studies The results for the 270 samples testing positive for TX-IgM collected from pregnant women 184 were (Fig 1): 185 -46/270 (17.78% (95% CI (0.14-0.23)) tested negative for TX-IgG, and 27/46 (58.70%) 186 subsequently seroconverted and were considered to be cases of CPI; 187 -224/270 (82.96%, 95% CI (0.78-0.87)) tested positive for TX-IgG. In 59/224 (26.34%, 188 95% CI (0.21-0.32)) cases, TX-IgG avidity was low. In 22/224 (9.82%, 95% CI (0.07-189 0.14)) cases, an analysis of TX-IgG kinetics confirmed CPI.  days)) and TX-PCR was also performed on blood for 78.00% (39/50) (Table 2). Overall, 16/50 227 (32.00%) children had CCI (Fig 2).    Given the large numbers of missing data for children born to women with PPI and NEPI, 240 only a minimum incidence of congenital TX was calculated, based on the demographic records 241 of children born to women with CPI. We estimated that 17,240 infants were born to the 16,952 242 mothers with available serological results during the study period. The minimal incidence was, 243 therefore, 9.28/10,000 liveborn infants (16/17,240, 0.09% (95% CI (0.0005-0.0015)). 244 245

246
For infants born to mothers with CPI, treatment information was available for 24.00% 247 (12/50) cases and imaging data were available for 98.00% (49/100). More than 90.00% of the 248 data were not reported in medical records (serological tests, PCR, treatment and imaging data) 249 for infants born to mothers with PPI or NEPI (Table 2). 250 251 Less than half the infants with CCI (43.75%, 7/16) underwent eye fundus examination, 252 and the results were normal in all cases (Table 3). More than half underwent transfontanelle 253 cerebral ultrasound examinations (62.50%, 10/16), the results of which were normal in 80.00% 254 of cases (8/10) (   Toxoplasmosis prevalence is lower in countries in which meat is eaten well cooked (United 276 Kingdom, North America). It is slightly higher in mainland France, where meat is often eaten 277 raw or smoked [9]. In Mayotte, meat is also often eaten raw or smoked, or cooked on a barbecue, 278 which does not destroy toxoplasmosis cysts. Another important factor influencing prevalence is socioeconomic conditions, including, in particular, access to clean water, and proximity to 280 animals (especially feral cats) that can serve as major reservoirs of T. gondii. Hot and humid 281 climatic conditions also favour the spread of toxoplasmosis [26,27]. One of the strengths of 282 our study is its exhaustivity of about 67% for the 2017/2019 period. However, prevalence may 283 well be underestimated in this study, as at least 20% of women in Mayotte are thought to illegal 284 immigrants, with an even lower socioeconomic status, and many of these women may not have 285 undergone serological testing at all during their pregnancy. 286

287
The incidence of TX during pregnancy varies with prevalence. The incidence of TX 288 during pregnancy is well-documented in mainland France, due to the mandatory serological 289 screening programme. This incidence has decreased over time, reaching a predicted 1.6/1,000 290 in 2020 [8,21]. In Mayotte, we found the incidence of maternal TX infection to be 2.89 to 291 4.72/1,000 per year. This higher incidence of maternal TX is consistent with the prevalence of 292 67.19%, and is significantly different from that of mainland France (p-value <0.05). However, 293 this rate is probably underestimated due to the large number of cases that could not be formally 294 classified. Indeed, one of the limitations of our study is the inconsistency of serological follow-295 up in pregnant women. In many cases, only one serological test was performed during 296 pregnancy, often more than four months after conception. This lack of follow-up is due to local 297 factors, such as a lack of awareness in women of child-bearing age, the late diagnosis of 298 pregnancy and frequent misunderstanding of the French language (French is the mother tongue 299 of only 10% of the mothers in Mayotte). This last factor is almost certainly also a major reason 300 for the lack of awareness concerning behaviours at risk for toxoplasma infection and for women 301 not following primary toxoplasmosis prevention rules. 302 For the management of maternal TX, our study highlights a greater acceptance of non-304 invasive examinations, such as ultrasound scans, rather than amniocentesis (in women with CPI 305 particularly: 49.98% (24/49; 95% CI (0.36-0.63) and 4.08% (2/49; 95% CI (0.01-0.14), 306 respectively). PCR on amniotic fluid is an excellent tool for confirming foetal infection and 307 ensuring optimal management of the child at birth, particularly in terms of reducing unnecessary 308 The worldwide incidence of congenital TX ranges from 1 to 30/10,000 live-born infants 317 [9, 31]. In mainland France, the incidence is about 2 to 3 cases/10,000, and this rate has 318 remained stable since 2007 [11,12]. In Mayotte, this incidence was estimated at 9.28/10,000 319 liveborn infants, which is three times the rate in mainland France (p-value <0.05). This 320 difference may, of course, be due to the higher incidence of maternal TX, but it may also reflect 321 a late introduction of treatment due to irregular and/or delayed serological follow-up, which 322 concerns more than just TX. One limitation of our results was the significant number of missing 323 data, but this lack of information may suggest that treatments were not prescribed or not taken 324 Slovenia, consider congenital TX to be a public health problem and deliver advice about its 340 prevention, but have yet to establish an effective congenital TX surveillance system. Austria This study is the first to report epidemiological data on maternal and congenital TX 365 management in Mayotte and highlights possibilities for improvements, although many issues 366 remain to be addressed. 367 368