| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The situation regarding infection with human immunodeficiency virus type 1 (HIV-1) in French Guiana is of great concern. This area reported an incidence for acquired immunodeficiency syndrome (AIDS) of 4,534 cases per million population in June 1998, which is higher than anywhere else in France (national rate = 799 cases per million).15 In contrast to HIV-1, where numerous reports have studied the incidence rates in various populations,16–18 HTLV-I incidence rates have been rarely reported.16,19–25 The goal of this study was to determine and to compare the seroprevalence trends and seroincidence of these two human retroviruses (which share most of their modes of transmission) in a large population of pregnant women belonging to different ethnic groups of French Guiana, and living in the same environment. We conducted two related studies. The first consisted of biennial cross-sectional evaluations of the serologic status (for both HTLV-I and HIV-1) of all deliveries during a 10-year survey in the obstetrics unit at the hospital in Saint-Laurent du Maroni. The second determined the incidence of these retroviruses in a dynamic cohort that included women who had at least two deliveries during the study period.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In this unit, maternal characteristics and serologic status are systematically recorded on the medical files for each pregnancy and delivery. We carried a dynamic cohort study including all the files for each hospital delivery from July 1, 1991 through June 30, 2001. After their consent was obtained, more than 90% of the women were tested for HTLV-I and HIV-1 when they came to the hospital to deliver their child during the survey period. In France, an HIV test is routinely conducted on every pregnant woman according to the recommendations of the French Ministry of Health. Zidovudine and more recently highly active anti-retroviral therapy have been recommended as treatment for all women found to be seropositive for HIV. Women seropositive for HTLV-I were informed of their HTLV-I infections and asked to bottle-feed their infants based on their socioeconomic status.
Serologic assays. Serum samples were screened for HTLV-I with an enzyme-linked immunosorbent assay (ELISA) (Cobas-Core enzyme immunoassay; Roche, Mannheim, Germany). All positive or borderline positive ELISA samples were then retested with a specific confirmatory Western blot test (Cambridge Biotech, Cambridge, MA). A Western blot result was considered positive for HTLV-I when it showed virus-specific bands corresponding to all major gag proteins (p19, p24, and p53), to the recombinant rgp21 protein, and the gp46 glycoprotein. A serum sample was considered indeterminate for HTLV-I when it showed only gag p19 and/or p24 and no env seroreactivity. The few samples indeterminate for HTLV-I were considered negative in the statistical analyses. For HIV, serum samples were screened using two ELISAs (Vidas HIV1/2 new; bioMérieux, Marcy l’Etoile, France and Recombigen HIV-1/HIV-2 RTDplus; Cambridge Biotech Diagnostics, Galway, Ireland). All positive or borderline ELISA results were retested by a confirmatory Western blot assay (NewLAV blot; Sanofi Diagnostic Pasteur, Marnela-Coquette, France). All HIV-seropositive women were infected with HIV-1; none were infected with HIV-2.
Statistical analysis.
To examine baseline differences between ethnic groups or seropositive and seronegative individuals, analysis of variance for continuous variables and Pearson chi-square statistic for categorical variables were used. While the overall serologic prevalence data were calculated based on all women tested during the survey (Tables 1
, 2, and 3), the evolution of the biennial seroprevalence rate (trend over the decade of observation) was calculated on all the mothers with a delivery during the period for which a serologic test was available (Table 4). The linear trends of seroprevalences were tested using the chi-square test for linear trend.
|
|
|
|
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) were enrolled since some had twins or two and more deliveries during the survey. Among them, 4,527 (65.4%) were Noir-Marron, 300 were Haitian, 576 were Amerindian, 352 were Creole, 163 were Hmong, and 1,003 were from miscellaneous ethnic groups including 215 French whites, 139 Guyanese, 67 Chinese, 108 Brazilian, 105 others, and 369 of unknown ethnic origin (Table 1). The mean ages, as well as the number of gestations and parities for each ethnic group, are presented in Table 1.
The serologic status for HTLV-I was available for 91.4% (6,331 of 6,921) of the mothers and HIV-1 serologic status was available for 95.1% (6,583 of 6,921) of the mothers (Table 1). There were 2,821 (40.8%) women with more than one delivery during the survey and 2,618 and 2,670 were tested for HTLV-I and HIV-1, respectively (Table 1).
Seroprevalence of HTLV-I and HIV-1.
Among all women tested, 218 (3.44%) of 6,331 were seropositive for HTLV-I (95% confidence interval [CI] = 2.99–3.89%) (Table 2), and 56 (0.85%) of 6,583 were seropositive for HIV-1 (95% CI = 0.63–1.07%) (Table 3). Seven women were coinfected with both viruses (five Noir-Marron and two Haitian).
The HTLV-I seroprevalence differed between ethnic groups and was mainly restricted to women of African origin, either Noir-Marron (181 of 4,266 = 4.24%; 95% CI = 3.64–4.84%) or Haitian (12 of 287 = 4.18%; 95% CI = 1.86–6.49%) (Table 2). There were no women seropositive for HTLV-I among the Amerindians, Creoles, or Hmongs, but 25 (3.12%; 95% CI = 1.92–4.32%) of other ethnic origins were seropositive. This group included 1 white woman, 7 from Guyana, and 17 of unknown ethnic origins. The seroprevalence of HTLV-I increased significantly with age in the Noir-Marron and Haitian women (Table 2).
During the 10-year survey period, based on deliveries, a significant (P = 0.037) decrease in the biennial seroprevalence of HTLV-I was observed among the Haitian women. However, this remained stable (P = 0.22) among the Noir-Marron women (Table 4).
The mean ages of the Haitian and Noir-Marron women were stable over time, as were the differences in mean age for women seropositive and seronegative for HTLV-I. Among the Noir-Marron, the annual trend of HTLV-I seroprevalence according to age group exhibited a decrease for those < 21 years old and those 27–32 years old (P = 0.004 and P = 0.025, respectively) and remained stable in the others age groups.
The seroprevalence of HIV-1 was higher in Haitian women (10 of 293 = 3.41%; 95% CI = 1.33––5.48%) than in Noir-Marron women (34 of 4,310 = 0.79%; 95% CI = 0.53–1.05%) and Amerindian women (4 of 552 = 0.72%; 95% CI = 0.01–1.43%) (Table 3). Among the women in the other ethnic groups, only 8 were infected with HIV-1, including 1 white woman from France, 4 from Guyana, and 3 of unknown origins (Table 3). Furthermore, a highly significant increase from 0.07% in 1991–1993 to 1.17% in 1999–2001) in the biennial seroprevalence of HIV-1 was observed among the deliveries of the Noir-Marron women (P = 0.0003), while a non-significant decrease was observed in the Haitian women (P = 0.44) (Table 4).
Seroincidence rates of HTLV-I and HIV-1.
The dynamic cohort included each woman with at least two deliveries during the survey period (Table 1). Among all women included in this prospective survey, 73% (2,064 of 2,821) were Noir-Marron (Table 1), and 1,966 and 1,961 were tested for HTLV-I and HIV-1, respectively. The follow-up for the calculation of the incidence was based on women at risk (initially seronegative): 1,872 for HTLV-I and 1,953 for HIV-1 (Table 1). Most, of the seroconversions (10 of 13 for HIV-1 and 15 of 17 for HTLV-I) occurred among the Noir-Marron women. Thus, the seroincidence rates of HTLV-I and HIV-1 were studied only in this specific population. None of the HIV-1 and HTLV-I seroconversions occurred in the same woman.
For HTLV-I, the total follow-up period was 8,283 person-years (median = 50 months) with 15 seroconversions (of 1,872 women), giving an incidence rate of 0.181 per 100 person-years (95% CI = 0.180–0.182). The incidence rate was 0.174 per 100-person years (95% CI = 0.173–0.175) for those 
25 years old at the last delivery and 0.196 per 100 person-years (95% CI = 0.195–0.197) for those > 25 years old at the last delivery.
For HIV-1, the total follow-up period was 8,295 person-years (median = 51 months) with 10 seroconversions (of 1,953 women), giving an incidence rate of 0.120 per 100 person-years (95% CI = 0.119–0.121). The incidence rate was 0.138 per 100 person-years (95% CI = 0.137–0.139) for those 25 years old at the last delivery and 0.018 per 100 person-years (95% CI = 0.0179–0.0181) for those > 25 years old at the last delivery.
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Among the five main ethnic groups studied, infection with HTLV-I was clearly restricted to the populations of African descent (Noir-Marron and Haitian), and did not spread to others ethnic groups who shared the same environment. Such ethnic/African restriction is probably a founder effect, followed by a high level of persistence of viral transmission.12–14,26 In this remote population, bottle-feeding, use of condoms, contraception, and blood transfusions were rare, and po1ygamy existed, which implies high levels of maternal and sexual transmission without the dilutional effect of migration or low mixing with other ethnic groups. Pregnancies and deliveries were the main opportunity for women of this population to be tested for retroviruses.
An HTLV-I incidence rate of 0.18% person-years in pregnant Noir women confirms the findings of our preliminary study, which was conducted for a shorter time period and on a much smaller population in the same area.14 Furthermore, this rate is similar to a rate (0.2% person-years) found in a population-based study of women > 40 years of age in an area of Japan endemic for HTLV-I.19 However, it is much lower than in groups with a high risk of sexually transmitted diseases (STDs), such as policemen in Guinea-Bissau (0.4% person-years)20 or women attending STD clinics in Jamaica (0.8% person-years), and mostly female sex workers (0.8% person-years in Japan, 1.6% person-years in Peru, and 0.7% person-years in Zaire).16,21–23 In a recently published study based on prospective work in Jamaican food handlers, van Veldhuisen and others estimated the seroincidence of HTLV-I to be 0.32% person-years for women.24 This is slightly higher than that in our study. One possible reason for this difference was that we studied a younger population (mean age = 25 versus 32 years). Nevertheless, women seropositive for HTLV-I were informed of their infections and asked to bottle-feed their infants, but not to terminate their pregnancies or reduce their numbers.
Several explanations for our finding of an age-related increase in HTLV-I seropreva1ence, as reported in many areas highly endemic for this virus,6,8,9 can be made. These include the accumulation of seroconversions over the lifetime of the individuals, an age cohort effect due to a rapid decrease in HTLV-I prevalence with very high levels over the past two or three generations, and a combination of both explanations. In Japan, where the westernization of life styles started in the 1950s and included bottle-feeding, shortening the duration of breast-feeding, and a higher frequency of condom use, several studies have reported a higher prevalence of HTLV-I in people > 40 years old.5,6,25,27,28 In our study, a significant decrease in annual HTLV-I seroprevalence was observed in the Noir-Marron women, especially in women < 21 years old (P = 0.004). This fact could reflect the emerging consequences of the recent modifications of the Noir-Marron life style that include a diminution of the duration of breast-feeding, a decrease in the number of blood transfusions, and better medical care. An underestimation of the prevalence of HTLV-I was unlikely because the rate of new pregnancies after serologic testing did not differ significantly for HTLV-I negative or HTLV-I positive women < 31 years old (58% versus 49%; P = 0.1)
The level of infection with HIV-1 in French Guiana is much higher than in all other areas of France.15 In this region, 80% of the AIDS cases are linked with heterosexual contacts and 52% of cases are in the Haitian community.29 In our study, the highest prevalence of HIV-1 was also found in Haitian women (3.41%), as observed in a previous study on all pregnant women in French Guiana in 1994–1995 (4.56%°).30 We also observed a decrease in the seroprevalence of HIV-1 in Haitian women, as seen in the entire pregnant population of French Guiana (4.6% in 1988 and 2.1% in 1996).30 This decrease in the prevalence of HIV-1 in Haitian women could be explained by 1) infections acquired in Haiti before migration, 2) a decrease in prevalence linked to deaths of infected women, 3) a decrease in reproductive activity following counseling, and 4) a low incidence rate of infection with HIV-1 in young Haitian women born in French Guiana or who had immigrated with their parents. In this context, the increase in the seroprevalence of HIV-1 in Noir-Marron women is striking and cause for concern. A 10-fold increase in the sero-prevalence of HIV-1 was observed over the last 10 years (from 0.07% in 1991–1993 to 1.17% in 1999–2001; Table 4), with a high incidence rate in the youngest women. An underestimation of such prevalence rates could be the result of counseling to terminate pregnancies and condom use when positive results were found. In our study, the rate of new pregnancies after serologic testing vary significantly between HIV-negative and HIV-positive women < 31 years old (50% versus 30%; P = 0.03). Such findings clearly indicate that a comprehensive HIV prevention program of condom use and pregnancy counseling should be more intensive in such culturally isolated populations with low socioeconomic levels. Only a strong and specifically adapted effort will result in a decrease in the prevalence of HIV-1 in this large ethnic group whose population is rapidly increasing both in French Guiana and Surinam.
In conclusion, these data demonstrate the differences in the spread of two human retroviruses among pregnant women in different ethnic groups living in the same environment. Unlike HTLV-I, which is highly endemic in populations of African origin and shows a slight decrease over time with a suspected cohort effect, HIV-1 spreads rapidly on an epidemic mode, especially in the populations of the lowest socioeconomic level, in which a major prevention effort is urgently needed.
Received October 28, 2004. Accepted for publication December 15, 2004.
Acknowledgements: We are indebted to the Service de Protection Maternelle et Infantile de la Direction Départementale des Affaires Sanitaires et Sociales–Guyane (Dr. J. Suzanon-Sankale), the Centre de Prévention et de Vaccination of St. Laurent du Maroni (Dr. M. Diop, D. Gaquière, G. Gauthier, Dr. D. Guillot, N. Joly, and Dr. Z. Tlilane), the staff of the Gynecology-Obstetrics department, especially Olivier Pavot and Diane Vernon, and the staff of the Central Laboratory of Biology of André Bouron Hospital for providing technical assistance and help in the collection of information.
* Address correspondence to Patricia Tortevoye, Unité d’Epidémiologie et Physiopathologie des Virus Oncogénes, Département d’Ecosytéme et Epidémiologie des Maladies Infectieuses, Batiment Lwoff, Institut Pasteur, 25–28 Rue du Dr. Roux, 75724, Paris, Cedex 15, France. E-mail: ptortevo{at}pasteur.fr 
Authors’ addresses: Patricia Tortevoye and Antoine Gessain, Unité d’Epidémiologie et Physiopathologie des Virus Oncogènes, Département d’Ecosytème et Epidémiologie des Maladies Infectieuses, Batiment Lwoff, Institut Pasteur, 25–28 Rue du Dr. Roux, 75724, Paris, Cedex 15, France, Telephone: 33-1-45-68-89-37, Fax: 33-1-40-61-34-65, E-mails: ptortevo{at}pasteur.fr and agessain{at}pasteur.fr. Philippe Tuppin, Etablissement Français des Greffes, Paris, France. Gabriel Carles, Service de Gynécologie-Obstétrique, Centre Hospitalier Frank Joly, 97320, Saint Laurent du Maroni, Guyane Française, France. Christian Peneau, Service de Biologie Médicale, Centre Hospitalier Frank Joly, 97320, Saint Laurent du Maroni Guyane Française, France.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. L.-D. Etenna, M. Caron, G. Besson, M. Makuwa, A. Gessain, A. Mahe, and M. Kazanji New Insights into Prevalence, Genetic Diversity, and Proviral Load of Human T-Cell Leukemia Virus Types 1 and 2 in Pregnant Women in Gabon in Equatorial Central Africa J. Clin. Microbiol., November 1, 2008; 46(11): 3607 - 3614. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
