Bridging the Gap: Conference Record [Abstract book, International Conference on AIDS (12th: 1998: Geneva, Switzerland)]

398 Abstracts 23283-23287

12th World AIDS Conference

23283 1 Prevention of vertical HIV transmission: Limits of
success at the population level
Christian Kind. The Swiss Neonatal Study Group; Division of Neonatology,
Kantonsspital CH-9007 St. Gallen, Switzerland
Background: In mid 1994 a recommendation for zidovudine (ZDV) prophylaxis
according to the ACTG 076 protocol for all pregnant, HIV-infected women was
issued in Switzerland. The effect of this recommendation was assessed on a
population basis and the circumstances of still occurring vertical transmission
characterised.
Methods: Data from an ongoing nationwide prospective study of children born
to HIV-infected mothers and from a national registry of children with vertically
transmitted HIV-infection were analysed. Two three-year periods were compared.
Period A:children born 7/1/1991 to 6/30/1994 and period B: 7/1/1994 to 6/30/1997.
Results: In period A 167 children of HIV-infected mothers were identified at
birth and 26 (15.6%) of them were found to be infected. Until 6 months after the
end of period A (up to 12/31/1994) 7 additional children born in period A and
diagnosed with HIV-infection later in life were announced to the registry. In period
B 138 children were identified at birth and 9 (6.5%) found to be infected (p = 0.014
vs. period A). 5 additional infected children born in period B were registered only
after birth until 6 months after the end of period B (up to 12/30/1997). Thus the
minimum estimated annual incidence of vertical HIV infection dropped from 1.4
per 10,000 to 0.6 per 10,000 (p = 0.009). In period B 54% of women known to be
HIV infected had a complete regimen of ZDV prophylaxis (>2 weeks antepartum,
iv intrapartum and 6 weeks postnatally to the newborn) and 85% had at least
some ZDV exposure. In 72% incomplete prophylaxis consisted of lacking or
late antepartum ZDV adminstration. Of the 9 transmitting mothers known to be
HIV-infected during pregnancy 4 refused ZDV-prophylaxis (2 of them breastfed
against medical advice) and 3 started antenatal ZDV late, shortly before giving
preterm birth. Among 45 women with complete ZDV prophylaxis and delivery by
elective cesarean section no case of transmission was observed. The frequency
of elective cesarean section in HIV-infected women increased from 24% in period
A to 50% in period B (p < 0.001).
Conclusion: The increased use of preventive measures (ZDV prophylaxis and
elective cesarean section) decreased the estimated annual incidence of vertical
HIV infection in Switzerland by more than 50%. More than a third of the remaining transmissions occur in women not identified to be HIV-infected before birth.
Intensified efforts for voluntary HIV screening early in pregnancy and counselling
for timely start of antiretroviral prophylaxis offer the best prospects for further
improvements.
23284 1A community based study of perinatal transmission of
HIV-1 and HIV-2 in the Gambia
Hilton Whittle', K.A. Ariyoshi', P.T.N. Ngom', P.M. Milligan', R.S.N. Sarge-Njie2,
D.O. O'Donovan', H.A. Wilkins'. 'MRC Laboratories, Banjul; 2Royal Victoria
Hospital, The Gambia
Objectives: To determine rates of, and risk factors for, mother to child transmission (MCT) of HIV-1 and HIV-2 in The Gambia.
Methods: From January 1993 to March 1995 we screened 29,670 pregnant
women attending antenatal clinics in The Gambia. Mothers and babies were
visited and examined by a clinician at 2. 9. and 18 months of age, when bloods
were taken for immunology and virology.
Results: We enrolled 109 HIV-1 seropositive women, 250 HIV-2 seropositive
women and 448 seronegative controls. Of the children born to HIV-1 seropositive
women. 7 of 64 were HIV-1 PCR positive at 9 months of age. and 6 of 11 children
who were dead or lost to follow-up by 9 months were HIV-1 PCR positive at 2
months of age. Four children who were PCR negative at 9 months of age were
HIV-1 seropositive and PCR positive at 18 months of age. MCT rate was 17/81
i.e. 21.0% (95% CI 2.7, 31.5).
Of the children born to HIV-2 seropositive women, 5 of 170 were HIV-2 PCR
positive at 9 months of age. None of the 22 children who were dead or lost to
follow up at 9 months of age and who were tested at 2 months were HIV-2 PCR
positive. Three children who were PCR negative at 9 months of age were HIV-2
seropositive and PCR positive at 18 months of age. MCT rate was 8/201 i.e. 4.0
(95% Cl 1.7, 7.7.).
A low CD4% and birth during the malaria season were risk factors for transmission. Plasma RNA HIV-2 viral load was higher in mothers who infected their
babies than in those who did not (GM 4722 vs 973 copies/mi respectively, p =
0.03). The HIV-1 plasma RNA viral load is being measured.
Mother to child transmission rate, which was influenced by season and state
of disease as determined by CD4 level or plasma viral load, was higher in HIV-1
than HIV-2 infected mothers. Late postnatal infection occurred in both infections.
|23285 Comparative importance of obstetric factors for vertical
HIV-1 transmission in Malawi and Brazil
Beatriz H. Tess', J.P.A. loannidis2, L.C. Rodrigues3, L. Mtimivalye4,

N. Kumwenda4, D.G. Contopoulos-loannidis5, R.J. Biggar'. Viral Epidemiology
Branch, NCI EPN/434 6130 Executive Bldg, Rockville, MD; 2HIV Research
Branch, NIAID, NIH, Rockville, MD; 5George Washington University,
Washington, DC, USA; 3London School of Hygiene Tropical Medicine, London,
UK; 4 University of Malawi, Blantyre, Malawi
Objectives: To assess the importance of obstetric factors on vertical transmission
of HIV-1 comparing results from two large cohort studies in Malawi and in Brazil.

Methods: A cross-validation approach was used to evaluate obstetric predictors similarly defined in Malawi (1359 mother-child pairs) and Brazil (434
mother-child pairs). A predictive model was built upon the Malawi data and was
tested independently in the Brazilian data, controlling for the effect of maternal
HIV disease. We compared the risk of vertical HIV-1 transmission according to
different indications of cesarean section (c-section) in the two settings.
Results: Duration of ruptured membranes (ROM) (OR per hour 1.02, 95% CI
1.00-1.04), gestational age (OR per week 0.93, 95% Cl 0.85-0.99) and maternal
stage of HIV disease (OR 1.68, 95% Cl 1.20-2.35) remained independently
associated with child's HIV-infection status in a multivariate analysis based on the
Malawi data and the model was fully validated in the Brazilian cohort (p = 0.01)
where the effects of these three predictors were similar. 14% of the deliveries
were by c-section in Malawi vs. 35% in Brazil. In Malawi, the transmission risks
for vaginal and c-section delivery were 26% (95% Cl 24-29) and 19% (95% Cl
14-25), respectively, whereas in Brazil these figures were 15% (95% Cl 11-19)
and 19% (95% CI 13-26). In Brazil, transmission risk was 21% (95% Cl 13-30)
for babies born by c-section for indications rarely used in Malawi but common
in Brazil (tubal ligation, prolonged ROM, multiple birth, previous c-section, postdate delivery, fetal distress and HIV infection); for indications which were equally
common in the two countries (dystocia, cephalo-pelvic disproportion, hypertensive conditions, abnormal fetal presentation, placental and cervical emergencies),
transmission risk was 13% (95% Cl 4-27).
Conclusions: The effects of ROM, gestational age and maternal disease stage
are consistent in Malawi and Brazil. Differences in the crude effect of the mode
of delivery on the risk of transmission may be partially due to differences in the
indications for c-section in the two settings. Such differences in indications need
to be considered before addressing the issue whether c-section per se may or
may not affect the risk of transmission.
S23286 The trophoblastic epithelial barrier is not infected in
full-term placenta of HIV-seropositive mothers
undergoing antiviral therapy
Charlotte Tscherning', N. Papadogiannakis2, S. Liindgren3, A.B. Bohlin4,
M. Anvret5, J. Albert6, E.M. Feny61. IMTC Karolinska Institute Box 280, 17177
Stockholm; 2Pathology Department Huddinge Hospital, Stockholm; 3Dpt of
Obstetrics Huddinge Hospital, Stockholm; 4Dpt of Pediatrics Hudding Hospital
Stockholm; 5Clinical Genetics Karolinski Hospital, Stockholm; 6Center of
Disease Control, Stockholm, Sweden
Aim: To determine if cells in the full-term placenta of HIV-seropositive mothers
are infected, and if so, which cell type is affected. This may provide important
information for the timing and mechanism of vertical transmission.
Material and Methods: Nine term placentas from HIV-1 seropositive mothers
and one term placenta from an HIV-2 infected mother have been prospectively
collected since october 1996 from delivery wards in Stockholm. The mothers
were all asymptomatic and treated during pregnancy according to the protocol
ACTG076. Freshly collected placentas were washed and digested with trypsin
and DNAse. The resulting cell suspensions were submitted to Percoll gradient
centrifugation to enrich for trophoblasts (85-92%). Depletion of T-lymphocytes
(CD3), Hofbauer cells and macrophages (CD14), granulocytic cells (CD45), endothelial cells (CD31) was performed with immunomagnetic beads coated with
the corresponding monoclonal antibody. FACS analysis showed that 95-98% of
the cells were positive for the GB25 marker. DNA was extracted separately from
the different cell populations and HIV-DNA PCR was performed. Up to date, all
children remained HIV-negative.
Results: All placentas were macroscopically and microscopically normal. Immunostaining for p24 antigen was negative. HIV-DNA was found in the enriched
trophoblastic cell populations and in the placental T-lymphocytes, but was regularly undetectable in the purified trophoblasts. Semi-quantitative PCR showed
a similar amount of HIV-DNA in the mother's PBMC and in the T-lymphocytes
of the placenta. Admixture of maternal cells to placental T-lymphocytes could be
demonstrated by the genomic microsatellite analysis.
Conclusion: Trophoblastic cells are not infected in full-term placenta of HIVseropositive mothers undergoing antiviral therapy. HIV-DNA was detected in
placental T-lymphocytes probably due to admixture of maternal cells. Considering that placental T-lymphocytes of maternal origin are HIV-DNA positive, the
trophoblastic barrier must have protected the fetus. Transmission, if it occurs,
is more likely towards the end of the pregnancy, when leakage in the placental
barrier frequently appears.
23287 | Evaluation of ZDV administration to pregnant women and
their children born in 1993 through 1996 in New Jersey
Sindy Paul, Helene Cross, L.G. Dimasi, S.J. Costa, J.K. Beil, D.H. Morgan. New
Jersey Dept of Hith & Senior SCS PO Box 363 Trenton NJ 08625-0363,
Issue: Perinatal HIV transmission can be prevented with the use of ZDV.
Project: NJ ranks third in the United States (US) in the number of pediatric
AIDS cases and has the highest percentage of women among its AIDS cases

in the US. 94% of the pediatric AIDS cases and 98% of HIV-infected children
in NJ are due to perinatal transmission. Information on children known to be
exposed to HIV perinatally, including pre-natal, delivery, and neonatal ZDV use is
collected. Data for years of birth 1993-1996, as reported through 12/31/97 were
analyzed. We also examined children known to serorevert since birth, but limited
our analyses to birth years 1993-1995. Children born in 1996 may not have had
the follow-up testing completed yet to confirm lack of infection.