Abstract Book Vol. 2 [International Conference on AIDS (14th: 2002: Barcelona, Spain)]

XIV International AIDS Conference

Abstracts ThPeB7234-ThPeB7236 395

objective was to compare accrual of total body bone mineral content (TBBMC) in
HIV+ children with and without LD.
Methods: TBBMC accrual was measured in grams by dual energy x-ray absorptiometry in prepubertal HIV+ children ages 4-12 yr between two visits separated
by 1.2+0.5 yr. All subjects were receiving antiretrovirals; 12 were taking protease
inhibitors. LD was defined as an increase in truncal fat with a decrease of combined arm and leg fat as measured over the two visits. Crude differences between
the LD+ and LD- groups were assessed by Student t-tests. Multivariate analyses
were used to control for potential confounders (age, sex, race, weight, and height).
Results: Eight children (29%) were classified as LD+. The mean unadjusted
TBBMC was lower in LD+ children at both visits 1 and 2 (888.0~197.4 vs
902.2~314.0 g and 964.5~238.7 vs 1022.1~371.2 g, respectively), but these differences were not significant. No significant differences in bone mineral accrual
from visit 1 to visit 2 were observed (76.5~114.0 g in LD+ vs 119.9~84.7 g in
LD-). At both visits, the mean unadjusted total bone density (TBD) was lower in
LD+ versus LD- children but the differences were not significant (0.84~0.03 vs
0.85~0.08 and 0.84+ 0.05 vs 0.88~0.10). The differences in log TBBMC and log
TBD remained insignificant after adjusting for age, sex, race, height, and weight
in the multivariate regression models.
Conclusions: TBBMC accrual does not differ in HIV+children with and without
LD. Using objective measurements and criteria are necessary to determine the
relationships between LD and bone abnormalities and associated risk factors.
Presenting author: Stephen Arpadi, St.Luke's-Roosevelt Hospital Center, 1111
Amsterdam Ave., New York, NY, 10025, United States, Tel.: [phone],
Fax: [phone], E-mail: [email protected]
ThPeB7234 Growth, morbidity and mortality in a cohort of
institutionalised HIV-1- infected African children
L.M.K. Nerlander', L.M. Nathan2, J. Dixon1, R. Barnabas1, B.E. Wholeben3
R. Musoke4, T. Palakudy4, A. D'Agostino4, R. Chakraborty'. 'Oxford University,
Balliol College, Oxford, OX1 3BJ, United Kingdom; 2Rush University, Chicago,
United States; 3Northern Illinois University, DeKalb, United States; 4Nyumbani
Children's Home, Nairobi, Kenya
Background: As a result of the HIV epidemic in Africa, much debate exists on
whether institutionalised compared to community-based care provides optimum
management of infected children. Previous reports calculated 89% mortality by
age 3 amongst out-patients in Malawi. No similar data is available for infected children in institutionalised care. We characterised patterns of morbidity and mortality
among a cohort of HIV-1 infected children residing at an orphanage in Nairobi.
Methods: 174 children were followed longitudinally over 5 years. Mortality was
analysed by Cox proportional hazards model to account for survival within the
community before admission. Anthropometric indices were calculated using standardised CDC growth curves and included mean z-scores for weight-for-length
and length-for-age. Low indices reflected wasting and stunting, respectively. Opportunistic infections and nutritional intake were documented. All children were
naive to antiretroviral therapy.
Results: 64/174 children died during follow-up. Survival was 70% at age 3 and
29% at age 10 (fig. 1) Mean z-scores for length-for-age suggested overall stunting
(z = -1.65). Wasting was not observed (z = -0.39). All children received adequate
nutrition and micronutrient supplementation. Morbidity included recurrent respiratory tract infections, gastroenteritis, parotitis and LIP, and candidiasis. No new
cases of TB disease were noted after admission to the home.

ThPeB7235 A successful program to prevent vertical
transmission of HIV in Sorocaba, Brazil
M.A.O. Oliveira, I.G. Gongorra, M.FEL. Mendonga, R.L. Louris, E.A. Adade,
V.L. Lippe, J.R.P. Marins. C/linica municipal STD/AIDS, Sorocaba, Brazil
Background: Since ACTG 076 demonstrated that vertical transmission can be
prevented, guaranteeing access to this intervention has been a worldwide challenge. In 1998, we implemented a program to screen pregnant women for HIV
and prevent vertical transmission in the public health services of Sorocaba, a ctiy
of 500,000 in Sao Paulo State.
Methods: We trained personnel in the city's 26 public clinics in the importance
of and procedures for HIV testing with pre and post counseling for all pregnant
women. All HIV+ women identified were guaranteed access to infant formula and
full medical care, including CD4 and viral load testing and antiretroviral treatment
to prevent vertical transmission and for themselves. Clinical data were included
through medical record review. This study did not include women already known
to be HIV+ prior to pregnancy
Results: The public services attended 7010 births out of 12,236 total in Sorocaba
in 2000. 6804 received prenatal care (97%), and 6481 of these (95%) received
HIV testing. 28 (0.4%) were HIV+; of these, all but one (96%) completed the
ACTG 076 protocol. The woman who refused had a viral load of 240,000. Despite receiving AZT after birth, her baby died at two months of an opportunistic
infection and was presumably HIV infected. Of the women discovered to be HIV+,
17 underwent CD4 testing while still pregnant; 4 were <200, 3 were 201-350, 5
were 351-500, and 5 were >500.
Conclusions: This program identified 28 HIV-infected pregnant women, resulting
in substantial health benefit for them and their babies. Only one case of vertical
transmission occurred compared to nine that would have been expected without
treatment based on a 30% rate. Several of the women had advanced disease
that required treatment, and all received counseling to avoid transmission to others. Although this program required substantial effort, we believe the results more
than justify the resources invested.
Presenting author: maria jose souza, rua da penha, 770, centro, sorocaba, sao
paulo, 18030-210, Brazil, Tel.: [phone], Fax: [phone], E-mail:
[email protected]
ThPeB7236 Reduced odds of livebirth associated with HIV-RNA
viral load: Rakai, Uganda
R.H.N. Nguyen1, S.J. Gange1, D. Serwadda2, F. Wabwire-Mangen2,
N.K. Sewankambo2, M.J. Wawer3, T.C. Quinn4, R.H. Gray'. 'Bloomberg School
of Public Health, Johns Hopkins University, Baltimore, MD, United States;
2Makerere University Kampala, Uganda; 3 Mailman School of Public Health,
Columbia University, New York, United States; 4School of Medicine, Johns
Hopkins University, Baltimore, United States
Background: HIV status and duration of infection are associated with reduced
likelihood of becoming pregnant and maintaining pregnancy The role of HIV-RNA
viral load on pregnancy rates and outcome is unknown.
Methods: A case-control nested in a population cohort used data from Rakai,
Uganda. Cases were women who delivered a livebirth during study follow-up
(n=270). Controls were randomly selected women who never became pregnant
during the study follow-up despite being sexually active and non-contracepting
(n=263). Controls were frequency matched by age (~5 years) to cases. Level
of serum HIV-RNA was determined using Roche Amplicor, version 1.5 and log
transformed for distribution normalcy Log viral load was categorized into 0.25 log
viral load increments for analysis. Chi-square statistics were used to test differences between viral load medians. Adjusted odds ratios (adj ORs) for livebirth
were calculated by conditional logistic regression models adjusting for age. The
lowest viral load category, 2.5 - 2.74, was the referent.
Results: The median age of women who had a livebirth was 24 years (range:
15 - 43). 26 of 270 (9.63%) cases and 28 of 263 (10.68%) controls had a viral
load below the level of detection (400 copies/mL). The median viral load for cases
was 13,250 copies/mL and 25,900 for controls (p-value<0.0001). Significant differences in median viral load remained after age stratification in all but 15-19 and
25-29 year olds. A progressive decrease in the odds of pregnancy was observed.
For each 0.25 log increment in viral load above 4.75 log serum HIV-RNA, the
adjusted odds ratios were:
Log Serum HIV-RNA Adj Odds Ratio 95% CI
4.75 - 4.99 0.65 0.30 - 1.38
5.00 - 5.24 0.64 0.25 - 1.61
5.25 - 5.49 0.11 0.02 - 0.54
5.5+ 0.08 0.02 - 0.36
Conclusions: Median viral loads were substantially lower among women who
had a livebirth versus those who did not become pregnant. A marked reduction
in adjusted relative odds of livebirth with increasing viral load categories greater

than 4.75 log viral load (56,234 copies/mL) was observed.
Presenting author: Ruby Nguyen, Room W4030 Hygiene Bldg, 615 N. Wolfe
Street, Baltimore, Maryland, 21205, United States, Tel.: [phone], Fax:
[phone], E-mail: [email protected]

Proportion surviving
Pointiwise 96% onfidence limits
1,I
1 1
Years
Survival, accounting for admission rate.

Conclusions: Our survival data suggested that the optimal form of care for HIVinfected children in resource-poor settings may be the development of similar
homes to the one described. Absence of TB disease among long-standing residents (probably due to decreased exposure of droplet nuclei from adults) may
have contributed to improved survival. Stunting in the absence of wasting implied
that growth was compromised by opportunistic infections. Similar studies elsewhere are needed to confirm our findings.
Presenting author: Lina Nerlander, Balliol College, Oxford, OX1 3BJ, United
Kingdom, Tel.: [phone], Fax: [phone], E-mail: lina.nerlander
@balliol.ox.ac.uk