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

XIV International AIDS Conference

Abstracts ThPeB7216-ThPeB7219 391

ThPeB721616 Brazilian network for drug resistance surveillance:
results from the first national survey
A. Tanuri1, M.A. Soares1, R.M. Brindeiro1, R.S. Diaz2, L.J. Costa2,
E.C. Sabino3, N. Gaburo Jr.3, L. Brigido4, R. Rodrigues4, I.L. Pires5,
F. Esperanza5, M.M. Morgado6, J. Fernandez6, C. Lauria7, M.C. Dantas8,
P.R. Teixeira8. 1Universidade Federal do Rio de Janeiro, Depto de Genetica -
UFRJ, CCS - BI. A - sala A2-121, Cidade Universitaria - 1/ha do Fundao,
21944-970 Rio de Janeiro - RJ, Brazil; 2Universidade Federal de S&o Paulo, S&o
Paulo, Brazil; 3Universidade de S&o Paulo, S&o Paulo, Brazil; 41nstituto Adolfo
Lutz, S&o Paulo, Brazil; 51nstituto de Biologia do Exdrcito, Rio de Janeiro, Brazil;
6Fundagqo Oswaldo Cruz, Rio de Janeiro, Brazil; 7 Hospital Pedro Ernesto, Rio
de Janeiro, Brazil; 8 CN-DST/AIDS, Brasilia, Brazil
Background: The Brazilian Ministry of Health (BMoH) has sponsored universal free access to HIV antiretroviral drugs for AIDS patients since 1996. This
widespread use poses major concerns regarding the emergence of drug-resistant
viral strains. BMoH, therefore has implemented a National Network for Drug Resistance Surveillance in the drug-naive population. This network aims to measure
the prevalence of drug resistance mutations in the protease (PR) and reverse
transcriptase (RT) isolated from antiretroviral (ARV) nafve individuals.
Methods: Whole-blood was collected from 224 HIV-infected individuals originated from 21 cities in 9 different Brazilian states. Plasma RNA was isolated from
these samples, and at least one genomic region (PR or RT) was amplified in 154
(68.8%) of them. For most of the samples, both genes were successfully amplified
and sequenced. Sequences were subtyped and genotyped for drug resistance for
all commercially available ARV.
Results: One hundred twenty seven samples were sequenced for the RT, and
122 for the PR region. Clade distribution was B (55%), C (26%), F(9%), and mosaic genomes composed by B/C and B/F subtypes (10%). A number of primary
and secondary resistance mutations were found for NRTI, NNRTI and PI. Overall, resistance frequencies were 1.9%, 3.9% and 1.3% for those three classes
of drugs, respectively A number of polymorphisms in positions associated with
resistance were also observed.
Conclusions: This is the first national-based surveillance in Brazil documenting primary resistance mutations for antiretrovirals in drug-naive subjects. These
findings strengthen the need for monitoring drug resistance in sentinel populations where treatment is provided in large scale. On the other hand, despite the
widespread use of ARV in Brazil, this study has shown that the frequency of ARVrelated mutations is still low in recently infected naive individuals when compared
to those of developed countries.
Presenting author: Amilcar Tanuri, Depto de Genetica - UFRJ, CCS - BI. A - sala
A2-121, Cidade Universitaria - Ilha do Fundao, 21944-970 Rio de Janeiro - RJ,
Brazil, Tel.: [phone], Fax: [phone], E-mail: atanuri@biologia.
ufrj.br
ThPeB7217 The pattern of HIV-1 primary resistance in Italy after
HAART therapy
M. Violin1, A. Cozzi-Lepri2, R. Piscopo3, A. Giacometti4, C.F. Perno5,
A. d'Arminio Monforte6, C. Balotta6. 1lnstitute of Infectious and Tropical
Diseases, University of Milan, Institute of Infectious and Tropical Diseases,
University of Milan, Via GB Grassi 74, 20157 Milan, Italy 2Royal Free Centre for
HIV Medicine London, London, United Kingdom; 3Infectious Diseases
Department, 'Galliera' Hospital, Genova, Genova, Italy; 4Institute of Infectious
Diseases, University of Ancona, Ancona, Italy; 51RCCS 'L Spallanzani, Rome,
Rome, Italy; 6Institute of Infectious and Tropical Diseases, University of Milan,
Milan, Italy
Background: The expected prevalence of primary resistance to antiretrovirals
(ARV) is a function of the proportion of patients achieving a sustained control of
viral replication on ARV, actual levels of transmitted drug mutations (mts) and fitness of resistant mutants. The aim of this analysis was to assess the prevalence
of transmitted mts over the time period 1996-2000 in a population of recently
infected individuals (SCs) in Italy.
Methods: We studied 95 SCs (male,81%; HE, 40%, HO, 32%; IDU, 21%) of the
Italian Cohort Naive for Antiretrovirals. Pre-therapy plasma sample were analysed
by sequencing of protease and RT regions using an Applied Biosystems kit. Clade
assignment was conducted by phylogenetic analysis tools of the Phylip package.
Proportions of patients with resistant mts were compared using a chi-square test
and confidence intervals for proportions.
Results: Overall, 9 NRTIs-and 5 PIs associated mts were detected. No NNRTI
mts were present in the study period. The prevalences of SCs with primary mts by
calendar year to either NRTIs or PIs were 9.1, 35.5, 3.8, 5 and 0 % from '96 to'00.
Concomitantly, in the whole I.Co.N.A. the proportions of patients with a plasma
viral load below 500 copies/mL increased from about 40% before '97 to 65% after
97 and patients with primary resistance decreased from 17.1 to 1.9% for NRTIs
(p=.04) and from 9.8 to 1.9 for PIs over time. The SCs with non-B clade HIV-1
increased from to 2.4 to 11.1% over time. HIV-1 variants with resistance to ARV

were more frequent in males compared to females (82.4 vs 17.6%) and in IDU
and HO compared to HE (23.9 and 12.9 vs 10%).
Conclusions: These data show no tendency for an increase of primary resistance mts from '96 to'00 in Italy. Interestingly, a concomitant increase of the
proportion of patients with suppressed viremia on therapy was observed in this
popoulation of SCs. These data are consistent with recent models suggesting
that the two phenomenon may be causally linked.

Presenting author: Michela Violin, Institute of Infectious and Tropical Diseases,
University of Milan, Via GB Grassi 74, 20157 Milan, Italy, E-mail: michela.violin
@unimi.it
ThPeB7218 Comparison of absolute CD4 and CD8 T-cell
counts, multi-platform versus single-platform
methods
G.D. Cross, J.H. Handsfield. Centers for Disease Control and Prevention,
Atlanta, GA, United States
Background: The CDC Model Performance Evaluation Program for T-lymphocyte
immunophenotyping (TLI) was initiated to evaluate the performance of labs conducting CD4+ T-cell testing related to HIV infection. Traditionally, absolute CD4+
T-cell counts have been derived using multi-platform methods (MPM) [hematology result coupled with flow cytometry derived CD4 percent]. Recently, singleplatform methods (SPM) [single instrument or test kit] have emerged. Development and use of these SPM have been particularly important for developing countries which have limited resources. The CDC MPEP for TLI has proven to be an
excellent medium to compare these two methods of CD4 T-cell enumeration in a
performance evaluation setting.
Methods: Utilizing results returned by MPEP TLI participant laboratories over
6 shipment periods, mean absolute CD4 and CD8 T-cell counts obtained using
MPM were compared to the mean absolute CD4 and CD8 T-cell counts obtained
using SPM.
Results: Absolute CD4 and CD8 T-cell counts obtained using MPM were generally greater than these counts obtained using SPM. Variability of absolute CD4
and CD8 T-cell counts obtained across laboratories was greater for results obtained using MPM methods than results obtained using SPM. The distribution of
MPM has a long tail and the association between MPM and the likelihood of having a high value is very strong. MPM are more at risk for producing a falsely high
CD4 result. The odds ratio for MPM versus SPM adjusting for the association
with shipment indicates that MPM are slightly over four times as likely to return a
falsely high result.
Conclusions: Single-platform methods for absolute CD4 T-cell enumeration have
greater precision than multi-platform methods. Clinicians should note the dissimilarity in the results between the methods used for CD4 T-cell count and the greater
precision of CD4 T-cell enumeration afforded by single-platform methods.
Presenting author: G. David Cross, Centers for Disease Control and Prevention, 4770 Buford Highway, N.E., Mailstop G-23, Atlanta, GA 30341-3717, United
States, Tel.: [phone], Fax: [phone], E-mail: gcross@cdc.
gov
ThPeB72199 Pediatric HIV-1 in Kenya: Predictors of early
mortality
E. Obimbo1, D.A. Mbori-Ngacha 1, P.A. Otieno 2 R.K. Bosire1, J.O. Ochieng1,
C. Farquhar3, J.K. Kreiss3, G.C. John-Stewart 3. University of Nairobi, Nairobi,
Kenya; 2Kenya Medical Research Institute, Nairobi, Kenya; 3University of
Washington, Seattle, United States
Background: A bimodal clinical course has been observed among perinatally
HIV-1 infected children, with rapid disease progression (RP) and death occurring
in 10%-45% of children by the age of 2 years, and the remaining progressing
more slowly. Determination of predictors of RP in children is important for defining guidelines for their care, and interventions to delay disease progression. A
prospective observational study was undertaken to determine maternal and infant correlates of mortality during the first two years of life in a cohort of HIV-1
infected African infants.
Methods: Infants identified to be HIV-1 infected by DNA PCR in a perinatal cohort
were followed monthly to 1 year, then quarterly to 2 years or death.
Results: Fifty HIV-1 infected infants were identified, and followed for a median
duration of 7 months (range 1-22). The median birth weight was 3.0 kg (range
2.0-4.0 kg), with a male:female ratio of 1.1:1, and a breastfeeding rate of 80%
(40/50). Seventeen infants (34%) died during follow up; median age at death was
4 months (range 0.01-18 months). In univariate analysis, increased risk of infant
mortality was associated with low maternal CD4 count [RR 6.7 (95% CI 1.7-26.3)
p=.007], low birth weight [RR 4.3 (1.4-13.9) p=.01], length < 5th centile [3.8 (1.2 -5.2) p=.02] and head circumference < 5th centile [3.6 (1.0-12.5) p=.05]. Low infant CD4 count at age 6 months (p=.09) showed a trend towards increased risk for
mortality. In multivariate analysis maternal CD4 count < 200 was independently
associated with infant mortality[RR 8.6 (95% CI 1.9-38.5)p=.005]. Conclusion:
Low maternal CD4 count, low birth weight and growth faltering are predictors of
early mortality, and may be used to identify HIV-1 infected infants requiring early
treatment in resource poor settings.
Presenting author: Elizabeth Obimbo, PO. Box 19676, Nairobi, Kenya, Kenya,
Tel.: [phone], Fax: [phone], E-mail: [email protected]