Abstracts Vol. 1 [International Conference on AIDS (11th: 1996: Vancouver, Canada)]

Monday, July 8, 1996

Mo.C.325 - Mo.C.333

rate of CD4 cell loss for men with and without symptoms during SC, simultaneously controlling for confounders including age and treatments as time dependfent cowari tes. Using logistic regression, we also attempted to predict the occurrence of synptomatic pri ary infection considering potential demographic, clinical and behavioral factors prior to SC.
Results: Of 328 men who were followed a median of 8 years, 1I4 (3/.8") developed
AIDS and 91(27.7%) died.The incidence rates of feverwere 45 ntmes hihet during SC
than during the prior interval (p <.00 I) and 3.4 times higher than attit. (p. 001).
More than seven days of fever during SC was reported by 13.8% aw,] was associated with
faster progression to AIDS (relative hazard (RH): 2.6; 95% confidenc interval I).6 -
4.2), to death (RH: 2. I; 95% Cl: 1.3 - 3.6) and faster rate of CD4 c 1I-ns- (p.0I6). None
of the other symptoms were associated with progression to AIDS aind Iatfi. although the
incidence rates during SC were all increased compared to prior and subsequent intervals.
Fever, nor any of the other symptoms during SC could be predicted by age, # CD4 cells
prior to SC, history of sexually transmitted diseases, sexual behavior and use of alcohol,
tobacco and recreational drugs.
Conclusions: A substantial proportion of seroconverting homosexual men experience prolonged periods of fever. Fever, alone, during SC predisposes to an increased risk to AIDS
and death, and for this group, early treatment should be considered. Whether fever during
SC is a marker, or an inductor of faster disease progression, -nd whether sippression of
fever may have a beneficial effect on the course of HIV- I infection requires further study
PJ.Veugelers, Municipal Health Service, Nieuwe Achtergracht 100, 10 8 \AT Amsterdam,
Netherlands.Telephone: [phone] Fax: [phone] E ci: paulus(tsara.nl
Mo.C.325
PREDICTORS OF DISEASE PROGRESSION AND SURVIVAL AMONG AZT-TREATED
INDIVIDUALS FOLLOWED AFTER HIV SEROCONVERSION.
Rezza G*, Galai N**, Pezzotti P*,Vlahov D**, Graham NMH**. the HIV-ltalia
Seroconversion Study group. *lstituto Superiore di Sanita, Roma, Italy and ** School of
Public Health, Johns Hopkins University Baltimore, USA.
Objective: The aim of the study was to evaluate indicators of respise to antiretroviral
treatment with zidovudine (AZT) in terms of progression to AIDS and death in a cohort of
HIV-serconverters.
Methods: We selected a treated cohort of individuals from a large group of 1024 seroconverters belonging to three different exposure groups, both genders, and a wide range of
ages at seroconversion.Time zero of the analysis was the date of initiation of AZT therapy
Analyses were performed using two different outcomes: AIDS (I 987 CDC definition), and
death. Response variables were: duration of HIV infection, acute retroviral syndrome, clinical
status, baseline level and changes of CD4 cells count and p24-antigen. Statistical analysis
used standard survival techniques as Kaplan-Meier curves, log-rank test, and Cox proportional hazards regression model.
Results: At the univariate analysis, individuals who demonstrated larger declines in CD4 following initiation of therapy (exceeding 30% in six months) had an increased risk for AIDS
progression (RH=2.89) and mortality (RH=4.16). Presence of p21 Antigen was strongly
associated with worse prognosis with RH=2.58 for AIDS and RH=2.,4 for death.
Individuals who remained p24-Antigene positive after AZT treatment were at higher risk
for AIDS and death relative to those who remained negative (RH=3./9 for AIDS, RH=3.54
for death). Acute retroviral syndrome and duration of HIV infection dii not predict
response to treatment. At the multivariate analysis, there was a clear independent effect of
CD4 at initiation of treatment, percentage of CD4 cell lost in the next six months, symptoms at index visit and presence of p24-Ag at the same visit.Those individuals with less
than 200 CD4 cells at first visit, who lost more than 30% of initial 3CD4 cells, had symptoms,
and p24-Antigen detectable at initiation of therapy were 425 tirmes snore likely to develop
AIDS than those who had all these parameters going in the opposite direction. Using death
as an end-point the results were very similar
Conclusion: CD4 and symptoms at start of therapy CD4 cell decline after initiation of AZT
and p24-antigenemia are early indicators of disease progression in treated patients.The
combined use of these indicators may help to better predict who wil i espond to AZT or
other antiretroviral therapy.
Giovanni Rezza, Istituto Superiore di Sanita,Viale Regina Elena, 299, 00 16 Roma, Italy
tel: (39) (6) 4990 2337; fax: [phone]; e-mail: PATRIZIO(IS. It
Mo.C.330
A BIRTH COHORT ANALYSIS OF 75,097 EUROPEAN AIDS CASES
Houweling, Hans, Heisterkamp SH*, Jager JC*, Coutinho RA**. European Research Team
on AIDS scenarios *National Institute of Public Health and the Environrnent, Bilthoven, the
Netherlands; *Municipal Health Service, Amsterdam, the Netherlands
Objectives: To study AIDS incidence in the European project on Multinational AIDS scenarios (contract BMH I-CT-941723) with particular reference to the youngest groups at risk
through sexual and needle-sharing risk behaviours.
Methods: AIDS incidence data (European AIDS Centre, Paris) for Aus, Bel, Gei Gre, Ita, Ire,
Net, Po Spa and Swi, 1978-1994, were adjusted for reporting delay Yearly AIDS incidence
was analysed by exposure group (homo/bisexual men-tHBM, injecting drug users-IDU, heterosexual contact cases=tHET), gender
and S-year birth cohorts 1925- 1975.
Results: The overall leveling off of AIDS _ i0 Eu,,.. oun8e
incidence among HBM is caused by a. 1955-1954
maturing of the epidemic among men - _born before 1960.The 1960-1964 -.
cohort is the irst to be shifted to the n 97i
right, but incidence reaches levels high 89 87 i, 91,i
er than in the pre-1960 cohorts. Y sorins ae
Comparing the slope of the 1965-1969
curve for cases diagnosed 199 I -1994 with periods t-5 yrs. for siiceis e older cohorts,
there was evidence for an only slightly lower incidence among HBM rho ave become
sexually active in the AIDS era. Among IDU AIDS incidence is still increasing among all birth
cohorts after 1950. A strong increase of AIDS incidence is seen in al 950- 9'! birth
cohorts of heterosexual contact cases, especially among women.

Conclusion: AIDS incidence is increasing in all cohorts of homo/bisexual men born after
1960, and injecting drug users and heterosexual men and women born after 1950.There
was only limited evidence for effective prevention among cohorts who have become sexually active or started injecting in the AIDS era.
H. Houweling, Dept Infect Dis Epidemiol, NaI Inst Public Hlth (RIVM), PO Box I, the
Netherlands, tel. [phone], fax [phone], email ciehharivm.nI
Mo.C.33 I
HIV PREVALENCE AMONG U.S. CHILDBEARING WOMEN, 1989-1994
Davis, Susan Fischer, Steinberg S, Jean-Simon M, Rosen, Gwinn M. Centers for Disease
Control and Prevention, Atlanta, GA, USA.
Objective:To describe trends in HIV prevalence among childbearing women in the
United States.
Methods: In an anonymous, population-based national sernosurvey, 2 million newborn
specirnemis were tested annually for maternal antibody to HIV. Based on data from 35 states
conducting the survey continuously since 1989, national and regional seroprevalence estimates were made by standardizing to the known distribution of births and extrapolating
using the distribution of reported perinatal AIDS cases. Regions were defined according to
the U.S. Bureau of Census.
Results: From 1989 through 1994, estimated HIV prevalence remained relatively stable
nationwide, with 6000 to 7000 HIV-infected women delivering live-born infants annually.
Seroprevalence rates declined in the Northeast, but they continue to be the highest in the
United States. In the South seroprevalence rates increased and then stabilized.
Rate per 1000 live births

Year
1989
1990
1991
1992
1993
1994

Northeast South

4.1
4.0
3.9
3.7
3.4
3.2

1.6
1.8
2.0
2.0
2.0
1.9

Midwest West All U.S.
0.5 0.5 1.6
0.5 0.6 1.6
0.5 0.6.7
0.6 0.6 1.7
0.6 0.5 1.6
0.6 0.6.5

\,O
0)
j
0
no
u
O---
cc
0
0
U
01)
0)
cc
c0
c
Conclusions: Distinct regional HIV seroprevalence trends among U.S. childbearing women
may reflect different phases of the HIV epidemic.The epidemic may have started earlier in
the Northeast than in other regions. Despite observed trends, seroprevalence rates in all
regions of the United States remain high and reflect ongoing HIV transmission and the need
for resourceful HIV prevention efforts.
Susan Fischer Davis, CDC, DHA, MS-E46, Atlanta, GA 30333, USATeI: [phone],
Fax: [phone], email: sxd I@cdchiv I.em.cdc.gov
Mo.C.332
TRENDS IN AIDS INCIDENCE IN THE UNITED STATES, 1990-1994
Ward ohn W, Karon J, Fleming P, Gayle H. Centers for Disease Control and Prevention,
Atianta, GA, United States
Objective: To describe national trends in estimated AIDS incidence rates (A-I) from
1990 through 1994.
Methods: We examined cases of persons > 13 years with AIDS-opportunistic illnesses (Ol)
diagnosed (dx) in 1990-1994 and reported through September 1995. Data were adjusted
for reporting delays, unreported risks, and AIDS-OI dx dates were estimated for cases
reported with on!y CD4+ values.The 1990 and 1992 census were used to obtain A-I rates
per 1 00,000 persons for 1990--1991I and 1992-1994, respectively
Results: From i990 through 1994, the national A-I increased 36% from 22.3 (n-45,000) to
30.4 (n- 63.000); but annual increases in A-I slowed from 15% in 1991 to 6% in 1994. From
1990- 1994, the increase among A-I for women (89%) was three times that among men
(29%) and the ratio of A-I fell firom 7.5:1, (40.5 men:5.4 women) in 1990 to 5. I: I
(52. I:10.2) in 1994. A-I grew 68% among blacks (BI) and 37% among Hispanics (Hp) compared with 14% among whites (Wh); in 1994, A-I among BI (106. I) and Hp (65.4) was 6
and 4 times greater, respectively than among Wh (I 6.5). A-I was highest in BI men; in 1994,
I in 590 BI men had an AIDS-OI dx (A-I, 171). From 1990-1994, the A-I among BI women
doubled, from 25. I to 50.5, and in 1994 was twice that among Hp women (24.5) and 17
times that among Wh women (3.0). By risk, A-I rose 20% among men who have sex with
men (MSM), 42% among injecting drug users (IDU), and 156% among persons with heterosexual risks (hetsex). A-I increased 6% among white MSM. Among BI men, growth in A-I
was similar among MSM (53%) and IDU (54%). Among women, Hetsex A-I grew more
than IDUI A-I (146% vs. 59%) and exceeded the IDU A-I in 1994 (5. I vs. 4.6).
Conclusions: A-I continues to increase in the USA but less so than in earlier years.The
slowing of the national epidemic largely reflects trends among white MSM. A-I is rising
fastest among women, particularly black women, and increasingly represents heterosexual
HIV transmission. Racial and ethnic minorities represent an increasing proportion of
persons with AIDS-Ol.
John W Ward, Centers for Disease Control and Prevention, 1600 Clifton Road, MS-E-47,
Atlanta, CA, 30333,USA.Tel: [phone], Fax: [phone],
e-mail: jww4@(cidhiv I.em.cdc.gov
Mo.C.333
ENHANCED SURVEILLANCE FOR HIV INFECTION IN
BRITISH COLUMBIA, CANADA
Cook DA, Patrick DM, Rekart ML, Middleton PJ, Strathdee SA, Spencer D, Rees T,
MacDougall R,. Marcin K. British Columbia Centre for Disease Control,Vancouver Canada.
Objective: Tor make better use of data from diagnostic HIV testing by eliminating duplicate
tests and collecting complete risk and ethnic information while protecting the confidentiality
ol HIV seropositive patients.
Methods: From J-snuary I to September 30, 1995 868 HIV positive specimens were identified. Physicians were telephoned to verify or complete missing information, to inquire about
previous test results and to request other information such as most recent CD4 count.The
call provided an opportunity to offer resources and referrals for physicians with new HIV