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

XIV International AIDS Conference

Abstracts WePpD2105-WePpD2107 153

WePpD2104 HIV risk behavior change over 24 months of
follow-up in a phase-Ill HIV vaccine efficacy trial
B.N. Bartholow1, S. Buchbinder2, V. Golil, C. Celum3, B. Koblin4, M. Marmor5,
P. Cronin6, V. Popovic6. 1Centers for Disease Control and Prevention, CDC,
MS-e45, 1600 Clifton RD., Atlanta, GA., United States; 2San Francisco
Department of Public Health, San Francisco, United States; 3University of
Washington, Seattle, United States; 4New York Blood Center, New York, United
States; 5New York University, New York, United States; 6VaxGen, Brisbane,
United States
Background: Increased risk behavior associated with efficacy trial participation
has been a concern related to the design, safety, and ethical conduct of trials. To
address these concerns, longitudinal risk behavior was evaluated in the context
of a phase III HIV vaccine efficacy trial.
Method: 5,109 HIV-negative MSM and 309 women were enrolled into a placebocontrolled efficacy trial of AIDSVAX B/B from 61 sites, primarily in North America. Sexual risk behavior data were collected at baseline (BL), 6-, 12-, 18-, and
24-mos. Participants' perception of vaccine efficacy was assessed at BL and perceived assignment to vaccine or placebo at 12- and 24-mos.
Prevalence of sex risk behavior in prior 6 mos BL 6m 12m 18m 24m
MSM: unprotected anal sex (UA) 58% 49% 50% 51% 52%
Women: unprotected vaginal sex 55% 44% 41% 45% 52%
Results: For MSM and women, risk behavior decreased from BL to 6-mos
(p<0.001), then increased from 6- to 24-months (p <.001 and p <.04, respectively). At 24-mos UA remained below BL among MSM (p <.001), while UV returned to BL among women (p =.31). MSM perceiving assignment to vaccine at
12-/24-mos (13%) were consistently more likely to report UA at each visit than
MSM perceiving assignment to placebo or who were unsure of their assignment
(OR=1.30 95% CI=1.13-1.50, p < 0.001), and were more likely to report UA at 24 -mos than 6-mos (59% vs. 53%, p < 0.001). Perceived assignment was unrelated
to UV sex for women.
Conclusions: Risk behavior remained at or below baseline levels at 24-mos.
Counseling for trial participants should consistently address assignment perceptions to guard against increases in risk.
Presenting author: Bradford Bartholow, CDC, MS-e45, 1600 Clifton RD., Atlanta, GA., United States, E-mail: bnbl @cdc.gov
WePpD2105 Post-exposure prophylaxis among HIV-uninfected
participants in a phase III HIV vaccine efficacy trial
M.L. Ackers1, S. Buchbinder2, D. McKirnan3, K.H. Mayer4, R. Novak',
C. Harro6, V. Popovic7, W. Heyward7, M. Para8, J. Fuchs2, S. Tabet9,
Vax004-Vision Study Team0. 1Centers for Disease Control and Prevention,
1600 Clifton Rd., DHAP, MS E-45, Atlanta, GA 30333, United States; 2San
Francisco Dept. of Public Health, San Francisco, United States; 3Howard Brown
Health Center, Chicago, United States; 4Fenway Community Health, Boston,
United States; 5Univ of Illinois at Chicago, Chicago, United States; 6John
Hopkins University, Baltimore, United States; VaxGen, Inc., Brisbane, United
States; 8Ohio State University, Columbus, United States; 9Univ. of Washington,
Seattle, United States; 10, United States
Background: HIV-uninfected participants in the AIDSVAX~ B/B phase III HIV
vaccine efficacy trial may use antiretrovirals obtained from other sources for postexposure prophylaxis (PEP). The frequency of PEP use and its impact on HIV
incidence and risk behavior within an efficacy trial are unknown.
Methods: Men who have sex with men (94%) and high-risk heterosexual women
(6%) were enrolled at 61 clinical sites in North America and Europe into the
36-month vaccine efficacy trial. All participants provided concomitant medication
data, including PEP use, at 6-month intervals. Supplemental data on PEP knowledge, perceptions, and use were obtained from trial participants at six sites.
Results: Overall, 108 (2%) of 5418 participants reported PEP use. These 108
participants were from 36 trial sites in 27 US cities, Puerto Rico, 3 Canadian cities,
and the Netherlands. In multivariate analysis, significant predictors (p< 0.05) of
PEP use included California site enrollment (n=7 sites) (Odds ratio [OR]=3.2), age
< 32 years (OR=2.4), having a known HIV-positive sex partner (OR=2.0), higher
education level (OR=1.4) and greater recreational drug use (OR=1.2). Analysis
of the supplemental PEP data from the six sites revealed 260 (46%) of 560 participants had heard of PEP, and trial PEP-users were more likely to believe PEP
prevents HIV infection than were non-PEP users (81% vs. 59%; p=0.05).
Conclusions: Trial PEP use was limited and does not appear to have an impact
on the conduct of this vaccine efficacy trial. As knowledge of and access to PEP
expands, its use among trial participants may increase and affect HIV incidence
and risk behavior in future HIV vaccine efficacy trials.
Presenting author: Marta-Louise Ackers, 1600 Clifton Rd., DHAP, MS E-45, Atlanta, GA 30333, United States, Tel.: [phone], Fax: [phone],

E-mail: [email protected]

WePpD2106 HIV testing outside of protocol during an HIV
vaccine efficacy trial: Comparison of direct
interview and computer assisted self-interview
(CASI) assessments
M.F. Para1, J. Neidig1, B. Bartholow2, M. Durham2, D. Brashers3, M. Longhi4
'Ohio State University, 456 West 10Oth Avenue, Room 4725, Columbus, OH
43210, United States; 2CDC, Atlanta, GA, United States; 3University of Illinois,
Champaign, IL, United States; 4 VaxGen, Brisbane, CA, United States
Background: In Phase III HIV vaccine efficacy trials, placebo and vaccine arms
must have comparable HIV risks. If volunteers discover their randomization assignment (unblinding) via HIV testing outside of the trial (OT) they may differentially change risks and alter study results.
Methods: 5418 Volunteers at-risk for HIV infection were enrolled in a 36-month
placebo controlled HIV vaccine (AIDSVAX~ B/B gpl20) efficacy trial at 61 sites in
N. America and Europe. All HIV testing is to be performed at trial sites to minimize
risk of unblinding. Volunteers were interviewed at the 12 and 24-month visit about
OT during the past year. At six CDC-sponsored VISION sites, study participants
were also asked the same questions 2 weeks later using CASI methodology
Results: Overall 14% of volunteers with data at 12 and 24 months reported OT in
the prior year during either interview, including 492 of 3,753 (13%) volunteers at
non-Vision sites and 105 of 658 (16%) at VISION sites (p=0.05). Among a subset
of VISION site volunteers with 12 and 24-month assessments of OT by interview
plus CASI, 78 of 359 (22%) reported OT during interview or CASI. Of these 36
(46%) acknowledged OT with both methods, 20 (26%) only during direct interview
and 22 (28%) only by CASI. The most common reasons for OT testing listed on
CASI included high-risk exposure (n=12), drug program (n=10), travel (n=1 1) and
medical (n=9). None of these subjects acknowledged unblinding as a reason for
OT.
Conclusions: In this HIV vaccine efficacy trial, OT was acknowledged by 14%
of volunteers over two years. At VISION sites, CASI elicited additional positive
OT responses after interviews. Most reasons for OT were appropriate and none
apparently resulted in unblinding. Although current EIA assays are unlikely to unblind a gpl20 vaccine recipient, more complex HIV vaccines may be at greater
risk for unblinding on routine testing.
Presenting author: Michael Para, 456 West 10th Avenue, Room 4725, Columbus, OH 43210, United States, Tel.: [phone], Fax: [phone],
E-mail: para.1 @osu.edu
WePpD2107 The private demand for an HIVIAIDS vaccine in
Thailand: Results from a household survey
M. Ainsworth1, C. Suraratdecha2, D. Whittington3, V. Tangcharoensathien4
C. Vasavid5. 1 World Bank, 1818 H St. NW Washington, DC 20433, United
States; 2Sukhothai Thammathirat Open University, Bangkok, Thailand;
3University of North Carolina, Chapel Hill, United States; 4Health Systems
Research Institute, Bangkok, Thailand; 5National Statistical Office, Bangkok,
Thailand
Background: Thailand, with one of the worst AIDS epidemics in Asia, is the
site of the first Phase III AIDS vaccine trial in a developing country. In light of
Thailand's successful policy response that has resulted in widespread behavior
change and a reduction in risk behavior, would there be private demand for a new
vaccine as an alternative preventive method? How would the price and efficacy
of the vaccine affect private demand?
Methods: We conducted a household survey of 2,524 adults aged 18-60 in a
random sample of 1,235 households in 9 of the 76 provinces representing four
regions and Bangkok, and high and low HIV prevalence areas. The households
were drawn from a sample previously interviewed by the 2000 Socioeconomic
Survey conducted by the National Statistical Office. Respondents were randomly
asked whether they would be willing and able to pay for an AIDS vaccine of either
50% or 95% efficacy, that lasts 10 years, is safe, and has no side effects, at prices
ranging from $5 to $1,500.
Results: The private demand for an AIDS vaccine is sensitive to price, efficacy,
household income, age, gender, and risk behavior. At a given price, the demand
for a 95% effective vaccine is greater than for a 50% effective vaccine. Controlling
for price and efficacy, demand rises with income and perceived risk of lifetime exposure to HIV, and is higher among men and those under 35. 78% of respondents
would be willing to be vaccinated if the vaccine were available for free. Willingness
to be vaccinated with a free vaccine is higher for vaccines of higher efficacy, but
has no relation with household income.
Conclusions: The results suggest that in introducing AIDS vaccines, government
can satisfy both equity and efficiency objectives by targeting subsidies to those at
high risk of infection with the lowest income while making the vaccine available at
cost to those who can afford it.
Presenting author: Martha Ainsworth, 1818 H St. NW, Washington, DC 20433,
United States, E-mail: [email protected]