566
Evaluation of the Safety of Carraguard for Vaginal Use by HIV-infected Women: Results from a Phase I, Randomized, Placebo-controlled, 3-Treatment Cross-over Study
Catherine McLean*1, H Jones2, J McNicholl3, J van de Wijgert4, T Medtanavyn5, W Uthaiworavit6, S Whitehead3, J Karon7, P Kilmarx1, and L Markowitz1
1CDC, Atlanta, GA, US; 2Population Council, New York, NY, US; 3Thailand Ministry of Publ Hlth-US CDC Collaboration, Nonthaburi; 4Academic Med Ctr, Univ of Amsterdam, The Netherlands; 5Chiang Rai Provincial Hlth Office, Thailand; 6Chiang Rai Hosp, Thailand; and 7Emergint Corp, Louisville, KY, US
Background: Candidate microbicide safety, including effect
on cervicovaginal, cell-free HIV, needs evaluation in HIV-infected women. We
conducted a randomized, controlled, crossover study of Carraguard (Population
Council, New York City) in HIV-infected women.
Methods: Eligibility required CD4+ count
of 50 to 500 cells/mm3 and not currently using antiretrovirals. Women
were randomized to 6, 1-month-treatment (Carraguard, methylcellulose gel,
no-product) sequences. During product-months, women inserted gel vaginally once
daily for 7 days. At baseline (pre-gel), 15 minutes, day 7, and day 14,
incident genital epithelial findings were documented and cervicovaginal lavages
for cell-free HIV-1 RNA (Roche Amplicor HIV-1 Monitor 1.5) were collected. cervicovaginal
lavages with <40 and <80 copies/mL were defined as non-detectable and
non-quantifiable, respectively; Log10 of total viral copies was
calculated. Logistic regression models were used to test for variations in
outcomes among the 3 treatment arms.
Results: We enrolled 60 women: median age, 34
years; CD4+ counts, 296 cells/mm3; plasma HIV viral
loads, 4.6 log10 copies/mL; cervicovaginal lavage viral loads, 3.1
total copies; and similar sexual behaviors among randomization groups. Overall,
594 of 600 scheduled study visits were completed. Incident epithelial findings
at any location and all visits were similar across treatment arms (range 10 to 18%,
p = 0.6). Median cervicovaginal lavage viral loads were lower with
Carraguard and placebo gel than with no-product at 15 minutes (non-detectable,
non-detectable, and 2.8 log10 viral copies, respectively; p <0.001)
and at day 7 (non-quantifiable, non-quantifiable, and 2.7 log10
viral copies, respectively; p <0.001) but were similar at day 14 (2.9
log10, non-quantifiable, and 2.8 log10 viral copies,
respectively; p = 0.22).
Conclusions: Cervicovaginal cell-free HIV decreased
in Carraguard and methylcellulose gel-use arms after 15 minutes and 7 days of
exposure. While gel effects on HIV detection must be excluded, our findings
suggest that Carraguard is safe for vaginal use in HIV-infected women; gel use
may decrease HIV shedding and transmission; however, the clinical correlation
is unknown.
|
