1	"LESSONS LEARNED FROM THE PHASE..." LESSONS LEARNED FROM THE PHASE III TRIAL OF COL-1492, A NONOXYNOL-9 VAGINAL GEL, ON HIV-1 TRANSMISSION IN FEMALE SEX WORKERS
2	UNAIDS studies with COL-1492 COL-1492 = Advantage S^® 52.5 mg N-9 bio-adhesive carrier Ù good coverage of vaginal walls and cervix Ù immediate availability of N-9 Placebo = Replens^® vaginal moisturiser more viscous than Advantage S^® lower pH
3	Safety trials with COL-1492 N = 534 COL-1492, placebo and no-treatment arm
4	Phase III trial: HIV incidence*
5	Methodological and Ethical Issues Methodological Issues: -placebo -no treatment arm -behavioural data collection -safety assessment -study population -p-value
6	Placebo “Ideal” placebo: no impact on HIV incidence indistinguishable from experimental product Has been developed and is currently being tested for safety by CONRAD
7	No treatment arm Open-label study prospective bias impact on loss-to-follow-up impact on recruitment potential sharing of products
8	Behavioral Data Collection Necessary ? Intent-to-Treat analysis Exploratory analyses How ? Coital Log Chart Direct interview ACASI
9	Assessment of Safety and Preliminary Effectiveness In COL-1492: risk of HIV when lesion with breach twice the risk without such a lesion Determination of net importance of findings: only in phase III
10	Study Population Generalizability of results Withhold potential beneficial product when trials done among high frequency users? High frequency use assessment: VERY important
11	p-value
12	CONRAD’s Approach 2 arms RCT HEC placebo 80% power .05 significance level 50% effectiveness 1-year retention: 80% 1 year follow-up
13	Informed Consent Process Not a “once and for all” event Questions to assess comprehension after obtaining consent during trial participation Some beliefs extremely difficult to change
14	HIV Treatment STI treatment for all women at screening and follow-up ART? Make a fund available to investigators for seroconvertors during follow-up Prevention of opportunistic infections