Two months of AIDS reports: critical lessons for HIV biology and medicine.

VACCINES “Translating uncertain data into policy is always a difficult issue” is how Gary Nabel, director of the Vaccine Research Center at the NIH, succinctly characterized the new issues complicating the design of AIDS vaccine candidates.1 Concerns arose following not only failure of the latest product but also its association with an increased risk of HIV acquisition. The STEP study (also known as the HVTN 502 or Merck V520-023 study) was a multicenter, randomized, double-blind, placebo-controlled phase 2b AIDS vaccine trial, which began in December 2004.2,3 The study enrolled 3000 HIV-negative volunteers aged 18 to 45 years at high risk for HIV infection and was conducted in 15 US and 3 Canadian cities as well as study sites in Peru, Brazil, Haiti, the Dominican Republic, Jamaica, and Australia.2 Participants were given 3 doses of vaccine over 6 months and followed up for a mean of 13 months. Most of the men in this study were men who have sex with men.3 They were evenly divided between the experimental and control groups based on circumcision status (about 65%) and prevalence of unprotected receptive or insertive anal intercourse. A few months later in South Africa, another trial—Phambili—enrolled 801 men and women, aged 18 to 35 years, most of whom were heterosexual.3 The vaccine used in both trials was a replication-defective adenovirus type 5 (Ad5) vector including, in a 1:1:1 admixture, 3 Ad5 viruses, each containing codon-optimized, near consensus transgenes for 1 of 3 HIV-1 clade B gene products, Gag, Pol, or Nef.4 The placebo was the vaccine dilution buffer. The STEP Study Oversight Committee halted all trials of the vaccine on September 21, 2007, after it became clear that it failed to protect against infection and may have even promoted it.1 Among the 741 volunteers in the V520 trial who received at least 1 dose of the 3-dose vaccine series, 24 cases of HIV infection were observed versus 21 cases in the 762 volunteers who received placebo.2 In the subgroup, those who had received at least 2 vaccinations and who were HIV-negative for at least the first 12 weeks of the trial, 19 infections were seen in the 672 who received vaccine versus 11 among 691 who received placebo. In addition, the vaccine did not reduce the viral load in those who became infected. The mean plasma HIV RNA levels 8 to 12 weeks after infection were 40,000 copies/mL in the vaccine group and 37,000 copies/mL in the placebo group.2 Of additional interest, even though about 40% of the trial participants were female, only 1 woman developed HIV infection.5 These data present some extraordinary challenges, none with particularly salutary consequences. First, there is the potential for heightened difficulty in recruitment of future volunteers, related not only to the unexpected outcome of this trial but also to the adverse outcomes associated with another biomedical means of HIV prevention, microbicides. Trials of 2 vaginal microbicide gels, nonoxynol-9 and cellulose sulphate, led to more infections among those using the products than among those receiving placebo. “This is my worst nightmare,” said Glenda Gray, the principal investigator for the Phambili trial.3 “I haven't slept for days. I have a headache. I'm ready to resign from trials for the rest of my life.” Second, there is the issue as to whether ongoing trials of other viral-based vectors used to augment immunity to HIV gene products might carry a similar risk. It is possible that vaccination in those with preexisting immunity may have triggered production of activated CD4+ T cells, the preferred host