The need to develop additional HIV prevention tools, especially those that women can use, is urgent . Microbicides are being developed for topical application inside the vagina or rectum to prevent infection with HIV and possibly other sexually transmitted infections. Women are often limited in their ability to abstain from sex or convince their male partners to adopt safer sex behaviours owing to social, cultural and economic gender inequalities. Furthermore, the importance of having children is a major obstacle to condom use for many women and couples, and noncontraceptive microbicides would give them an option to protect themselves from HIV while trying to conceive.
In January 2007, the Contraceptive Research and Development Program (CONRAD) and Family Health International announced that they had halted two phase 3 clinical trials of the candidate microbicide cellulose sulfate gel (Ushercell, Polydex, Canada) . The independent Data Monitoring Committee for the CONRAD trial in South Africa, Benin, Uganda and India had detected a trend towards increased HIV risk associated with cellulose sulfate use, whereas that for the Family Health International trial in Nigeria had not detected any increased risk. At the time of the closures, cellulose sulfate was one of four vaginal microbicide candidates in phase 2B/3 efficacy trials for HIV prevention (Table 1) . The other three candidates (Carraguard gel, PRO-2000 gel and BufferGel) are currently still in clinical development.
The news about the premature closure of the cellulose sulfate trials was disappointing and unexpected. Cellulose sulfate, a polyanion, showed good activity in preclinical studies against HIV, herpes simplex virus type 1 and 2, human papillomavirus, Neisseria gonorrhoeae, Chlamydia trachomatis and Gardnerella vaginalis, and no activity against lactobacilli [4,5]. Furthermore, it was found to be safe in several safety and contraceptive efficacy trials involving 518 women (including HIV-positive women) and 48 men in the treatment arms that were conducted before the phase 3 trials were initiated [6–10]. Most studies used the sexual lubricant KY jelly (Personal Products Co, Skillman, New Jersey, USA) as the placebo gel, with gel used one to four times daily for 6 days to 6 months. No or minimal differences between the cellulose sulfate group and the placebo jelly group were found in self-reported symptoms, genital epithelial findings assessed by colposcopy, vaginal infections, proinflammatory cytokines in genital fluid and systemic safety laboratory findings. One study, however, found changes in the vaginal flora, with an increase in Escherichia coli in the cellulose sulfate group and decreases in hydrogen peroxide-producing lactobacilli in both groups .
The phase 3 trial of cellulose sulfate for HIV prevention was placebo-controlled, randomized and blinded . The targeted sample size was 2574 HIV-negative women based on an assumed 50% effectiveness of cellulose sulfate and 4% annual HIV incidence in the control arm. A total of 66 HIV endpoints were expected and clear stopping rules were defined. When the Data Monitoring Committee met, 1333 women had been enrolled and 35 seroconversions had taken place at the African sites, where the annual HIV incidence was higher than 4%. As many before us have pointed out, the trial investigators and teams deserve praise for immediately closing the trials, recalling study gels as fast as possible, ensuring that women who tested HIV positive in the trials received good HIV-related care and transferring participants from trial services to adequate services in the community.
Cellulose sulfate is not the first product to fail in the microbicides field; nonoxynol-9 (N-9), a nonionic surfactant, preceded it. When N-9 products were first tested for anti-HIV activity in clinical trials, they had already been used for over 25 years in spermicides and sexual lubricants (in concentrations ranging from 2 to 12%), they were cheap, and data on anti-HIV activity from in-vitro and animal models were promising [11,12]. In the 1990s, safety studies of different formulations (such as film, gel, and suppository) of N-9 were conducted [11–14]. Low-dose N-9 was considered safe and two phase 3 trials were launched. A phase 3 trial of 70 mg N-9 film showed no effect against HIV . A phase 3 trial of 52.5 mg N-9 gel showed no effect against HIV at low frequency of use and an increased risk of HIV when used more than three to four times per day . In-vitro studies that were conducted after these phase 3 trials had been completed confirmed that the therapeutic window of N-9 is very narrow (N-9 displays anti-HIV activity at doses that are cytotoxic for epithelial cells and lymphocytes); that N-9 cytotoxicity varies by concentration, duration of exposure and repeated exposure; and that N-9 use is associated with vaginal colonization of Escherichia coli, Enterococcus and anaerobic Gram-negative rods [11,17]. Additional studies showed that N-9 induced an inflammatory response characterized by increases in levels of cytokines, chemokines and other inflammatory mediators; recruitment of neutrophils and monocytes into the genital tract; and activation of HIV transcription in infected cells by upregulation of transcription factors NF-κB and activator protein-1 [11,18,19]. In the current scientific and regulatory environment, it is unlikely that N-9 would have progressed further than preclinical evaluation.
C31G or Savvy
Another surfactant, C31G or Savvy, is also no longer being developed as a vaginal microbicide. Two phase 3 trials of Savvy, in Ghana and Nigeria, were discontinued in 2005/2006 because HIV incidence at the sites was too low.
Potential explanations for trial results in the direction of harm
Microbicide trial results in the direction of harm can be explained in multiple ways. The most obvious explanation is that the candidate microbicide increases women's vulnerability to HIV. However, results in the direction of harm could also be caused by differences between study arms in nonadherence with study products and/or condom use, particularly if the candidate microbicide is coitally dependent and less efficacious than condoms. For example, if a candidate microbicide gel is more acceptable to women than the placebo gel, gel use may be higher and condom use lower in the microbicide arm compared with the placebo arm. Finally, the placebo could be more protective than the candidate microbicide or trial results could be explained by statistical chance. The last two explanations are unlikely and the remainder of this review will, therefore, focus on the first two.
Increased vulnerability: are pieces of the safety puzzle missing?
After vaginal exposure, HIV is capable of establishing infection through multiple pathways involving a variety of target cells and receptors . The relative importance of each pathway is not yet clear. The virus can cross the epithelium by infecting epithelial cells, by transcytosis through epithelial cells, by epithelial transmigration of infected donor cells, via uptake by Langerhans cells, and by direct entry through epithelial disruptions. Crossing the epithelium is likely easier when there are fewer layers of epithelial cells, such as the one layer of columnar epithelium in the cervix as opposed to the multilayer squamous epithelium in the vagina . In the submucosa, the virus can infect CD4 T cells, dendritic cells and macrophages. It can use the gp120 receptor in conjunction with the CXCR4 or CCR5 coreceptor, but it can also attach to mannose-binding C-type lectins such as DC-SIGN. The ideal microbicide should be able to cope with this variety of HIV infection mechanisms but should also be able to block HIV replication and release once HIV is intracellular; this may require a combination of several active ingredients in one formulation .
Microbicide-induced enhancement of HIV entry
Cellulose sulfate is a polyanion. Polyanion acitivity was first identified in experiments using CXCR4 viruses, in which they were able to bind to positively charged regions of gp120, thereby inactivating cell-free virus [4,22,23]. However, recent studies showed that most sexually transmitted HIV-1 strains use the CCR5 coreceptor for transmission, and these CCR5 viruses have significantly lower levels of exposed positive charges on their gp120 . The current understanding is that polyanions are most effective against CCR5 viruses following conformational change induced by CD4 binding, which exposes the basic charge regions in gp120 (Fig. 1). Therefore, to be effective against CCR5 transmission, polyanions must be present at the site of virus–target cell interaction within mucosal tissue. Even then, such protection could be circumvented by the rapid transport of virus from mucosal sites by migratory dendritic cells . The degree of tissue penetration required to block viral attachment and fusion is not known. However, current polyanion candidates were selected on the basis of undetectable mucosal absorption to minimize systemic side effects, with cellulose sulfate (2000 kDa) being the largest in its class.
While these new insights may explain why cellulose sulfate lacked efficacy, they do not explain a potential increased susceptibility to infection. One of the ways in which a candidate microbicide could increase a women's vulnerability to HIV is by increasing HIV binding to target cells, increasing HIV entry in permissive cells and/or facilitating cell-associated transepithelial infection. To rule out these possibilities, the US Food and Drug Administration (FDA) recommends cell-based antiviral activity and cytotoxicity assays using a variety of cell lines (laboratory cell lines, peripheral blood mononuclear cells, primary macrophages and dendritic cells) and different laboratory and clinical virus isolates (including both cell-associated and cell-free systems, and CXCR4 and CCR5 viruses) under conditions likely to be encountered in real life (e.g., in the presence of semen and vaginal secretions) . In the case of cellulose sulfate, the FDA-recommended assays did not give cause for concern, but CONRAD is considering additional nonclinical studies to further investigate potential mechanisms of harm .
Ex-vivo human explant studies and in-vivo animal models are increasingly being used to compare the efficacy and safety of different candidate microbicides, and new in-vitro, ex-vivo and in-vivo models continue to be developed . For example, recent human cervical explant studies suggest that the smaller polyanion candidate PRO-2000 (5 kDa) can inhibit both localized infection and dissemination pathways . The most relevant model of human efficacy is likely to be the macaque model of vaginal CCR5 transmission. Recently, two compounds (PSC-RANTES and CMPD-167) have demonstrated good protection against CCR5-utilizing SHIV162P3 in this model [27,28]. Some argue that all candidate microbicides that are currently in phase 3 trials should first have been evaluated in this model . The problem with newer models is that their utility and reproducibility in different hands and under different study conditions have not yet been fully explored, and results have not yet been correlated with clinical trial results. The experiences with N-9 and cellulose sulfate may provide an opportunity to validate these models further.
Microbicide-induced changes in genital epithelial integrity and/or permeability
Repeated exposure of the cervicovaginal mucosa to a candidate microbicide may lead to changes in epithelial integrity and/or permeability, which could facilitate pathogen transmission instead of preventing it. In the microbicides field, in-vivo safety studies typically include rabbit and/or primate vaginal irritation models, employing histology, colposcopy and inflammatory cytokine assays to detect local toxic effects . More recently, the rabbit vaginal irritation model has been refined , and in-vivo models have been expanded to correlate toxic effects with increased susceptibility to infection with a variety of sexually transmitted pathogens [11,19,31].
Genital epithelial findings and clinical signs of inflammation have always been important safety endpoints in clinical trials of candidate microbicides . Attempts have been made to standardize the diagnosis and documentation of genital findings across trials, trial sites and trial clinicians [33,34]. In the case of cellulose sulfate, no evidence was found for visible genital irritation in any of the safety trials. However, microtrauma that is invisible to the naked eye was not evaluated in these trials but may also play a role in facilitating HIV transmission, but this was not evaluated in these trials .
Studies suggest that 12–45% of women not using any experimental vaginal products have visible genital epithelial findings upon pelvic examination . Sexual intercourse and several factors that are common in women of reproductive age (such as increased parity, tampon use, traditional vaginal product use and vaginal infections) are correlated with such findings. Given these high background rates, the contribution of experimental vaginal products to the prevalence of genital epithelial findings is difficult to assess. Furthermore, placebo products (especially gels) may also cause genital epithelial findings. The clinical significance of many of these findings, including their potential significance in heterosexual HIV transmission, remains unknown.
For all of these reasons, validated biomarkers of genital irritation could be useful. For example, microtrauma might be detected by measuring haemoglobin and red blood cells in cervicovaginal lavage specimens . Epithelial breaches of sufficient size to allow red blood cells out would be of sufficient size to allow the entry of HIV and possibly even HIV-infected cells. Furthermore, the upper genital tract cannot be assessed by colposcopy. These new techniques, however, are currently limited by the absence of clear normative values, intersubject variability and lack of standardization. Ultimately, studies are needed to determine the relationships between visual genital findings and these types of biomarker in the absence of experimental product use, as well as their relationships with HIV acquisition.
Experiments and studies with N-9 have shown that timing of safety measurements is important: for example, rectal application of N-9 leads to rapid exfoliation of sheets of epithelial cells (within 15 min of application) and regeneration of the epithelium takes place within hours . Susceptibility to infection is increased immediately after the rapid exfoliation but also during the regeneration of the epithelium. Nonclinical safety experiments, and possibly also early safety trials in humans, should, therefore, include acute as well as chronic safety measurements.
Microbicide-induced genital inflammation
Repeated exposure of the cervicovaginal mucosa to a candidate microbicide may also lead to a subclinical inflammatory reaction, increased immune activation and/or decreased innate immunity . Epithelial, dendritic and immune cells at mucosal surfaces produce chemokines and cytokines, which activate immune cells and recruit bactericidal and virucidal agents (such as defensins, other antimicrobial peptides and enzymes) . Activation of immune cells could be counterproductive, as CD4 cells are also target cells for HIV-1. The balance between proinflammatory and anti-inflammatory mediators determines whether inflammation occurs. Cytokine measurements were only recently introduced in microbicide trials but correlate poorly with clinical signs and self-reported symptoms of genital inflammation [11,38]. They were included in one phase 1 trial of cellulose sulfate with no evidence for concern . However, detailed analyses of inflammation and immune activation over and above cytokine measurements are not typically carried out in microbicide trials and were not done in any of the cellulose sulfate trials.
Cytokine concentrations in cervicovaginal lavage samples are difficult to interpret; different types of cytokine are released at different time points, and sometimes only after repeated exposure [38,39]. Furthermore, the cytokine equilibrium at most mucosal surfaces under normal conditions has not been fully characterized, and cytokine levels may be affected by infection, hormonal status, menstrual cycle, semen, use of vaginal products, genetic variability and other factors [38–41]. Interactions between different immune mediators are complex, and the clinical or biological significance of changes in individual or groups of cytokines is not known. It would, therefore, be useful to develop assays that evaluate the cumulative impact of candidate microbicides on mucosal immune function as a whole (as opposed to measuring levels of individual immune mediators) and that distinguish protective and increased-risk responses. A better understanding of mucosal immune responses to candidate microbicides is not only important from a safety perspective but could also be used to optimize immune responses, thereby merging mucosal vaccine and microbicide approaches.
Microbicide-induced vaginal flora changes
The last set of hypotheses relates to the effect of candidate microbicides on the cervical mucus and vaginal microflora. The quantity and composition of cervical mucus is not usually studied in phase 1 safety trials except for differentiating pus from regular mucus. However, the effect of candidate microbicides on lactobacilli is assessed in vitro and in clinical trials (the latter using microscopy). Changes in the vaginal flora have been reported for N-9, cellulose sulfate, and BufferGel [42–44]. Most of these changes were considered beneficial or minor, and the FDA considers some activity against vaginal lactobacilli acceptable if the product possesses sufficient desirable properties . However, the epidemiological evidence for bacterial vaginosis and yeast infections as risk factors for HIV acquisition in women is mounting, and even minor changes of the vaginal flora may matter in settings with high HIV prevalence . It is not clear what levels of vaginal flora changes are acceptable in different settings, and whether more sensitive methods of vaginal flora evaluation should be considered.
Microbicide trial challenges that may explain results in the direction of harm
No validated surrogate endpoints for the biological activity of microbicides currently exist. Therefore, the primary endpoint of microbicide efficacy trials must be HIV incidence . Testing of a vaginal microbicide thus requires the participation of thousands of women at risk of HIV infection through heterosexual sex. The higher the HIV incidence in a trial community, the lower the required sample size. For feasibility reasons, trials are typically conducted in communities that experience a large number of HIV infections each year . For ethical reasons, efficacy trials must measure the incremental effect of the potential microbicide over and above a package of already proven HIV prevention interventions that include HIV counseling and testing, condom promotion and treatment of curable sexually transmitted infections and vaginal infections . Measuring such an incremental effect may be particularly problematic for microbicides that have a lower expected method efficacy than condoms. Another microbicide trial challenge is that a perfect placebo does not exist [48,49]. Even though a placebo typically does not contain the active ingredient of the candidate microbicide, it may nonetheless reduce or enhance HIV transmission. For example, any vaginal gel may prevent trauma through its lubricating properties or it may cause irritation. A placebo gel may also have a different acceptability and adherence profile than the active microbicide owing to small differences in, for example, viscosity. Some have argued for a condoms-only control arm, in which women would not receive a vaginal product but would receive the currently available HIV prevention package in a clinical trial setting . A condoms-only arm cannot be blinded (which could lead to differentials in condom use, enrolment and retention rates, and product sharing) but would allow for comparisons between ‘best case scenario HIV prevention plus microbicide’ and ‘best case scenario HIV prevention without microbicide’. Having both types of control arm in parallel would be ideal but raises concerns about feasibility and cost.
Even in communities with high HIV incidence, one can never be sure which women are exposed to HIV through vaginal sex, through another transmission route, or never at all, without knowing the HIV status of their sex partner(s). While randomization minimizes systematic exposure differences between study groups, transmissions through anal sex or blood transfusions may occur disproportionately. Anal sex in particular is often highly stigmatized and, therefore, difficult to measure. In contrast to vaccines, but like condoms, most (but not all) microbicides would have to be used every time sex takes place, or at regular intervals over time, in order to be protective. Perfect adherence has not been achieved in any long-term microbicide trial to date and high pregnancy rates have resulted in large numbers of women being taken off the product [15,50]. Furthermore, extensive research has shown that there is currently no reliable way to measure sexual behaviour, including microbicide and condom use . Many of these microbicide trial complications are likely to result in underestimates of microbicide efficacy [46,52].
Research is ongoing to try to improve measurement of sexual behaviour and product use. Biomarkers of exposure to semen include the prostate-specific antigen test (multiple manufacturers) and the rapid stain identification of human semen (RSID-Semen; (Independent Forensics, Hillside, Illinois, USA). Unfortunately, prostate-specific antigen can only be detected 24–48 h after semen exposure [53,54]. The International Partnership for Microbicides, in collaboration with Paragon (Invensys Controls, Carol Stream, Illinois, USA), is developing a ‘sexometer’, a vaginal ring containing electronic devices capable of registering when sex takes place . The Population Council developed an applicator-staining test to determine whether an applicator has been exposed to the vagina; it cannot, however, determine whether the applicator was emptied inside the vagina or whether sex took place . When the candidate microbicide is absorbed by the cervicovaginal mucosa (as may be the case with some of the next generation antiretroviral drug-containing microbicides), adherence data may be enhanced by measuring levels of the active ingredient in genital tract tissues. Finally, some of the newer microbicide formulations, such as vaginal rings, no longer have to be inserted every time sex takes place but can be left in place for several weeks or months at a time .
When the cellulose sulfate trial closures were announced, one of the most frequently asked questions by women participating in trials, journalists and others was whether (some of) the 35 women who had become infected during the trial would not have become infected if they had not participated. It should, therefore, be emphasized that all cellulose sulfate trial participants were offered a package of HIV prevention interventions, which probably reduced the total number of new HIV infections during the trial compared with before the trial or in the community outside the trial. Unfortunately, the trial data that have been released to date do not yet provide a definitive answer to this important question.
An in-depth discussion of microbicide trial ethics is beyond the scope of this review and can be found elsewhere . Recently, several trials on the use of tenofovir for preexposure prophylaxis of HIV in Africa and Asia were stopped prematurely because of ethical concerns . The negative consequences included, among others, a delay in the development of new HIV prevention interventions, a waste of scarce resources and decreased trust between research teams and trial communities. It is, therefore, more important now than ever to ensure that microbicide trials are carefully designed and ethically sound.
Future microbicide trial designs may have to be adapted in light of recent prevention successes with male circumcision [60–62]. They may have to be adapted further if recently completed or currently ongoing trials of first-generation microbicides show positive results, or other new HIV prevention strategies (such as preexposure prophylaxis with antiretroviral drugs) are proven to be effective. Whether a new prevention tool become part of the standard prevention package that is offered to all trial participants, or becomes the comparator product in equivalence or superiority trials, will most likely depend on the robustness of the efficacy results, whether the results can be extrapolated to other populations and settings, and on availability. Debates on this topic are to be expected in the near future.
What the field of microbicide development needs most is a variety of microbicide candidates in the pipeline, proof of concept, true placebos and validated safety and surrogate efficacy endpoints. The wealth of cellulose sulfate data now available should be used to move the last forward. For example, it would be useful to determine the efficacy of cellulose sulfate in the CCR5 macaque challenge model to determine if this model would have predicted lack of efficacy in clinical trials . Research on biomarkers of sexual behaviour and adherence, and new user-independent delivery mechanisms, should continue. Microbicide researchers and advocates have done a good job so far in designing ethical trials and addressing ethical issues. However, trial designs and implementation should continue to be debated and adjusted as the field develops and new efficacious HIV prevention methods become available.
While the issue of potential harm may puzzle those working with microbicides for some time, it is now clear that cellulose sulfate was not a potent inhibitor of sexual transmission of HIV-1 infection. This has intensified the ongoing discussion about selection of candidate microbicides to move forward to phase 3 [26,63,64]. A continued pattern of well-publicized trial failures may lead to a loss in confidence in donors, regulatory authorities and potential trial volunteers. The current generation of microbicides included several polyanions. Whether the failure of cellulose sulfate reflects product-specific traits (such as molecular structure) as opposed to class-specific traits can only be determined after completion of ongoing Carraguard and PRO-2000 trials. The next generation of microbicides to be ready for large-scale testing includes at least four gels containing nonnucleoside reverse transcriptase inhibitors. A number of questions arise. Should one or two of these be selected for initial further testing? Should each of these be tested in separate trials right away? Should they be compared head-to-head? And/or, should they be compared in different formulations (e.g., vaginal gel, vaginal ring and oral tablet)? Some argue that microbicide research requires a mechanism to help in making rational choices about the best candidates to move forward [26,63,64]. Others counsel patience, arguing that the development of new products is always an unpredictable process. Most, however, agree that microbicide development should continue.
1. van de Wijgert J, Coggins C. Microbicides to prevent heterosexual transmission of HIV: Ten years down the road. AIDS 2002; 15:23–28.
3. Alliance for Microbicide Development. Microbicide Research and Development Database (MRDD)
. Silverspring, MD: Alliance for Microbicide Development. http://www.microbicide.org.html
[Accessed 21 April 2007].
4. Anderson RA, Feathergill KA, Diao XH, Cooper MD, Kirkpatrick R, Herold BC, et al
. Preclinical evaluation of sodium cellulose sulfate (Ushercell) as a contraceptive antimicrobial agent. J Androl 2002; 23:426–438.
5. Balzarini J, van Damme L. Microbicide drug candidates to prevent HIV infection. Lancet 2007; 369:787–797.
6. Mauck C, Weiner DH, Ballagh S, Creinin M, Archer DF, Schwartz J, et al
. Single and multiple exposure tolerance study of cellulose sulfate gel: a phase I safety and colposcopy study. Contraception 2001; 64:383–391.
7. Malonza IM, Mirembe F, Nakabiito C, Odusoga LO, Osinupebi OA, Hazari K, et al
. Expanded phase I safety and acceptability study of 6% cellulose sulfate vaginal gel. AIDS 2005; 19:2157–2163.
8. El-Sadr WM, Mayer KH, Maslankowski L, Hoesly C, Justman J, Gai F, et al
. Safety and acceptability of cellulose sulfate as a vaginal microbicide in HIV-infected women. AIDS 2006; 20:1109–1116.
9. Schwartz JL, Mauck C, Lai JJ, Creinin MD, Brache V, Ballagh SA, et al
. Fourteen-day safety and acceptability study of 6% cellulose sulfate gel: a randomized double-blind phase I safety study. Contraception 2006; 74:133–140.
10. van Damme L. Update on current microbicide trials: CONRAD's Cellulose Sulfate HIV prevention trial.14th Conference on Retroviruses and Opportunistic Infections
. Los Angeles, February 2007 [Oral presentation on late-breaking trials of new antiretroviral drugs and microbicides]. http://www.retroconference.org/2007/data/files/webpage_for_CROI.htm
[Accessed 23 April 2007].
11. Hillier SL, Moench T, Shattock R, Black R, Reichelderfer P, Veronese F. In vitro and in vivo: the story of nonoxynol-9. J Acquir Immune Defic Syndr 2005; 39:1–8.
12. World Health Organization and Contraceptive Research and Development Program. Technical Consultation on Nonoxynol-9: Summary Report
. Geneva: World Health Organization; 2001.
13. van Damme L, Nirhuthisard S, Atisook R, Boer K, Dally L, Laga M, et al
. Safety evaluation of nonoxynol-9 gel in women at low risk of HIV infection. AIDS 1998; 12:433–437.
14. van Damme L, Chandeying V, Ramjee G, Rees H, Sirivongrangson P, Laga M, et al
. Safety of multiple daily applications of COL-1492, a nonoxynol-9 vaginal gel, among female sex workers. COL-1492 Phase II Study Group. AIDS 2000; 14:85–88.
15. Roddy RE, Zekeng L, Ryan KA, Tamoufe U, Weir SS, Wong EL. A controlled trial of Nonoxynol-9 film to reduce male-to-female transmission of sexually transmitted diseases. New Eng J Med 1998; 339:504–510.
16. van Damme L, Ramjee G, Alary M, Vuylsteke B, Chandeying V, Rees H, for the COL-1492 Study Group. Effectiveness of COL-1492, a nonoxynol-9 vaginal gel, on HIV-1 transmission in female sex workers: a randomised controlled trial. Lancet 2002; 360:971–977.
17. Beer BE, Doncel GF, Krebs FC, Shattock RJ, Fletcher PS, Buckheit RW, et al
. In vitro preclinical testing of nonxynol-9 as potential antihuman immunodeficiency virus microbicide: a retrospective analysis of results from five laboratories. Antimicrob Agents Chemother 2006; 50:713–723.
18. Fichorova RN, Tucker LD, Anderson DJ. The molecular basis of nonoxynol-9-induced vaginal inflammation and its possible relevance to human immunodeficiency virus type 1 transmission. J Infect Dis 2001; 184:418–428.
19. Galen BT, Martin AP, Hazrati E, Garin A, Guzman E, Wilson SS, et al
. A comprehensive murine model to evaluate topical vaginal microbicides: mucosal inflammation and susceptability to genital herpes as surrogate markers of safety. J Infect Dis 2007; 195:1332–1339.
20. Moench TR, Chipato T, Padian NS. Preventing disease by protecting the cervix: the unexplored promise of internal vaginal barrier devices. AIDS 2001; 15:1595–1602.
21. Shattock RJ, Moore JP. Inhibiting sexual transmission of HIV-1 infection. Nat Rev Microbiol 2003; 1:25–34.
22. Moulard M, Lortat-Jacob H, Mondor I, Roca G, Wyatt R, Sodroski J, et al
. Selective interactions of polyanions with basic surfaces on human immunodeficiency virus type 1 gp120. J Virol 2000; 74:1948–1960.
23. Fletcher PS, Wallace GS, Mesquita PM, Shattock RJ. Candidate polyanion microbicides inhibit HIV-1 infection and dissemination pathways in human cervical explants. Retrovirology 2006; 3:46.
24. Margolis L, Shattock R. Selective transmission of CCR5-utilizing HIV-1: the ‘gatekeeper’ problem resolved? Nat Rev Microbiol 2006; 4:312–317.
25. Lard-Whiteford SL, Matecka D, O'Rear JJ, Yuen IS, Litterst C, Reichelderfer P, for the International Working Group on Microbicides. Recommendations for the nonclinical development of topical microbicides for prevention of HIV transmission: an update. J Acquir Immune Defic Syndr 2004; 36:541–552.
26. Check E. Scientists rethink approach to HIV gels. Nature 2007; 446:12.
27. Lederman MM, Veazey RS, Offord R, Mosier DE, Dufour J, Mefford M, et al
. Prevention of vaginal SHIV transmission in rhesus macaques through inhibition of CCR5. Science 2004; 306:485–487.
28. Veazey RS, Klasse PJ, Schader SM, Hu Q, Ketas TJ, Lu M, et al
. Protection of macaques from vaginal SHIV challenge by vaginally delivered inhibitors of virus-cell fusion. Nature 2005; 438:99–102.
29. Shattock RJ, Doms RW. AIDS models: microbicides could learn from vaccines. Nat Med 2002; 8:425.
30. Doncel GF, Chandra N, Fichorova RN. Preclinical assessment of the proinflammatory potential of microbicide candidates. J Acquir Immune Defic Syndr 2004; 37:S174–S180.
31. Cone RA, Hoen T, Wang X, Abussuwa R, Anderson DJ, Moench TR. Vaginal microbicides: detecting toxicities in vivo
that paradoxically increase pathogen transmission. BMC Infect Dis 2006; 6:90–106.
32. Mauck C, Rosenberg Z, van Damme L, for the International Working Group on Microbicides. Recommendations for the clinical development of topical microbicides: an update. AIDS 2001; 15:857–868.
33. World Health Organization. Manual for the Standardization of Colposcopy for the Evaluation of Vaginally Administered Products: An Update
. Geneva: World Health Organization; 2000.
34. Bollen L, Kilmarx PH, Wiwatwongwana P. Photo Atlas for Microbicide Evaluation. Bangkok: MOPH–US Centers for Disease Control and Prevention Collaboration; 2002.
35. Moench T. Sensitive methods to detect epithelial disruptions: tests for microhemorrhage in cervicovaginal lavages. J Acquir Immune Defic Syndr 2004; 37(Suppl 3):S194–S200.
36. Fraser IS, Lahteenmaki P, Elomaa K, Lacarra M, Mishell DR, Alvarez F, et al
. Variations in vaginal epithelial surface appearance determined by colposcopic inspection in healthy, sexually active women. Hum Reprod 1999; 14:1974–1978.
37. Phillips DM, Taylor CL, Zacharoupolos VR, Maguire RA. Nonoxynol-9 causes rapid exfoliation of sheets of rectal epithelium. Contraception 2000; 62:149–154.
38. Fichorova RN. Guiding the vaginal microbicide trials with biomarkers of inflammation. J Acquir Immune Defic Syndr 2004; 37(Suppl 3):S184–S193.
39. Wira CR, Fahey JV, Sentman CL, Piolo PA, Shen L. Innate and adaptive immunity in female genital tract: cellular responses and interactions. Immunol Rev 2005; 206:306–335.
40. Keller MJ, Zerhouni-Layachi B, Cheshenko N, John M, Hogarty K, Kasowitz A, et al
. PRO-2000 gel inhibits HIV-1 and herpes simplex virus infection following vaginal application: a double-blind placebo-controlled trial. J Infect Dis 2006; 193:27–35.
41. Keller MJ, Herold BC. Impact of microbicides and sexually transmitted infections on mucosal immunity in the female genital tract. Am J Reprod Immunol 2006; 56:356–363.
42. Watts DH, Rabe L, Krohn MA, Aura J, Hillier SL. The effects of three nonoxynol-9 preparations on vaginal flora and epithelium. J Infect Dis 1999; 180:426–437.
43. Gupta K, Hillier SL, Hooton TM, Roberts PL, Stamm WE. Effects of contraceptive method on the vaginal microflora: a prospective evaluation. J Infect Dis 2000; 181:595–601.
44. Clarke JG, Peipert JF, Hillier SL, Heber W, Boardman L, Moench TR, et al
. Microflora changes with the use of a vaginal microbicide. Sex Transm Dis 2002; 29:288–293.
45. Myer L, Kuhn L, Stein ZA, Wright TC Jr, Denny L. Intravaginal practices, bacterial vaginosis, and women's susceptibility to HIV infection: epidemiological evidence and biological mechanisms. Lancet Infect Dis 2005; 5:786–794.
46. van de Wijgert J, Jones H. Commentary: challenges in microbicide trial design and implementation. Stud Fam Plann 2006; 37:123–129.
47. World Medical Association. Declaration of Helsinki: ethical principles for medical research involving human subjects
48. Kilmarx P, Paxton L. The world needs a true placebo for vaginal microbicide efficacy trials. Lancet 2003; 361:785–786.
49. Jones H, van de Wijgert J, Kelvin E. The need for a condoms-only control group in microbicide trials. Epidemiology 2003; 14:505–506.
50. Kilmarx P, van de Wijgert J, Chaikummao S, Jones H, Limpakarnjanarat K, Friedland B, et al
. Safety and acceptability of the candidate microbicide Carraguard in Thai women: findings from a phase II clinical trial. J Acquir Immune Defic Syndr 2006; 43:327–334.
51. Hewett PC, Mensch BS, Erulkar AS. Consistency in the reporting of sexual behaviour by adolescent girls in Kenya: a comparison of interviewing methods. Sex Transm Infect 2004; 80(Suppl 2):ii43–ii48.
52. Trussell J, Dominik R. Will microbicide trials yield unbiased estimates of microbicide efficacy? Contraception 2005; 72:408–413.
53. Steiner MJ, Feldbloom PJ, Padian N. Invited commentary: condom effectiveness: will prostate-specific antigen shed new light on this perplexing problem? Am J Epidemiol 2003; 157:298–300.
54. Macaluso M, Blackwell R, Jamieson DJ, Kulczycki A, Chen MP, Akers R, et al
. Efficacy of the male latex condom and of the polyurethane female condom as barriers to semen during intercourse: a randomized clinical trial. Am J Epidemiol 2007; 166:88–96.
55. Rosenberg Z. International Partnership for Microbicides Activities/MDS Priority Actions.Alliance for Microbicide Development Annual Meeting.
Washington, DC, March 2007. http:www.microbicide.org/meetings/meeting10.shtml
[Accessed 16 August 2007].
56. Wallace A, Thorn M, Maguire RA, Sudol KM, Phillips DM. Assay for establishing whether microbicide applicators have been exposed to the vagina. Sex Transm Dis 2004; 31:465–468.
57. Woolfson AD, Malcolm RK, Morrow RJ, Toner CF, McCullagh SD. Intravaginal ring delivery of the reverse transcriptase inhibitor TMC120 as an HIV microbicide. Int J Pharm 2006; 325:82–89.
59. Singh JA, Mills EJ. The abandoned trials of preexposure prophylaxis for HIV: what went wrong? PloS Med 2005; 2:e234.
60. Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, Puren A. Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 Trial. PloS Med 2005; 2:e298.
61. Bailey RC, Moses S, Parker CB, Agot K, Maclean I, Krieger JN, et al
. Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial. Lancet 2007; 369:643–656.
62. Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, Nalugoda F, et al
. Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. Lancet 2007; 369:657–666.
63. Gross M. HIV topical microbicides: steer the ship or run aground. Am J Public Health 2004; 94:1085–1089.
64. Editorial. Trial and failure