Surfactants and detergents
Microbicides were originally intended to have broad-spectrum activity against all STIs and contain non-drug compounds for over the counter use without a prescription (2). The first generation microbicides were non-specific agents comprising surfactants and detergents that directly inactivate viral and bacterial STIs (28). The nonionic surfactant, nonoxynol-9 (an over the counter spermicide) and the detergent, sodium dodecyl sulphate (sodium lauryl sulphate or SAVVY), dissolve the lipid envelope and denature membrane proteins of sexually-transmitted pathogens, respectively. These agents have a low therapeutic window and frequent use of nonoxynol-9 causes vaginal inflammation and disruption of the vagina epithelium, as observed in vitro and in a mouse model (29, 30). Consistent with these findings, a Phase II/III clinical trial in sex workers demonstrated that vaginal application of nonoxynol-9 several times a day increased risk of HIV acquisition compared to women using placebo gel (31). Trials to evaluate the efficacy of SAVVY in preventing HIV transmission were ceased due to futility (32, 33). These agents are no longer considered safe or viable options for microbicides.
Acid buffering agents
A second class of nonspecific first generation microbicides is acid buffering agents. The acid buffering agents BufferGelÒ and ACIDFORM are designed to maintain the low antimicrobial pH of the vagina (pH 3.5 - 4.5) in the presence of semen (which transiently increases the vaginal pH to neutral) thus preventing the window of opportunity for acid labile pathogens such as HIV and other sexually-transmitted pathogens to infect the mucosa (34, 35). BufferGelÒ was found to be safe in a phase IIb clinical trial; however it failed to prevent HIV transmission in women (36). In contrast to BufferGelÒ, which solely relies on protons for lowering the pH, ACIDFORM (now called Amphora®) contains a combination of acids, including lactic acid, which is produced by protective vaginal Lactobacillus spp. Lactic acid at concentrations found in the cervicovaginal fluid from women (37, 38) has orders of magnitude greater HIV virucidal activity compared to protons alone (39). Lactic acid is also active against Chlamydia trachomatis and Neisseria gonorrhoeae in vitro (40). ACIDFORM is currently being progressed as a non-hormonal contraceptive, for the prevention of urogenital chlamydia and gonorrhoea in women, and as an antimicrobial MPT vaginal gel to prevent recurrent bacterial vaginosis (41) (http://www.evofem.com/products/).
Moderately specific microbicides
Members of the moderately specific microbicide class are mainly macromolecular linear anionic polymers such as cellulose sulfate (UsherCell), carrageenan (Carraguard), PRO 2000 and cellulose acetate phthalate (CAP) (28). Linear polyanions generally have broad-spectrum activity against several viral and bacterial pathogens including HIV, HSV, human papillomavirus, Neisseria gonorrheae and Chlamydia trachomatis (28). Linear polyanions block host cell viral entry via electrostatic interactions with the viral envelope proteins and/or host cell receptors (42, 43). Linear polyanions usually lack potency against HIV compared to ARVs and more potent inhibitors of HIV that uses the CXCR4 chemokine receptor compared to CCR5 (44), with the latter HIV strains accounting for most cases of HIV transmission via the mucosa (45).
The performance of linear polyanion-based microbicides has been disappointing in HIV efficacy trials. Carraguard failed to prevent the sexual transmission of HIV in women, while cellulose sulfate showed a trend towards increased HIV acquisition (46, 47). Despite PRO 2000 demonstrating 30% efficacy in a phase IIb trial (36), it failed to prevent HIV transmission in a follow up phase III study (48). The lack of in vivo efficacy demonstrated by linear polyanions against HIV transmission can be ascribed to several factors. These include attenuation of their anti-HIV activity by semen (49) as well as bioavailability in the female genital tract before and after coitus (50), which is compounded by their low potency relative to specific ARVs.
The dendrimer (dendri- = tree, -mer = branching), SPL7013, is a distinct synthetic macromolecular class, which in contrast to linear polyanions, comprises a defined chemical structure and similar potency against CCR5 and CXCR4-using HIV isolates (51, 52). 3% SPL7013 Gel (VivaGel®) has progressed through extensive phase I safety, acceptability and vaginal retention studies (53-60). However, its further development as a vaginal microbicide to prevent HIV infection was put on hold due to a catalytic shift in the microbicide field prompted by the clinical failure of the early microbicide classes (61), although other classes of dendrimers have been pursued in the development of vagina and rectal microbicides (reviewed by das Neves et al 2016) (62). There has been a move away from coitally-dependent gels to sustained release formulations due to issues with poor adherence driving lack of efficacy in women, even with a more potent ARV-based microbicide (63).
1% Tenofovir Gel
Given the lack of efficacy observed with nonspecific and moderately specific microbicides, the microbicide field pivoted towards the development of potent ARVs, used to treat HIV infection, for microbicide development (2). These antivirals were mainly inhibitors of the HIV reverse transcriptase, that block an early stage of the virus life-cycle (2). The ARV-based microbicide gel comprising 1% tenofovir, a nucleotide analogue reverse transcriptase inhibitor, was developed. Tenofovir was selected due to the long intracellular half-life of the active form (tenofovir diphosphate), its high genetic barrier for development of drug resistance, tolerability, and the relatively low prevalence, at the time, of circulating tenofovir-resistant HIV harbouring the K65R mutation (2, 64).
While 1% tenofovir gel demonstrated efficacy in the phase IIB CAPRISA 004 trial in preventing the sexual transmission of HIV in women compared to a placebo gel (65), these findings were not confirmed in the follow up phase III FACTS 001 study (63). These results can be explained in part due to lack of adherence, confirmed by monitoring of tenofovir levels. The poor adherence is likely due to the impractical BAT24 dosing regimen (applying gel twice not more than 12 hours before and after coitus) and the difficulty of integrating this regimen into the daily lives of younger women (65, 66). The VOICE study, that included arms assessing daily non-coitally dependent intravaginal dosing of 1% tenofovir gel, also failed to show efficacy due to lack of adherence (20). 1% TFV gel for intravaginal use is no longer being pursued as a microbicide for women.
Tenofovir based IVRs and films are under various stages of preclinical/clinical development in addition to other ARV-based microbicides and delivery modalities. In addition, despite the disappointing phase III clinical data for 1% tenofovir gel in women (63), rectal microbicide gels containing 1% tenofovir are still being pursued (4). Please refer to the following website for a more information on the ARV-based prevention preclinical and clinical pipeline (https://www.avac.org/infographic/arv-based-prevention-pipeline).
Given the disappointing clinical trial outcomes with 1% tenofovir gel, the field shifted its attention away from gels towards sustained release modalities such as IVRs (67). IVRs releasing the nonnucleoside RT inhibitor, dapivirine, were evaluated in parallel in phase III studies (The ASPIRE and the RING studies) (12) (68). Both studies demonstrated modest yet significant dapivirine ring efficacy (27% and 31%) in preventing HIV infection in women compared to IVR without drug (12, 68). These were landmark studies in the microbicide field representing the first phase III clinical trials showing significant HIV protective effect of a topical microbicide in women. Dapvirine rings are currently being evaluated in the DREAM and HOPE open label studies and moving towards licensure.
HIV-specific monoclonal antibodies (mAb) are also being advanced for HIV prevention and treatment, recently reviewed by Anderson and colleagues (69). These broadly neutralizing monoclonal antibodies (bNAbs) are derived from HIV positive individuals and target various epitopes on the HIV envelope (69). They can be divided into three generations based on potency and neutralisation breadth (69). In addition to direct HIV-1 neutralizing activity, many of these antibodies are reported to block HIV infection through several additional Fc-mediated antiviral effector functions such as antibody-dependent cell-mediated cytotoxicity (ADCC) (69)
Numerous studies have evaluated either intravenous administration of various HIV mAbs or topical administration of HIV mAbs in protecting against simian-HIV (SHIV) mucosal challenge (vaginal or rectal) in monkeys (69-71) (72) (73) (63, 74). Collectively these studies demonstrated the viability of both parenteral and mucosal delivery of mAbs to prevent vaginal and rectal transmission of SHIV and underpin current clinical studies evaluating bNAbs for HIV prevention in humans (69). Efficacy of the vaginally administered mAb VRC01 against HIV-1 transmission has also been demonstrated in the humanised mouse model (75).
Currently, there is a major focus on delivering bNAbs systemically (passive immunization) for preventing the sexual transmission of HIV-1 with several clinical trials planned or ongoing (avac.org) (69). In contrast, there is relatively less attention on the development of mAbs for use as a topical microbicide. The cost of production of clinical grade human monoclonal antibodies is likely a major hurdle (69). To address cost and scale of production of mAbs, transgenic technologies have been pursued with mAb production in plants being the most prominent in the context of HIV prevention strategies. Two groups, one in the USA and the other in Europe have progressed mAbs produced in plants to phase I clinical trials (69, 76) that include successful delivery via an IVR (77). A phase I study of a topical microbicide gel comprising human anti-HIV-1 mAbs has also been completed and found to be safe over a 12 day period of daily dosing (78).
Another strategy representing an inexpensive source of antibodies for vaginal or rectal microbicides is generating mAbs targeting the HIV envelope in cows (79). While there are currently no clinical trials of bovine mAbs there is an interest in this technology as demonstrated by a published study from a leader in the bNAb field based in the United States (80) and pioneering studies performed in Australia (79, 81, 82).
Rectal Microbicides and Douches
While shown to be safe in clinical trials performed in women (65, 83), the vaginal tenofovir gel formulation is hyperosmolar and disrupts epithelial integrity (84, 85) raising concerns of safety for use in the rectal compartment, which is relatively more fragile than the vagina, comprising a single layer of columnar epithelium. Accordingly, a 1% tenfovir gel formulation with 73% lower osomolality was developed for dual compartment use (86).
The reduced glycerine content 1% tenofovir gel modified for rectal use was safe when used daily or before and after receptive anal intercourse in MSM and transgender women from Thailand, South Amercia, mainland United States and Peru as demonstrated in the MTN-017 phase II study (87). Based on convergence interviews, 83% of participants had ≥80% adherence to the daily gel and 93% had ≥80% adherence to twice-weekly use of the gel before and after RAI (87). Adherence to RAI-associated rectal gel use was as high as adherence observed in the same study for daily oral PrEP (88). These studies demonstrate that rectal microbicides have promise for use in MSM and transgender women and await testing in efficacy trials.
To promote adherence, microbicide strategies should ideally be compatible with normal sexual behaviours. The use of cleansing enemas prior to RAI is an example of a typical behaviour in MSM. A study demonstrated that the use of a hypo-osmolar douche containing tenofovir achieved rapid and therapeutically relevant concentrations of the active form of this ARV in mouse colorectal tissue. This promising strategy awaits confirmation in non-human primates and humans (89).