The development of safe topical microbicides that effectively prevent human immunodeficiency virus (HIV) infection is a major goal in curbing the human immunodeficiency virus pandemic. A number of past failures resulting from mucosal toxicity or lack of efficacy have informed the field. Products that caused toxicity to the female genital tract mucosa, and thereby increased the likelihood of HIV acquisition, included nonoxynol 9, cellulose sulfate, and C31 G vaginal gel Savvy ® . Topical products that were ineffective in preventing HIV infection include BufferGel ® , Carraguard ® , and PRO 2000 ® . Antiretroviral drugs such as tenofovir and dapivirine formulated into microbicide products have shown promise, but there is much to learn about ideal product formulation and acceptability, and drug distribution and disposition (pharmacokinetics). Current formulations for water-soluble molecules include vaginally or rectally applied gels, vaginal rings, films and tablets. Dosing strategies (e.g. coitally dependent or independent) will be based on the pharmacokinetics of the active ingredient and the tolerance for less than perfect adherence.
Introduction
In order to be effective, an ideal microbicide should not cause mucosal irritation, should distribute into the tissues that are associated with HIV transmission, be metabolised efficiently to their active components (if necessary), and achieve adequate and sustained concentrations for protection against human immunodeficiency virus (HIV) transmission. Absorption into the systemic circulation, and subsequent exposure in blood plasma, should be minimal to diminish systemic adverse effects and the potential for the virus to become drug resistant if HIV infection should occur. A microbicide must also be easy to use and well accepted by the target population. In this chapter, we discuss the lessons learned from past microbicide failures, review successes and outline current knowledge of microbicide formulations in development.
Microbicide formulation
Product considerations
Microbicide vehicles are important, as they allow the active drug to reach the proper site of activity, to distribute throughout the luminal surface (vagina or colorectum), and remain at the surfaces for an adequate (as yet undefined) period of time. Vehicles can include aqueous gels, reservoir rings, and locally applied solid films and tablets.
Physiochemical characteristics of a gel vehicle that must be optimised include volume, viscosity, yield stress, sheer rate, pH and osmolality. Reservoir rings must be formulated to provide long-term, predictable, sustained release of an active component throughout the vaginal compartment. Film polymers should be non-toxic, non-irritant, have adequate wetting, spreadability, disintegration over time, and strength. Solid vaginal tablets must also have adequate disintegration and dispersion throughout the vaginal compartment without causing irritation. For all formulations, it is important that the pharmaceutically active compound be stable in the vehicle in which it is formulated, and uniformly distributed to allow for broad distribution in the vaginal compartment. A full list of the advantages and disadvantage of each of the formulations in development is presented in Table 1 .
| Formulation | Advantages | Disadvantages |
|---|---|---|
| Gels | Female controlled. Discrete. Lubricating properties. Limited systemic absorption. Low side-effect potential. | Effective compound must be stable in aqueous solution Stability. Applicators required. Leaking is possible. Frequent application may be required. Adherence disadvantage. Uneven distribution possible, depending on gel properties. |
| Rings | Female controlled. Discrete. No applicator required. No leakage. Long term. Adherence benefit. Limited systemic absorption. Low side-effect potential. Can be combined with contraception. | Sustained delivery must occur. Distribution may not be uniform throughout genital compartment. May be uncomfortable or difficult to insert for some. More complex manufacturing. |
| Films and tablets | Female controlled. Discrete. No applicator needed. Viable option for drugs that would be degraded in aqueous solution. Could be formulated for immediate for timed release. No leakage. Low side-effect potential. Can be combined with contraception. | Absorption may be dependent upon local hydration. Local irritation could result from contact with solids. Concern for uniformity of distribution. |
Microbicide formulation
Product considerations
Microbicide vehicles are important, as they allow the active drug to reach the proper site of activity, to distribute throughout the luminal surface (vagina or colorectum), and remain at the surfaces for an adequate (as yet undefined) period of time. Vehicles can include aqueous gels, reservoir rings, and locally applied solid films and tablets.
Physiochemical characteristics of a gel vehicle that must be optimised include volume, viscosity, yield stress, sheer rate, pH and osmolality. Reservoir rings must be formulated to provide long-term, predictable, sustained release of an active component throughout the vaginal compartment. Film polymers should be non-toxic, non-irritant, have adequate wetting, spreadability, disintegration over time, and strength. Solid vaginal tablets must also have adequate disintegration and dispersion throughout the vaginal compartment without causing irritation. For all formulations, it is important that the pharmaceutically active compound be stable in the vehicle in which it is formulated, and uniformly distributed to allow for broad distribution in the vaginal compartment. A full list of the advantages and disadvantage of each of the formulations in development is presented in Table 1 .
| Formulation | Advantages | Disadvantages |
|---|---|---|
| Gels | Female controlled. Discrete. Lubricating properties. Limited systemic absorption. Low side-effect potential. | Effective compound must be stable in aqueous solution Stability. Applicators required. Leaking is possible. Frequent application may be required. Adherence disadvantage. Uneven distribution possible, depending on gel properties. |
| Rings | Female controlled. Discrete. No applicator required. No leakage. Long term. Adherence benefit. Limited systemic absorption. Low side-effect potential. Can be combined with contraception. | Sustained delivery must occur. Distribution may not be uniform throughout genital compartment. May be uncomfortable or difficult to insert for some. More complex manufacturing. |
| Films and tablets | Female controlled. Discrete. No applicator needed. Viable option for drugs that would be degraded in aqueous solution. Could be formulated for immediate for timed release. No leakage. Low side-effect potential. Can be combined with contraception. | Absorption may be dependent upon local hydration. Local irritation could result from contact with solids. Concern for uniformity of distribution. |
Toxicity to vaginal mucosa
Above all, the field has learned that microbicides must demonstrate no mucosal toxicity, which can lead to irritation, epithelial disruption, release of cytokines, immune activation and changes to flora, which increase the risk of HIV acquisition. In order for a product to be compatible with the mucosal environment, physiologic pH must be maintained and the product should be iso-osmolar. Past microbicide gel failures owing to local mucosal toxicity included those containing nonoxynol 9, cellulose sulfate, and Savvy ® . The following is a summary of the findings from those studies.
Nonoxynol-9
COL-1492 was a phase III, randomised, placebo-controlled study in 765 female sex workers using 1.5 g of carbomer gel (a polymer with bio-adhesive properties), containing 3.5% nonoxynol-9 (a spermicidal surfactant). A 16% seroconversion rate was seen in the active arm of the study, compared with 12% in the placebo arm. Women who used the product more frequently had higher rates of HIV infection and were found more likely to have lesions with epithelial disruption. Those women reporting a mean applicator use greater than 3.5 per working day had a hazard ratio of 1.8 (1.0 to 3.2) for HIV incidence ( P = 0.03) compared with placebo. Subsequently, it was determined that nonoxynol-9 causes cytokine release, immune-cell recruitment, and alterations in vaginal flora, all of which may make women more susceptible to HIV acquisition. In particular, release of the cytokine interleukin-1 activates nuclear factor-kappa B, leading to chemokine-induced immune cell recruitment, which provided host cells for HIV-1 infection and increased HIV replication. In another study, after the use of nonoxynol-9 containing diaphragms, women had vaginal colonisation of Escherichia coli , Enterococcus species, and anaerobic gram-negative rods, which could also lead to immune activation and immune-cell recruitment.
Cellulose sulfate
Cellulose sulfate is a polyanion entry inhibitor with in-vitro activity against HIV-1 and HIV-2, as well as Neisseria gonorrhoeae , and Chlamydia trachomatis . It was found to be well tolerated in early phase safety studies using single and multiple daily doses for up to 2 weeks. A phase III study compared 3.5 ml of 6% cellulose sulfate gel with placebo gel in 1398 women 1 h before intercourse. Interim analysis found 24 seroconversions in the cellulose sulfate arm and 11 in the placebo arm (hazard ratio 2.23; P = 0.02) leading to premature discontinuation of the study. Final analysis revealed 25 out of 706 seroconversions in the cellulose sulfate arm and 16 out of 692 in the placebo arm (hazard ratio 1.61; P = 0.13). Possible explanations of the results include increased inflammatory reactions and immune cell recruitment. Although not statistically significant, one of the safety studies had shown slightly more women with signs of increased inflammation on colposcopy after using cellulose sulfate gel twice daily for 14 days.
SAVVY ®
The chemical compound C31 G (SAVVY ® ) is a surfactant that disrupts the outer membrane of HIV with a similar mechanism to nonoxynol-9. In multiple phase I studies, signs of irritation with repeat dosing of 1% SAVVY ® were reported, but considered less toxic than nonoxynol 9. One colposcopy study found 1.2% SAVVY ® to cause similar epithelial disruption as 2% nonoxynol-9. Subsequently, a phase III study enrolling 2153 women using 3.5 ml of product in a coitally dependent manner observed more seroconversions in the active arm (21 out of 1076) than in the placebo arm (12 out of 1077). The study was stopped early because of an overall lower than expected HIV incidence: mid-point analysis determined that an additional 1980 participants would be needed to reach the overall 66 events required to show efficacy. The true hazard ratio and reasons for the increased number of infections in the SAVVY ® arm could not be determined. An additional parallel phase III study of 2142 women in Ghana was also stopped early because of a lower than expected HIV incidence rate.
Vivagel ®
Vaginal irritation and inflammation was found to occur when 3% Vivagel ® (SPL 7013) was used twice daily for 14 days. Two out of 35 women discontinued the gel as a result of genital irritation, confirmed on colposcopy. A second study of 3% Vivagel ® evaluated inflammatory markers in the vaginal mucosa of women using the gel twice daily for 14 days. By day 7, interleukin-6, CD8+ and CD69+ cells, CD4+ and CD69+ cells and CD4+ and CCR5+ cells were all elevated ( P < 0.05) in women using the gel compared with placebo. By day 14, interferon gamma, interleukin-5 and interleukin-10 ( P = 0.001) were elevated. All were reversible with discontinuation of VivaGel ® , and this product is continuing in Phase II studies for the treatment and prevention of bacterial vaginosis.
Lack of efficacy
A number of topical microbicides have failed to show efficacy at preventing HIV infection, including BufferGel ® , PRO 2000 ® , and Carraguard ® . By maintaining vaginal pH 4.0–4.5 in the presence of semen, BufferGel ® was intended to protect against HIV infection by broadly acidifying pathogens. A phase II/IIb, four-arm study in 3101 women comparing BufferGel ® with 0.5% PRO 2000 ® gel, placebo gel, and no gel found no difference in HIV infection rates between BufferGel ® and placebo gel (hazard ratio 1.10; P = 0.63), and BufferGel ® and no gel (hazard ratio 1.05, P = 0.78). This same study showed modest benefit of a 0.5% PRO 2000 ® gel over placebo gel (hazard ratio 0.7, P = 0.10) and no gel (hazard ratio 0.67, P = 0.06).
PRO 2000 ® gel is a synthetic naphthalene sulfonate polymer that inhibits viral attachment and entry. It demonstrated antiviral activity in macaques and in vitro. A subsequent phase III study showed that 2% and 0.5% of PRO 2000 ® gel were ineffective at preventing HIV infection compared with placebo. The 2% PRO 2000 ® gel arm was stopped prematurely due to futility, and the 0.5% PRO 2000 ® gel arm failed to show a difference in HIV incidence compared with placebo (hazard ratio 1.05; P = 0.71).
Carraguard ® contains the polyanion carrageenan. The mechanism is not fully understood, but it may bind the positively charged regions of the viral envelope and act as a barrier to host cells. It was previously shown to prevent effectively HIV infection in vitro and in macaques. In a phase III study conducted in 6005 women in South Africa, 4 ml of 3% Carraguard ® failed to demonstrate efficacy compared with placebo (covariate adjusted hazard ratio 0.87, 95% CI 0.69 to 1.09).
Owing to the significant number of failures of surfactants, polyanions, and general antimicrobials, a number of small molecule antiretroviral drugs are under investigation. A review of the microbicide formulations that are currently under development follows. A summary of these data are presented in Table 2 .
| Study | Product/formulation | Methods | Efficacy and pharmacokinetic results |
|---|---|---|---|
| Tenofovir with or without emtricitabine | |||
| Schwartz et al. | Tenofovir 1% gel. | Single and multiple dosing of once or twice daily 4 ml applications for 2 weeks in 49 women. | Concentrations ranged from 1.2 × 10 4 to 9.9 × 10 6 ng/ml after a single dose, and 2.1 × 10 2 to 1.4 × 10 6 ng/ml after multiple dosing in fluid samples with no difference in concentrations between proximal and distal tissues. Tenofovir diphosphate levels ranged from 7.1 × 10 3 to 8.8 × 10 ng/ml in endocervical cells, and drug concentrations were minimal in blood plasma with a median maximum concentration of 4.0 ng/ml after a single dose and 3.4 ng/ml after multiple dosing. |
| Hendrix et al. MTN 001 | Tenofovir 1% gel, tenofovir oral tablet, or both. | Crossover design to compare acceptability, tolerability and pharmacokinetics in 144 women. Daily doses of either oral tenofovir, tenofovir gel, or both, were administered daily for 6 weeks in each period. | African women preferred the gel form; American women preferred the oral form; most adverse effects were associated with oral tenofovir. Tissue concentrations of the active tenofovir triphosphate were 2352 fmol/mg after gel administration compared with 17 fmol/mg after oral administration. Peripheral blood mononuclear cell tenofovir–diphosphate concentrations were 52 fmol/million cells after oral administration and 5 fmol/million cells after gel administration. |
| Anton et al. MTN 006 | Tenofovir 1% gel. | 18 healthy participants received a single dose of oral tenofovir, then a single dose of either vaginally formulated gel used rectally or placebo gel rectally (2:1), followed by 7 days of daily gel dosing. | Rectal tolerability and acceptability was low after both single and multiple doses. Ex-vivo human immunodeficiency virus challenge of rectal biopsies showed protection after multiple dosing with rectally applied tenofovir gel compared with placebo, but not single gel dosing. |
| Abdool Karim et al. CAPRISA004 | Tenofovir 1% gel. | 843 women used one application (4 ml) of gel up to 12 h before and after sex for an average of 18 months of follow up. | Tenofovir was 39% effective at preventing HIV infection with average adherence and 54% effective in the highest of adherers. The tenofovir concentrations correlated with likelihood of infection, with women with cervicovaginal fluid concentrations greater than 1000 ng/ml more likely to be protected and less than 1000 ng/ml less likely to be protected. |
| VOICE | Tenofovir 1% gel; tenofovir 300 mg oral tablet; tenofovir plus emtricitabine oral tablet. | Once-daily dosing in 5000 women (approximately 1000 women in the gel arm). | Gel arm stopped early due to futility in November 2011 after 2 years of follow up. Full results expected in early 2013. |
| FACTS | Tenofovir 1% gel. | Approximately 2200 women randomly assigned to tenofovir gel or placebo gel used one application of gel up to 12 h before and after sex. | Trial is ongoing. |
| Dapivirine | |||
| Nel et al. | Dapivirine 0.001%, 0.002%, 0.005%, 0.02% gel. | 2.5 ml applied vaginally twice daily for 42 days in 119 women. | All doses were found to be safe, with dose proportional plasma concentrations across the dosing range. The maximum observed plasma concentration was 474 pg/ml in the 0.02% dose group. |
| Nel et al. | Dapivirine 0.001%, 0.005%, 0.02% gel. | 2.5 ml applied vaginally in 18 women for 10 days; once daily on days 1 and 10 and twice daily on days 2–9. | All doses were safe and well tolerated. Mean peak values in cervicovaginal fluid were 4.6–8.3 × 10 6 pg/ml on day 1 and 2.3–20.7 × 10 6 pg/ml on day 10. Dapivirine concentrations in vaginal tissue biopsies collected at day 10 were 1.0–356 × 10 3 pg/mg. |
| Nel et al. | Dapivirine 0.05% gel. | 2.5 g applied vaginally daily for 11 days in 36 women. | Cervicovaginal fluid concentrations were at least 5 logs higher than the in-vitro inhibitory concentrations and maximum plasma concentrations did not exceed 1.1 ng/ml. |
| Romano et al. | Dapivirine 25 mg, and 200 mg ring. | 25 women inserted the ring for 7 days, with blood plasma and vaginal fluid sampling from the introitus, cervix, and ring area. | Both concentrations of rings were found to have concentrations throughout the genital tract that exceeded 1000 times the EC50. The lowest concentration in any of the tissue samples was over 30-fold higher than the in-vitro EC90. Mean plasma concentrations were less than 50 pg/ml. |
| Nel et al. | Dapivirine 25 mg Ring. | 48 women randomised (3:1) to dapivirine ring or placebo ring. Group A inserted a first ring for 28 days then a second ring on day 31 for another 28 days. Group B inserted a first ring for 35 days, then a second ring on day 38 for 21 days, then a third ring on day 59 for 24 h. | Dapivirine drug concentrations in vaginal fluid peaked within 24 h remained consistent for up to 35 days at levels well above the IC50, with even distribution throughout the genital tract. The maximum plasma concentration measured was 553 pg/ml. |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
