DRUG DELIVERY: FUNDAMENTALS AND APPLICATIONS

Delivery of drugs to the vagina was thought to be limited to topical administration until 1918, when Macht reported systemic absorption of vaginally dosed morphine, atropine, and potassium iodide. Nearly a century later, drug delivery to the female reproductive tract (FRT), most notably the uterus and vagina, has become an essential component of female reproductive health. Previous reviews largely focused on vaginal delivery; however, due to the connectivity of the reproductive organs, we believe the whole FRT should be considered.

A diverse array of active pharmaceutical ingredients (API) has been delivered via the FRT for both topical and systemic effect. Clinically, these most commonly include estrogens, progestins, and anti-infective agents. The HIV prevention field has extensively investigated the prophylactic delivery of antiviral drugs to the FRT since the turn of the twenty-first century, catalyzing innovation in vaginal dosage form design and development. Vaginal delivery has many advantages, including ease of use, painlessness, privacy, reversibility, avoidance of hepatic first-pass metabolism in the case of systemic delivery, and direct drug elution to the site of action in the case of topical delivery. Despite its importance to women’s health, FRT drug delivery has not been a popular area of study for pharmaceutical scientists, likely because of an increasingly outdated social stigma surrounding women’s health and sexuality. Nonetheless, several important historical “firsts” in pharmaceutical science have been achieved in response to problems in women’s health, including the first commercially available zero-order release implant (Norplant^®) and the first melt-extruded controlled-release device (NuvaRing^®).

In this chapter, we provide the reader with broad overview of the principles and applications of drug delivery to the FRT, including a brief overview of anatomy and physiology, a description of drug transport principles and physiological effects, a summary of the benefits afforded by FRT delivery, and a survey of existing and upcoming drug delivery systems (DDS) and technologies. We focus largely on systems that have entered clinical development or have a large body of preclinical data to support clinical viability.

12.2 ANATOMY AND PHYSIOLOGY OF THE FRT

The FRT is a continuous passageway from the ovaries to the vaginal introitus and consists (traveling distally) of the fallopian tubes, uterus, cervix, and vagina (Figure 12.1).

images

FIGURE 12.1 Sagittal view of the human female reproductive tract (FRT). (Modified from Tortora, G.J. and Derrickson, B.H., eds., Principles of Anatomy and Physiology, 14th ed. 2013. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

Gonadal differentiation occurs during the eighth to the ninth week of gestation. In the female embryo, the mesonephric (Wolffian) ducts completely regress and the paramesonephric (Müllerian) ducts develop into the proximal vagina, uterus, and fallopian tube. The two paramesonephric ducts meet the posterior wall of the urogenital sinus to form the sinovaginal bulbs, which, in turn, form the lower vagina.

As a result of the different embryonic origin of the lower and upper vagina, both blood supply and innervation of the upper and lower vagina differ. The lower one-fifth to one-quarter of the vagina is innervated from the deep perineal nerve, which is a branch of the pudendal nerve. The remainder of the vagina is innervated by the uterovaginal nerve plexus. Similarly, blood supply to the upper vagina is supplied from the uterine artery and vaginal artery, while the lower vagina is supplied by the pudendal artery and middle rectal artery.

The epithelium of the upper and the lower portions of the FRT differs, again as a result of differing embryonic origin. The Müllerian epithelium in the upper FRT differentiates into a single layer of columnar epithelial cells, while the vaginal plate differentiates into stratified squamous epithelium, meeting at the transformation zone, whose location starts on the ectocervix but migrates proximally with age.

12.2.1 OVARIES

Ovaries are solid, hormone-secreting structures located inside the peritoneal sac that serve as the proximal termini of the FRT. The adult ovary is 2.5–5.0 cm long, 1.5–3.0 cm wide, and 0.6–2.2 cm thick. Ovaries have four layers:

1. Germinal epithelium, the outer layer (neither germinal nor an epithelium)

2. Capsule (white coat), located underneath the germinal epithelium, composed of fibrous tissue

3. Cortex, comprising the bulk of the ovary, where the ova develops

4. Medulla, located inside the cortex, containing blood vessels and connective tissue

The ovaries are generally not considered an important site for drug delivery, but many ovarian pathologies, particularly ovarian cancer, might benefit from local or targeted drug delivery.

12.2.2 FALLOPIAN TUBES

The fallopian tubes are a symmetric pair of seromuscular channels, approximately 7–14 cm in length and 0.01–1 cm in diameter, which run medially from each ovary and attach to the uterus. The fallopian tubes transport eggs from the ovary into the uterine cavity through a combination of peristaltic contraction and flow induced by ciliary motion. Thus, occlusion of the tubes results in sterilization.

12.2.3 UTERUS

The uterus is a roughly pear-shaped organ approximately 8 cm long and 5 cm wide, with involuntary smooth muscle walls approximately 1.25 cm thick. These smooth muscle cells are capable of hypertrophy and recruitment of new myocytes from stem cells within the connective tissue, in order to accommodate expansion during pregnancy. The walls of the uterus are lined with the endometrium, a glandular epithelium that changes during menstruation and serves as the site of ovum implantation. Importantly, the pH of the intrauterine environment is typically near neutral, in contrast to the typically (but not always) acidic pH of the vagina (discussed in detail in the following text).

Uterine drug delivery is primarily used for long-term contraception, but there is a need for development of FRT delivery systems to target other pathologies, including uterine fibroids and endometriosis. Uterine fibroids are benign tumors that arise from smooth muscle cells in the myometrium of the uterus and are a common indication for hysterectomy. They are quite prevalent and can result in painful menstruation, bleeding, and reduced fertility in women. Endometriosis occurs when endometrial tissue grows in locations outside the uterine cavity, resulting in a chronic immune response, and symptoms including pain, infertility, and fatigue.

12.2.4 CERVIX

The opening of the uterus forms the cervix, a narrow channel composed of connective tissue and smooth muscle that protrudes slightly into the upper vagina, terminating at the cervical os. The exterior epithelial surface of the vaginal canal inside the vagina is referred to as the ectocervix, and the luminal surface is referred to as the endocervix. Importantly, goblet cells exist in the cervical canal and secrete mucin-containing mucus that can act as a barrier to sperm and infectious pathogens. Mucus is secreted cyclically into the vagina, where it can also act as a barrier to infection. As sperm must traverse the cervix to achieve fertilization, all contraceptive barrier devices function by preventing the passage of sperm from the proximal vagina into the cervix. Some contraceptive barriers also incorporate a drug delivery component and are discussed later in the chapter.

12.2.5 VAGINA

The vaginal tract of an adult female is a collapsed pseudocavity with a slight “S” shaped structure approximately 2 cm wide and 6–12 cm long. A common misconception is that the vaginal tract is a straight tube, which often leads to a fear that items placed in the proximal vagina can easily fall out. However, there are two distinct portions of the vagina, where the proximal axis, nearly horizontal when a woman is standing, forms a 130° angle with the distal axis. The walls of the vagina are generally formed into laterally oriented 2–5 mm thick folds (rugae) that are in close apposition. The vagina is capable of holding a few milliliter of fluid or gel without leakage.

Histologically, the vaginal wall consists of the following:

• A superficial nonsecretory stratified squamous epithelial layer (with underlying basement membrane)

• The lamina propria

• A muscular layer consisting of smooth muscle fibers

• The tunica adventitia, consisting of areolar connective tissue

12.2.5.1 Vaginal Epithelium

The vaginal epithelium is made up of noncornified, stratified squamous cells (see Chapter 4, Figure 4.7), is on average 50–200 μm thick, and consists of five cell layers:

1. Superficial (approximately 10 rows of cells)

2. Transitional (about 10 rows of cells)

3. Intermediate (about 10 rows of cells)

4. Parabasal (about 2 rows of cells)

5. Basal (a single row of cells)

Intercellular channels pass through the vaginal epithelium, allowing for the transport of leukocytes and large proteins, as well as fluid exudate from the vesicular plexus. Immune system cells can be found below the lamina propria and bridging the basal layer.

12.2.5.2 Lamina Propria

The lamina propria is a dense layer of vascularized, collagenous connective tissue beneath the vaginal epithelium. A diverse array of cells can be found in this layer, including fibroblasts, macrophages, mast cells, lymphocytes, Langerhans cells, plasma cells, neutrophils, and eosinophils. Drugs delivered to the vagina gain access to the systemic circulation through blood vessels in the lamina propria. Additionally, the lamina propria contains a lymphatic drainage system, which feeds into four lymph nodes: the iliac sacral, gluteal, rectal, and inguinal.

12.2.5.3 Cervovaginal Fluid

The interior of the vagina is coated with a thin layer of a complex biological media known as cervovaginal fluid (CVF). Approximately 2 g of CVF is produced per day in the absence of sexual arousal. CVF contains secretions from the cervix in the form of mucins and secretions from the Bartholin’s and Skene’s glands, endometrial fluid, and fluid transuded from the vascular bed of the vaginal tissue. It also contains a large number of squamous epithelial cells and commensal bacteria. During the menstrual cycle, this fluid increases in volume through mixing with fluid from the uterus, ovaries, follicles, and peritoneum, permitting movement of spermatozoa.

CVF acts as a barrier against infections, not only by directly binding to microorganisms, but by maintaining an acidic pH, which inhibits pathogenic bacterial proliferation. Additionally, CVF contains antimicrobial substances such as defensins, lysozyme, lactoferrin, fibronectin, spermine, and secretory IgA and IgG. Naturally, the vaginal fluid may contain highly variable amounts of semen and spermatozoa following intercourse.

A wide range of aerobic and anaerobic bacteria normally are found in the vagina with concentrations of 10⁸–10⁹ colonies/mL of vaginal fluid. Lactobacilli (found in 62%–88% of asymptomatic women) that normally reside in the vagina break down secreted glycogen to form lactic acid, resulting in an acidic vaginal environment. Many reviews give healthy CVF pH ranges between 3.5 and 5.5. However, studies have shown an even larger range in women who show no clinical pathologies, perhaps indicating there is not a “normal” vaginal pH as commonly thought. The pH of the vagina differs depending on age, ranging from a more acidic level during sexually mature childbearing years to neutral or alkaline before puberty and after menopause. Shifts in the CVF pH can be, but are not always indicative of vaginal infection: a more alkaline environment can indicate bacterial vaginosis or trichomoniasis, while a more acidic environment can indicate fungal infection.

12.3 DRUG TRANSPORT IN THE FRT AND THE IMPACT OF VAGINAL PHYSIOLOGY

12.3.1 COMPARTMENTAL DESCRIPTION OF DRUG TRANSPORT

There are three basic compartments of importance when considering spatial drug pharmacokinetics (PK) in the FRT (Figure 12.2): (1) cervovaginal/uterine fluid, (2) epithelial and underlying tissues, and (3) systemic circulation. Cervovaginal/uterine fluid generally acts as a dissolution medium that relays drug molecules from dosage form to tissue. The fluid provides a sink-type boundary in the case of water-soluble drugs or can become a saturated (or partially saturated) rate-controlling impediment for release, in the case of less water-soluble drugs. In either case, the fluid serves as a source for drug transport throughout the tissues. Fluid convection in the FRT serves to effectively spread drugs distally from the device to the introitus, which is particularly advantageous for devices that provide a constant drug source from a fixed point in the FRT (such as an intravaginal ring [IVR] or intrauterine system). Also, it is likely that a fraction of any dose delivered vaginally remains dissolved in the fluid and is lost through the introitus, which could lead to a reduction in dose availability for bolus-type doses (such as gels and tablets). It has also been suggested that drugs can partition directly from dosage form to epithelium, thus some fraction of released drug molecules can bypass the fluid compartment completely.

The tissue compartment can be further stratified into various subcompartments. When considering continuum-scale drug transport, the tissue can be divided into the epithelium (non-vascularized) and the underlying lamina propria (vascularized). For vaginal delivery, a linear concentration reduction in drug concentration can be expected across the epithelium by Fick’s Vaginal fluid laws, approximating the epithelium as a thin plane, while combined diffusion and elimination to the bloodstream should result in an exponential decay of drug concentration throughout the vascularized tissues. For topical prophylactic applications, a sufficient drug source must be maintained in the fluid compartment to ensure an effective drug level surrounding infectable cells at various depths in the tissue. For drugs that are activated intracellularly (such as many antiretrovirals), the tissue can be further stratified at the cellular level and chemical reactions must be considered along with diffusion and elimination.

images

FIGURE 12.2 Generalized anatomy of the female reproductive tract and display of the various pharmacokinetic compartments. (a) An intravaginal ring is shown as a model dosage form and can elute drug into (1) vaginal fluid, (2) epithelial and underlying tissues, and (3) systemic circulation. (b) The vaginal ring is in contact with the walls of the vagina and is in contact with vaginal fluid. The virtual canal is anatomically connected through the cervical os to the uterus and fallopian tubes and ovaries. (Reprinted with permission from Kiser, P.F., Johnson, T.J., and Clark, JT. State of the art in intravaginal ring technology for topical prophylaxis of HIV infection, AIDS Rev. 2012; 14:62–77. Copyright 2012 Permanyer Publications.)

Blood circulation serves as an intermediate sink for topical FRT delivery, in contrast to its function as a source for oral and parenteral delivery. Once in circulation, drugs delivered via the FRT are distributed similarly throughout the body but likely eliminated from the body via the same hepatic and renal clearance mechanisms.

From a pharmacokinetic/pharmacodynamic (PK/PD) perspective, intravaginal delivery can have several advantages over delivery from other routes. Intravaginal delivery requires reduced systemic drug concentrations to achieve effective levels in the FRT fluid and tissue, in comparison to oral and transdermal delivery. For systemic delivery, the hepatic first-pass effect is avoided, reducing presystemic drug loss. Furthermore, the FRT provides multiple sites for the noninvasive placement of controlled-release devices (e.g., IVRs and intrauterine systems), thus allowing for constant or sustained drug concentration in the FRT and the blood plasma, which is impossible to achieve through intermittent oral dosing. In a study directly comparing the systemic absorption of estrogen following vaginal, oral, or transdermal delivery over a 21-day period, vaginal delivery provided comparable efficacy to the other routes while avoiding unnecessarily high plasma concentrations (Figure 12.3).

Finally, we have observed that drug delivered vaginally in nonhuman primates (NHP) can be found at lower concentrations in proximal parts of the FRT, including the uterus and ovaries, despite distal fluid convection. Some have proposed a counter current exchange mechanism whereby drug could exchange from the venous bed in the vagina to the uterus. It is also possible that material is moved into the upper reproductive tract by diffusion down a concentration gradient in the cervical canal.

images

FIGURE 12.3 Comparison of systemic plasma ethinyl estradiol (EE) concentrations for 21 days via (1) vaginal delivery (NuvaRing^®; shown in blue, 15 μg/day vaginal EE), (2) transdermal delivery (Evra^™ transdermal patch; shown in red, 20 μg/day transdermal EE), and (3) oral delivery (Microgynon^® combined oral contraceptive [COC] pill; shown in green, 30 μg/day oral EE). Systemic C_max values following vaginal delivery were 78% and 65% lower than via the oral and transdermal routes, respectively; furthermore, the characteristic peaks and troughs associated with oral delivery were avoided. (Modified from Contraception, 72, van den Heuvel, M.W., van Bragt, A.J.M., Mohammed, A.K. et al., Comparison of ethinyl estradiol pharmacokinetics in three hormonal contraceptive formulations: The vaginal ring, the transdermal patch and an oral contraceptive, 168–174, Copyright 2005, with permission from Elsevier.)

12.3.2 DRUG PROPERTIES

Several physicochemical properties are of interest for drugs delivered via the FRT, including aqueous solubility, tissue permeability, and hydrophobicity (see also Chapter 4, Section 4.3.4). Solubility in the vaginal fluid is important, since the concentration dissolved in the vaginal fluid functions as the boundary condition for transport through the tissues. Compounds also must have sufficient diffusivity in the tissue compartment for them to reach their potential sites of action or to reach circulation. Although hydrophobic compounds generally have better permeability across epithelial cell membranes, small hydrophilic compounds can rapidly diffuse in the extracellular space, quickly reaching continuum-level steady-state concentrations. The extracellular concentration at any point then serves as a drug source across individual cell membranes. Aqueous solubility is potentially more important for long-term delivery from a controlled-release device because local tissue and systemic concentrations should be directly related to the source concentration dissolved in the fluid and/or the device matrix.

Chemical stability of drugs dissolved in the CVF is an important factor in selection of appropriate drugs for vaginal delivery. As discussed previously, the CVF pH can exhibit intra- and interpatient variability, which must be considered for drugs with pH-dependent stability.

12.3.3 PHARMACOKINETIC MODELS

Several models of local PK for vaginal delivery have been reported. Saltzman used a steady-state solution to a one-dimensional diffusion–elimination model to predict the vaginal transport of ¹²⁵I-labeled IgG antibodies released from poly(ethylene-co-vinyl acetate) (EVA) matrix devices and also measured the concentration as a function of tissue depth experimentally. This treats the CVF as a thin conducting surface with constant API concentration. However, this model neglects the advective transport of drugs longitudinally through the vaginal tract. Saltzman later published an improved compartmental PK model for vaginal delivery that also considered fluid advection. Geonnotti and Katz constructed a finite-element model of a two-dimensional cross section of an IVR and the vaginal tract and surrounding tissues. The results of this model indicated that the thickness and fluid velocity of the vaginal fluid boundary layer have much greater impact on drug distribution through tissue than effective drug–tissue diffusivity.

12.3.4 CYCLIC AND INTERMITTENT CHANGES IN CERVICOVAGINAL MUCUS

The effects of menstrual cycle variation and sexual intercourse on vaginal drug delivery are often not considered. In cases where the flux of a drug from a device surface is limited by its solubility in the surrounding fluid, changes in the volume and/or composition of cervicovaginal mucus (CVM) could affect drug release. Some long-acting FRT dosage forms are left in place for one or more menstrual cycles. During this time, drug release rates could be modulated by the composition of CVM, which may contain various cyclically present drug-solubilizing factors, as well as by an increase in the fluid volume.

The epithelium of the vagina changes based on response to hormones (estrogens, progesterones, luteinizing hormone, and follicle-stimulating hormone), aging, sexual cycling, and pregnancy. These changes can all affect the vaginal absorption of drugs, rendering difficulty in achieving consistent drug levels. In the follicular phase, the basal and parabasal layers increase mitosis in response to higher levels of estrogen, resulting in increased epithelial thickness. Intercellular channels also increase in frequency, resulting in a more connected epithelium. Changes in drug transporter expression and distribution may also occur.

In contrast, during the luteal phase, desquamation occurs, causing epithelial thinning, loss of structure, and increased porosity of the epithelium as well as widening of the intercellular channels. After menopause, the vaginal epithelium becomes very thin, with indistinct cell boundaries and decreased microridging.

When selecting an API for formulation in an FRT-based dosage form, these changes should be considered as they can alter drug solubility, permeability, and absorption. This can especially confound the formulation of drugs with a low therapeutic index.

The drastic increase in vaginal pH following intercourse could also result in a temporary modulation of drug solubility, concentration, and transport.

12.3.5 PREGNANCY

During pregnancy, the epithelial wall of the vagina becomes thicker and more vascularized, and blood flow slows. Recovery from this state takes several weeks postpartum. The effects of pregnancy on FRT drug transport are not often discussed, as many systems are intended to prevent pregnancy, to aid in achieving or maintaining pregnancy, are intended for use in postmenopausal women, and/or are contraindicated in pregnant women. Although intrauterine drug delivery systems (IUS) are highly effective in preventing pregnancy, pregnancies can still occur, and in these cases, IUS use is associated with a higher risk of ectopic pregnancy. Many API that could be delivered by the vaginal route are contraindicated with pregnancy. For instance, topical delivery of the chemotherapeutic sensitization agent, 5-fluorouracil, is regarded as unsafe for use during pregnancy by the FDA.

12.4 WHY DELIVER DRUGS TO THE FRT?

FRT DDS are naturally suited for women’s health applications as they are only available to women. As discussed thus far, these systems can be designed for topical and/or systemic effect. Most FRT DDS to date are indicated for elimination of infection, postmenopausal hormone replacement, and contraception. As discussed in the previous section, a key limitation is that cyclical changes in the vaginal environment across multiple time scales can complicate drug dissolution/release and tissue absorption/transport.

12.4.1 UTILITY FOR TOPICAL DOSING AND FEASIBILITY FOR SYSTEMIC DELIVERY

Direct drug delivery to the FRT is particularly advantageous when a local effect is desired and high systemic drug levels are not needed. This is the case with anti-infectives, HIV/sexually transmitted infection (STI) prophylactics, and some contraceptives (particularly low-dose, progestin-only formulations). Systemic delivery through the FRT also has many advantages, including low metabolic activity, relatively high permeability, ease of administration, prolonged retention, avoidance of hepatic first-pass metabolism, and potential for sustained and controlled release from long-acting devices.

12.4.2 REDUCTION/ELIMINATION OF USER/PATIENT INTERVENTIONS

Poor compliance/adherence to traditional oral drug therapies is a universal detractor of effectiveness, especially for prophylactic applications. Delivery to the lower FRT with longer acting devices, such as IVR, minimizes user/patient intervention, while delivery to the upper FRT with intrauterine devices/systems (IUD/IUS) effectively eliminates these interventions for the total device duration (up to 10 years for some contraceptive IUD). Poor adherence is also associated with frequency of dosing and duration of treatment across all routes of administration, strengthening the case for FRT delivery using long-acting devices when appropriate.

12.4.3 FRT INFECTIONS

A good portion of vaginal dosage forms comprise an anti-infective. The most common sources of infection are trichomonal, bacterial, candidal (yeast), and gonococcal. Candida albicans is responsible for approximately 25% of cases of vaginitis, which can be treated with local intravaginal azole agents (clotrimazole, miconazole, butoconazole, terconazole, and tioconazole). Additionally, modification of vaginal pH using a boric acid suppository has been found successful in treatment of candida, though it is not advised for pregnant women.

Trichomoniasis is a common STI caused by the anaerobic protozoan Trichomonas vaginalis, responsible for 25% of all cases of vaginitis. Because the infection resides in the vagina as well as the urethra, Skene’s glands, and bladder, an oral dose of 2 g of metronidazole or tinidazole is typically prescribed, although these compounds are also administered vaginally.

Approximately, 40% of vaginosis is bacterial in nature and caused by an overgrowth of the lactobacilli responsible for maintaining vaginal pH. Bacterial vaginosis is not often treated, since it usually can resolve on its own, but several treatments are available, including metronidazole intravaginal gels, clindamycin intravaginal creams, and a combination of oral and intravaginal clindamycin.

12.4.4 MENOPAUSAL SYMPTOMS

Surrounding and following menopause, the concomitant decrease in estrogen levels may result in atrophic vaginitis, whereby the vulvar and vaginal tissue can become pale, thin, and dry, resulting in general discomfort and painful intercourse. An estrogen is often prescribed in the form of a vaginal cream, tablet, or ring, as local delivery is particularly desirable for estrogen to avoid endometrial stimulation.

12.4.5 CONTRACEPTION

Contraceptive hormones can be absorbed efficiently to systemic circulation from the vagina. Since low-dose progestin-only contraceptives are effective primarily through local effects, such as thickening of CVM, very low doses of progestins can be applied directly to the FRT to eliminate the need for estrogen delivery and greatly reduce systemic progestin exposure. This strategy is employed by the Mirena^® and Skyla^® IUS and in levonorgestrel (LNG)-releasing IVR. Reducing the dose also allows for higher duration in IVR and IUD, which have inherent size limitations.

12.4.6 TOPICAL PREEXPOSURE PROPHYLAXIS OF HIV

The continued absence of an HIV vaccine has motivated the investigation of woman-controlled strategies that can interrupt the early events of male-to-female sexual transmission. Early studies employed vaginal gels formulated with nonspecific polymeric agents (e.g., carrageenans, cellulose sulfate) designed to prevent infection and were categorically unsuccessful. More recent efforts have proven the concept of preexposure prophylaxis (PrEP), whereby ARV drugs are administered prior to sexual exposure. To date, clinical studies of vaginal HIV PrEP have reported mixed results, but overall this strategy shows promise to impact the pandemic if products are developed that women are highly motivated to use. As discussed in the following texts, the PrEP effort has spurred innovation in nearly every vaginal product category, most notably in IVR and vaginal gels. Vaginal delivery of ARV is thought to be advantageous because drugs can be delivered directly to the site of cells that are infectable by HIV with less systemic exposure. Nonetheless, only oral tablets have been approved by the FDA for prophylactic use to date. It remains uncertain whether oral delivery of ARV prevents HIV transmission in the mucosa or later events in the immune system required for dissemination of HIV.

The terms “PrEP” and “microbicides” are often used interchangeably. Here, we distinguish that PrEP refers to the administration of agents with specific pharmacological activity against HIV, whereas microbicides are a broad category of prophylactic antivirals and anti-infectives.

12.5 DRUG DELIVERY TO THE UPPER FRT

We will next discuss existing products and emerging technologies that utilize the FRT as a route for both topical and systemic drug absorption. A pictorial survey of commercially available vaginal delivery systems is shown in Figure 12.4.

12.5.1 DELIVERY TO THE OVARIES

Because the ovaries are located within the peritoneal sac, intraperitoneal injection of chemotherapeutics has been used to treat ovarian cancer. Additionally, ^99mTc-labeled human albumin microspheres deposited in the vaginal fornix have been found to concentrate in the ovaries. This has been used as a method of evaluating fallopian tube function and illustrates another potential route of ovarian administration. Little has been explored in the field of ovarian drug delivery, and it remains a potentially interesting field for study.

12.5.2 DELIVERY TO THE FALLOPIAN TUBES

Pellets of quinacrine were first used by Jaime Zipper in Chile as a form of nonsurgical sterilization in the late 1970s. Using a modified IUD insertion device, seven 36 mg pellets of quinacrine hydrochloride were placed close to the entry to the fallopian tubes. These pellets caused chemical burns and scarring, resulting in occlusion and permanent sterilization. While over a hundred thousand women were sterilized using quinacrine, primarily in developing countries, ethical controversy in the late 1990s over inadequate testing and possible carcinogenicity resulted in its declining use.

The Essure^® system is a permanent birth control option with a success rate of 99.8%. An Essure micro-insert is a stainless steel inner coil surrounded by a nickel–titanium outer coil and polyethylene terephthalate (PET) fibers. Upon release of each micro-insert into the proximal section of each fallopian tube, the outer coil expands from 0.8 to 2.0 mm diameter to anchor the device. The local inflammatory response to the PET fibers causes fibrosis and occlusion of the fallopian tubes, resulting in permanent sterilization.

12.5.3 INTRAUTERINE DRUG DELIVERY

IUD and IUS are T-shaped polymeric devices inserted in the uterus that can provide highly effective contraception for up to several years from a single device (Figure 12.4a). Traditionally, IUD did not contain API, but provided contraception solely by exhibiting a spermicidal foreign body response in the uterus. The modern IUD shape is likely derived from attempts in the early 1900s to create a stem-type pessary that extended from the vagina into the cervix. These devices were eventually abandoned due to their high rate of infection, as well as the controversial nature of contraception at the time. In fact, two scientists, Grafenburg of Germany and Oda of Japan, were exiled from their respective countries as a result of their research into contraceptives in the 1930s. In the late 1960s, the safety of combined oral contraception (“the pill”) had come into question, prompting the investigation and clinical use of IUD as an alternative. By the early 1970s, 17 IUD or IUS were under development by 15 different companies, including the progesterone-releasing Progestasert^® IUS, developed by Alejandro Zaffaroni and colleagues at the eponymous ALZA Corporation. Progestasert^® was one of the first drug/device combinations to be used clinically. Another of these IUD, the Dalkon Shield^®, was associated with extensive safety concerns and may have caused at least 18 deaths. It was eventually removed from the market in 1974. As one would expect, this resulted in a very sour public opinion of IUD. The reader is reminded that the Dalkon Shield^® was sold at a time when FDA approval was required only for drugs, and not for medical devices.

images

FIGURE 12.4 Pictorial survey of some female reproductive tract drug delivery systems. (a) Mirena^® intrauterine drug delivery system, (b) Today^® Sponge; (c) NuvaRing^®, (d) segmented tenofovir/levonorgestrel intravaginal ring; (e) Caya Contoured diaphragm (SILCS diaphragm), (f) vaginal contraceptive film; (g) pod-insert intravaginal ring, (h) Crinone^® gel applicator. (e: Photo courtesy of PATH. All rights reserved, Seattle, Washington; f: Courtesy of Apothecus Pharmaceutical Corp., New York. Copyrighted by Apothecus Pharmaceutical Corp., 2015; g: Photo courtesy of Marc Baum, Oak Crest Institute of Science, Pasadena, CA; h: Photo courtesy of Actavis plc, Parsippany, NJ.)

After a lull in IUD/IUS use in the 1980s and 1990s, primarily resulting from issues with the Dalkon Shield^®, IUD/IUS use has resurfaced in the past 15 years, resulting in FDA approval of two Bayer-marketed drug-releasing IUS. The Mirena^® and Skyla^® IUS, FDA approved in 2000 and 2013, respectively, both deliver “microdoses” (up to 14 and 20 μg/day, respectively) of the progestin LNG in the uterus for up to 3 and 5 years, respectively, and achieve contraception largely through local effects (e.g., cervical mucus thickening and inhibition of sperm function). Both Mirena^® and Skyla^® are over 99% effective in preventing pregnancy. Mirena^® is also coindicated for treatment of heavy menstrual bleeding.

An alternative to progestin-releasing IUS is the Paragard^® IUD, also known as the “Copper T.” Paragard^® was approved in 1981 but has seen a recent resurgence in popularity along with the IUD/IUS platform as a whole. Paragard^® is a copper-releasing IUD that is also over 99% effective and can remain in place for up to 10 or 12 years. Paragard^® use results in high concentrations of copper in the cervicovaginal fluids that are toxic to sperm cells and may aid in the prevention of implantation. Although Paragard^® is primarily used as a long-acting contraceptive, it has clinically been shown to be effective as a method for emergency contraception (EC). However, hormonal EC methods, such as high oral doses of LNG, are more commonly used.

Both LNG-releasing IUS and the Paragard^® IUD consist of a T-shaped inert (nondrug releasing) polyethylene frame varying slightly in overall dimensions, where Paragard^® is the largest (32 mm wide and 36 mm long) and Skyla^® is the smallest (28 mm wide and 30 mm long). In all three devices, barium sulfate is compounded into the polyethylene matrix as a radiopacifier. Furthermore, a polyethylene monofilament is tied to a loop at the end of the vertical stem, resulting in two threads that allow the user to confirm that her IUD/IUS is still in place. A silver ring was placed around the top of the vertical stem in Skyla^® to render the device visible by ultrasound. In Mirena^® and Skyla^®, the vertical stem is covered with a cylindrical polydimethylsiloxane (PDMS or “silicone elastomer”) reservoir-type cylinder, where the reservoir core contains a mixture of PDMS and LNG. Skyla^® was introduced in 2013 as a smaller, low-dose variant of Mirena^® specifically marketed for nulliparous women, whereas Mirena^® had been originally indicated only for use in women who had at least one child. Skyla^® has a lower overall drug load (52 versus 13 mg), shorter duration (3 versus 5 years), and lower final release rate (5 versus 11 μg/day). The lower release rate was likely achieved by using a thicker outer, rate-controlling membrane for Skyla^® and/or is due to the lower drug concentration in the core (assuming the drug is mostly dissolved in the core). Nonetheless, Skyla^® is nearly as effective as its predecessor for the indicated duration, with a cumulative 3-year pregnancy rate of 0.9% (compared to a 5-year pregnancy rate of 0.7% for Mirena^®). The horizontal and vertical stems of Paragard^® are partially wrapped with a total of 176 mg of copper wire, resulting in an exposed copper surface area of 380 mm². Frameless intrauterine implants also have been developed as alternatives to traditional T-shaped designs. These systems consist of a straight cylindrical segment containing an anchor system that is kept in place by suturing to the uterine fundus. A 5-year, copper-releasing frameless IUD (GyneFix^®) is available in Europe but has not been approved in the United States, and a 20 μg/day LNG-releasing variant (FibroPlant^™) is in development.

12.6 DRUG DELIVERY TO THE LOWER FRT

12.6.1 FIRST VAGINAL DRUG DELIVERY DEVICE: THE PESSARY

“Pessary” is a term used for devices placed into the vagina for structural support of the uterus of parous women but has been used also to refer to vaginal DDS. Pessaries have been used since the time of the ancient Greeks; Polybus (≈400 BC) recommended the use of a half pomegranate, and Soranus (110 AD) recommended a linen tampon soaked in vinegar with a piece of beef. While pessaries have primarily been used for treatment of pelvic organ prolapse and urinary incontinence, the term pessary has also been used to refer to IVR that deliver estrogen for the management of urogenital atrophy. While contemporary vaginal DDS have been described as pessaries, the term is archaic and should only be used to refer to devices for pelvic floor disorders and not in the description of drug formulations or devices.

12.6.2 VAGINAL ABSORPTION OF TOXINS

Prior to 1918, it was widely believed that any compounds administered to the vagina were capable of only local effect. In 1918, Macht reported systemic absorption of molecules through the vagina by administration of a number of compounds to dogs and cats. He was able to demonstrate that compounds could quickly produce effects, including vomiting and death, upon administration to the vagina, thereby demonstrating systemic absorption.

12.6.3 CERVIX AND COMBINATION BARRIER DEVICES

Female barrier contraceptives are designed to prevent sperm transport from the proximal vagina to the cervix. However, these devices do not form a perfect barrier and have often been formulated with spermicides to increase efficacy. Common designs of such devices are the diaphragm, and cervical cap, and the contraceptive sponge.

The diaphragm is a dome-shaped cup with a steel flexible rim that holds the device in place against the vaginal walls, thereby sealing the cervix (Figure 12.4e). The cervical cap, first introduced in 1838 by Friedrich Wilde, is comprised entirely of latex and is intended to fit snugly around the base of the cervix. Diaphragms and cervical caps are generally intended for use with a spermicide. Currently, the only available cervical cap is branded as FemCap^™.

The Today^™ sponge is a sustained-release system consisting of a circular polyurethane disc impregnated with nonoxynol-9 (N9), intended to release approximately 125 mg of N9 over 24 hours (Figure 12.4b). N9 is a nonionic surfactant that immobilizes spermatozoa by removing lipid components of the midpiece and tail. The sponge is inserted against the cervix up to 24 hours before intercourse and is left in place for 6 hours following intercourse. Despite the inclusion of N9, the sponge has a failure rate of 8%–13%, considerably higher than many modern contraceptive options.

The combination barrier devices discussed earlier can be modestly effective in preventing pregnancy but are limited in that, with the exception of the female condom, they offer no protection against HIV/STI transmission, in contrast to the male condom. In fact, N9 is now thought to carry increased risk for HIV/STI acquisition (discussed in the following texts). To address this gap, strategies are under investigations that form a physical barrier to sperm, as well as a chemical barrier to infection, using modifications of SILCS diaphragm (Figure 12.4e). SILCS, also known as the Caya Countoured diaphragm, is a modified silicone diaphragm designed for easy insertion/removal and better anatomical fit. Key features are the thermoplastic nylon 6,6 spring forming a cervical rim, as well as a recessed “removal dome” to aid in easy removal. One incarnation for drug delivery is a SILCS variant designed to carry a dose of the 1% tenofovir (TFV) gel, which has shown effectiveness against HIV infection. Major and Malcolm reported a long-acting combination SILCS device, whereby the nylon spring was replaced by an elastic polyoxymethylene matrix containing the anti-HIV drug dapivirine (DPV). If effective, this variant could protect against pregnancy and HIV infection for 6 months.

12.6.4 INTRAVAGINAL RINGS

12.6.4.1 Technical Description of Intravaginal Rings

IVRs are torus-shaped drug/medical device combinations used clinically for long-acting contraception and hormone replacement therapy and have garnered recent interest as HIV prophylactics (Figure 12.4c, d, and g). IVR dimensions are loosely constrained by the width of the vaginal canal, observed to vary by as much as 48–63 mm, although the fit and comfort of IVRs are also constrained by the overall balance of elastic properties between the compressible ring and the vaginal musculature. Clinical IVR outer diameters (OD) do not vary greatly, ranging from 54 mm (NuvaRing) to 56 mm (Femring^®); however, OD of 50 to 61 mm are thought to be well tolerated based on clinical data. Interestingly, the first silicone IVR tested by Mishell had OD of up to 80 mm. The cross-sectional diameters of clinical IVR vary significantly, ranging from 4 mm (NuvaRing) (Figure 12.4c), to 9 mm (Estring^®).

As with other drug/device combinations, most IVR can be classified as either matrix or reservoir devices. Matrix IVRs are simple, monolithic devices where drug is dissolved and/or dispersed evenly throughout the polymer matrix, whereas reservoir IVRs employ an outer polymeric membrane to control the rate of drug release (see also Chapter 2). Drug release from matrix IVR usually decreases significantly with time, due to an increase in the diffusion path length following progressive inward drug depletion. Drug release rates from reservoir IVR are linearly dependent on the dissolved core drug concentration and thus can be constant (if an excess of undissolved drug is present) or can decay in a first-order exponential fashion (if the entirety of drug is completely dissolved). Most clinical IVRs are reservoir devices, but matrix IVRs have seen significant investigation by the HIV PrEP field, likely due to their potential manufacturing simplicity and reduced cost. More advanced designs have been investigated that address the challenge of delivering multiple drugs simultaneously. A fixed-dose combination was achieved for the dual hormone contraceptive IVR NuvaRing by dissolving different concentrations of two drugs with similar chemical properties in the core. Advanced designs include (1) the segmented or multisegment IVR (Figure 12.4d), which theoretically can incorporate both matrix and reservoir components, and (2) the pod-insert IVR, wherein an injection-molded ring is fitted with several reservoir-controlled release inserts (Figure 12.4g). Teller et al. recently described an IVR containing a “flux-controlled pump” (FCP) for zero-order release of hydrophobic and macromolecular compounds. In an FCP, drug tablets formed from a hydrophilic gel-forming polymer matrix are incorporated in a hollow-tube segment. Upon hydration, a drug-containing gel is extruded through one or more orifices in the tubing wall of a rigid hollow segment, resulting in the pressure-driven release of hydrophobic drugs at rates independent of their molecular properties.

All commercially available IVRs are fabricated from either thermosetting silicone elastomers or thermoplastic EVA. The DPV IVR under clinical investigation for HIV PrEP is also synthesized from a silicone elastomer. Kiser and colleagues have extensively investigated thermoplastic polyurethane elastomers as IVR matrices, particularly poly(ether urethanes), which can be synthesized in water-swellable and nonswellable variants. Most IVR designs are limited to the delivery of small molecule drugs due to similar solubility/permeability constraints; however, several designs have been investigated, including the aforementioned pod-insert ring that could allow for delivery of larger molecules.

Drug release from IVR under sink conditions has been mechanistically described by solutions to diffusion equations, which assume a sink boundary between the device and CFV/surrounding tissues. It is unclear whether a biological sink is maintained for some hydrophobic antiviral compounds in the vaginal tract. It is thought that solubility-limited release occurs (i.e., CVF does not provide a sink condition for IVR drug release) in vivo for certain device/drug combinations where the vaginal fluid concentration remains saturated or supersaturated with drug throughout the dose duration.

In general, the efficacy of long-term controlled release drug–device combinations is dependent on their ability to remain in place for the duration of use. IVR differ from other long-acting drug/device combinations in that they are not surgically implanted and are designed to be inserted and removed by the user. If an IVR is too easily deformed, the ring may be expelled as a result of the user’s day-to-day activities. However, if the force required to compress an IVR is too great, it may be difficult for the user to compress the ring prior to insertion, or the presence of the ring may cause damage to the vaginal epithelium.

12.6.4.2 Clinical History of Intravaginal Rings

IVRs were first investigated for the sustained delivery of contraceptive hormones to systemic circulation in the late 1960s. Mishell and colleagues were the first to clinically demonstrate the potential of contraception via IVR by testing a silicone IVR impregnated with the progestin medroxyprogesterone acetate (the drug now used in the Depo-Provera^® contraceptive injection) in a small PK/PD pilot trial. In a series of seminal papers, Chien and colleagues used similar silicone IVR as a case study for the discussion of controlled drug release in the context of vaginal delivery. The WHO later developed a silicone IVR that released approximately 20 μg/day of LNG for topical contraception for up to 3 months. This LNG IVR was tested in a large clinical multicenter clinical trial and was shown to be over 95% effective in preventing pregnancy; however, development was discontinued, in part due to use of the IVR being associated with the formation of vaginal lesions, potentially due to excessive device stiffness. The first contraceptive IVR to market was the progestin/estrogen NuvaRing^®, an EVA-based reservoir ring that releases the progestin etonogestrel and ethinyl estradiol (EE) at near constant rates over 21 days. The inclusion of EE and the 3-week-in/1-week-out pattern generally maintains a normal menstrual cycle, as in progestin/estrogen oral contraceptives that adhere to a similar 3 week/1 week cycle. Several important insights to IVR design and development gained from the NuvaRing^® work were published by van Laarhoven and colleagues in the early 2000s.

Four silicone-based IVRs are currently available worldwide for various indications other than general contraception. Estring^® and Femring^®, both available in the United States, deliver estrogen analogs (estradiol and estradiol acetate, respectively) for postmenopausal hormone replacement. Progering^® and Fertiring^® both release natural progesterone and are approved only in a few Central and South American countries. Progering^® is indicated for contraception in lactating women, while Fertiring^® is indicated for hormone supplementation and pregnancy maintenance during in vitro fertilization.

IVR for HIV PrEP can be traced back to the N9-releasing silicone IVR reported by Malcolm and Woolfson at the Queen’s University Belfast in the early 2000s. This group went on to formulate DPV into a silicone IVR that lead to phase 3 trials of this antiviral IVR in African women. Use of the DPV IVR demonstrated a 27% overall efficacy in preventing sexual transmission of HIV and 65% efficacy in regular users.

In the mid- to late 2000s, Kiser and colleagues published several papers reporting the in vitro and in vivo investigation of polyurethane IVR for the delivery of several ARV, including multisegment IVR designs for the simultaneous delivery of multiple API with disparate properties. One of these designs was used in the development of the first clinically tested IVR for multipurpose prevention (discussed in the following texts). ARV-releasing IVR have achieved high rates of efficacy in NHP challenge studies, and have demonstrated adherence dependent efficacy in recent dapivirine IVR trials. An IVR-releasing MIV-150 (a nonnucleoside reverse transcriptase inhibitor) demonstrated 80% efficacy in preventing simian–human immunodeficiency virus (SHIV) infection in rhesus macaques, while another IVR-releasing tenofovir disoproxil fumarate (TDF) was 100% effective in preventing SHIV infection in pig-tailed macaques challenged weekly for 4 months during continuous IVR use.

IVRs have also been investigated for delivery of lidocaine as a cervical anesthetic and oxybutynin as a treatment of overactive bladder syndrome, with the latter recently completing a Phase II clinical study. Bayer is currently developing a combination anastrozole/LNG 28-day IVR for treatment of endometriosis. Anastrozole is a third-generation aromatase inhibitor that has been used for management of severe endometriosis through suppression of estrogen synthesis. LNG, which causes the endometrium to become atrophic and inactive, has also been reported to improve endometrial pain. This ring is currently in Phase III trials.

12.6.5 VAGINAL SEMISOLIDS (GELS AND CREAMS)

Vaginal semisolids, known colloquially as vaginal gels or creams, may be the broadest category of vaginal drug delivery system with respect to their formulation, physical properties, and clinical indication. Semisolid formulations are an attractive option for drug delivery as they are typically self-administered by a patient or user and afford the potential for on-demand or as-needed use. Vaginal semisolids have many drawbacks as well, such as messiness and, as with traditional oral formulations, a tendency for poor user adherence/patient compliance. Here, we use the word vaginal “gel” in a colloquial and not scientific sense; a true gel must contain long-lasting cross-links and therefore must retain a predominantly elastic character regardless of strain rate. Vaginal semisolids made from hydrocolloids generally do not have these properties and are therefore not strictly gels but are rather viscoelastic fluids.

A majority of vaginal semisolid formulations are prepared from viscoelastic water-soluble polymers with appreciable viscoelasticity. As with other dosage form classes, semisolids can contain dispersed and/or dissolved drugs. Viscoelastic semisolid “gels” are formed by dissolution of a relatively small amount of solid (hydrophilic polymers) in large amounts of liquid, resulting in “gels” with physically or chemically cross-linked, 3D polymer matrices that exhibit solid-like behavior. A majority of clinically available vaginal gels are formed from either a cellulose derivative (e.g., methylcellulose, hydroxyethylcellulose [HEC], carboxymethyl cellulose), a Carbopol^® brand cross-linked poly(acrylic acid), or poly(carbophil). The literature is replete with other examples of natural and synthetic polymers investigated as hydrophilic polymers used in these formulations. Typically, these matrix-forming polymers are classified as “excipients” as they do not contribute to the activity of the drug product. However, as discussed in the following texts, early HIV/STI prevention (“microbicide”) research employed vaginal gels where the gel-forming polymer was also the active agent. There are many claims stating that gelling polymers such as Carbopol and poly(carbophil) can also serve a bioadhesive function to aid in gel retention, but there is not a single convincing paper in the literature showing improved performance of these systems, and thus the concept of “bioadhesive vaginal gels” should be treated with skepticism.

Several additional excipient classes are commonly used in vaginal gels, including gelling agents, humectants, and preservatives. In general, excipients are considered to be pharmacologically and toxicologically inert, but this is not always the case; for example, excipients such as sodium dodecyl sulfate, carrageenans, cellulose acetate phthalate (CAP), and benzalkonium chloride may have antimicrobial properties. Important tests for excipient biocompatibility include cell growth/cell viability assays, cell proliferation assays, and cytotoxicity assays.

Vaginal semisolids are typically applied by the user with little or no instruction from a clinician. Disposable applicators are used to deliver a single dose, typically around 3 or 4 mL, to the proximal vagina. Ideally, this gel dose will spread and coat the vaginal mucosa to result in even drug distribution throughout the tract, although this property is more important in some applications than in others. Gel distribution can occur passively via gravity or through ambulation and even intercourse. Retention in the vaginal tract is also of key importance, both from the standpoint of PK and compliance/acceptability, as excessive gel leakage will result in large drug fractions being lost, as well as being an inconvenience to the user. Several semisolid characteristics affect distribution and retention. First of all, the appropriate volume of gel must be applied, as an excess may result in leakage, but a deficiency may result in inadequate distribution. Most importantly, the bulk viscoelastic properties of the gel formulation (e.g., viscosity, elastic modulus, or storage/loss modulus, depending on the mechanical behavior of the formulation) will influence distribution and retention. Ultimately, these properties are products of the strength and temporal nature of the physical and/or chemical interactions within the polymer. Traditionally, new vaginal semisolids were evaluated comparatively to existing products, but as with vaginal rings, the HIV prevention effort of the past two decades has catalyzed more rigorous study of composition–structure–property–performance relationships of vaginal semisolids.

As with traditional matrix and reservoir IVRs, the mechanical properties of semisolids are intimately coupled to their drug dissolution/release behavior. Vaginal semisolids are perhaps one of the more complicated dosage forms for the mechanistic description of drug release, as the time-dependent size and shape of the formulation confounds the already complex interplay between vaginal diffusion and convection. Changes to the nature of the polymer network are likely to simultaneously affect the viscoelastic behavior of the formulation, as well as the drug diffusivity in the formulation matrix, which in turn affects drug release rates.

Perhaps the most widely utilized and investigated agent in vaginal semisolid delivery systems is the aforementioned N9. An effective spermicide, N9 is included in many over-the-counter, vaginally administered, sodium carboxymethylcellulose-based contraceptives, such as Gynol^® and Conceptrol^®. However, the use of N9-containing contraceptives is considered by some to be controversial, especially in populations at high risk for STI acquisition. This is because some studies have implicated N9, which can weaken the natural epithelial barrier of the vagina, in increased risk of susceptibility to such STI. N9 was one of the first agents investigated as an HIV microbicide and was at one time lauded as a potential cure-all multipurpose prevention agent due to its activity against HIV and other STI. Subsequent HIV microbicide trials investigated vaginal semisolids formulated with nonspecific polymeric agents (e.g., carrageenans, cellulose sulfate) designed to prevent infection. These formulations were determined to be clinically safe for vaginal administration but were categorically unsuccessful in preventing HIV infection, leading to the investigation of antiretroviral-eluting prophylactics as already discussed. Most notably, a HEC-based vaginal gel containing the anti-HIV compound TFV has been both successful and unsuccessful in preventing HIV infection in clinical studies. Mixed results are thought to be due, in part, to low clinical trial adherence. Several ARVs, including the integrase inhibitor raltegravir, are still under preclinical investigation for delivery via vaginal gels. PrEP via a vaginal gel has proven successful in multiple SHIV (recombinant simian HIV) challenge studies in macaques, results that still continue to motivate the investigation of vaginal gels for HIV PrEP and the investigation of HIV PrEP in general.

Although contraception and microbicides have gathered most of the recent attention in regarding vaginal semisolids, several other products are available for a diverse array of indications. For example, semisolids are used for the delivery of antibiotics, such as metronidazole and clindamycin, for the treatment of bacterial vaginosis. Polycarbophil-based vaginal gels containing natural progesterone (i.e., Crinone^®, Prochieve^®) are indicated both for fertility assistance and secondary amenorrhea (Figure 12.4h). Prochieve^® has also been clinically studied for prevention of preterm birth but failed to receive FDA approval following lack of statistical significance in a clinical study. Prostaglandins are administered to induce labor via vaginal semisolids, such as the dinoprostone-releasing, silicon dioxide–based Prostin E2. Vaginal “creams” containing estrogens are available for management of vaginal atrophy and associated symptoms (e.g., vaginal dryness, painful intercourse) in postmenopausal women; however, there is increasing evidence that chronic dosing of unopposed estrogen (without a progestin) carries an increased risk of a variety of serious medical outcomes including endometrial and breast cancer, cardiovascular events (e.g., heart attack, deep venous thrombosis), and dementia.

Various vaginal gel formulations have been investigated for insulin delivery, but it has been difficult to achieve the sufficiently high systemic insulin levels necessary to warrant clinical development. An interesting platform under clinical investigation is a vaginal gel containing the dendrimer SPL7013, which has demonstrated in vitro inhibition of HIV, herpes simplex virus (HSV), and bacterial replication. This dendrimer has shown clinical efficacy in treating bacterial vaginosis and is also being considered for use as a condom coating.

Vaginal creams are employed in the treatment of vaginal intraepithelial neoplasia (VAIN), a premalignant lesion involving cells in the vaginal epithelium, thought to be associated with human papillomavirus (HPV) infection. Topical delivery is advantageous to avoid systemic exposure to compounds used in the treatment of VAIN, which are associated with a wide array of side effects. 5-fluorouracil is formulated in a topical cream for the treatment of diffuse VAIN, though it has been associated with ulceration and poor tolerance. It is also used in conjunction with surgery for vaginal cancer, both pre- and postoperatively. Aldara^™ (Imiquimod) is another topical cream for treatment of diffuse VAIN. Unlike 5-fluorouracil, Imiquimod is an immune response modulator that recruits cytokines with antiviral and tumoricidal effects.

12.6.6 ERODIBLE VAGINAL SOLIDS (TABLETS, SUPPOSITORIES, INSERTS, AND FILMS)

Thus far, we have discussed nonerodible solid devices (IUS/IUD, IVR, and cervical barriers) and semisolids (gels and creams). Between these two categories lie erodible/dissolvable solid formulations, which include vaginal tablets and suppositories, and the emerging vaginal film platform. In theory, vaginal solids are not fundamentally different from oral solid dosage forms, although the slow convection of the FRT (in comparison to the GI tract) lends the possibility for sustained-release tablets with high residence times, including the incorporation of bioadhesives to enhance retention, as with vaginal gels.

Solid vaginal tablets and suppositories are available for a variety of indications, somewhat mirroring the list earlier for vaginal semisolids. The majority of these formulations are designed for delivery of antifungals and/or antibiotics, including the Candizole-T^® tablet, which is a fixed-dose combination of two antifungals (miconazole and tinidazole) and an antibiotic (neomycin). Vaginal solids are also employed for vaginal delivery of progesterone, estradiol, and the spermicide N9 for similar indications as discussed earlier for gels.

There are many reports of off-label vaginal application of oral solid dosage forms in the clinic for various reasons. Many clinical studies have demonstrated the vaginal administration of the abortifacient misoprostol for cervical ripening and labor induction. Also, sildenafil (Viagra^®) has been used in fertility maintenance, and the NSAID indomethacin has been dosed vaginally for prevention of preterm birth. Other oral formulations, including oral contraceptives and hormone therapies, have been dosed vaginally in cases of intolerance to oral medication.

As with many of the platforms discussed thus far, the HIV PrEP effort has spurred innovation within the vaginal tablet platform. The quick-dissolve tablet platform has been adapted for vaginal use and is currently under clinical investigation for delivery of the HIV RT inhibitors TFV and emtricitabine.

Vaginal films have recently emerged as an alternative to vaginal gels. Vaginal films are typically thin, solid polymeric sheets formed from water-soluble polymers and can incorporate drug substances (Figure 12.4f). In contrast to semisolid formulations, films are less messy to apply and provide a formulation strategy for compounds that are unstable in aqueous solution. The concept of thin-film drug delivery is thought to have evolved from consumer products for the oral cavity, particularly the breath mint “patch” (see also Chapter 8, Section 8.5.2.2).

Vaginal films can be designed for rapid disintegration and complete drug release, or to be slow dissolving for more sustained release. The majority of vaginal films currently in use or development are based on poly(vinyl alcohol) (PVA) and hydroxypropylmethyl cellulose, whose mass fraction and molecular weight can control disintegration time. Films can include plasticizers such as glycerol to reduce mechanical rigidity. Films are typically formed by solvent casting but can also be formed by hot-melt extrusion and incorporate thermoplastic polymers. Again, drugs can be dissolved and/or dispersed in the polymer matrix but should, if possible, be completely dissolved in the solvent casting or extrusion phase, to ensure even distribution.

Vaginal drug delivery applications for thin films have already been discussed, such as contraception, treatment of bacterial vaginosis, and HIV/STI prevention. To date, the only vaginal film to market has been the Vaginal Contraceptive Film^®, which is another N9-containing contraceptive product marketed as a hormone-free contraceptive. Films have also been investigated for delivery of itraconazole for candidiasis and clindamycin phosphate for bacterial vaginosis. Rohan et al. have recently published several papers describing the in vitro evaluation of several antiretroviral-releasing vaginal films for HIV PrEP.

12.6.7 VAGINAL “INSERTS”

The term vaginal “insert” is somewhat ambiguous and has been used to describe vaginal solids not readily placed in other categories. For instance, the Endometrin^® “insert” is a progesterone tablet placed in the proximal vagina, near the cervix, with the aid of a specialized applicator. Cervidil^®, also called an “insert,” is a swellable, nondissolving, dinoprostone-releasing slab, composed of a cross-linked poly(ether urethane)-based copolymer. The slab is encased in a knitted polyester bag with a string for retrieval.

12.7 NEW TECHNOLOGIES FOR FRT DELIVERY

In the last few decades, a large body of work has been reported in the area of DDS in the FRT; in particular, vaginally applied concepts are at the cutting edge of this area of study.

12.7.1 MULTIPURPOSE PREVENTION TECHNOLOGIES

Biomedical interventions designed to simultaneously address multiple reproductive health needs are commonly termed “multipurpose prevention technologies” (MPT). MPT can include products that protect against multiple STIs or against unintended pregnancy and an STI. The only commercially available MPT are barrier contraceptives (e.g., male and female condoms, cervical caps), which exhibit reasonable efficacy in providing a barrier to both agents of infection (i.e., viruses and bacteria) and sperm cells, but are prone to misuse and inconsistent use. Specifically, a single product that combines HIV PrEP and contraception has been identified as an urgently needed MPT by the global health community. It has been hypothesized that the inclusion of a contraceptive into an HIV PrEP could bolster demand for unfamiliar PrEP products due to the widespread use of contraceptives.

Conceptually, MPT can take the form of any of the dosage forms already discussed. The first long-acting MPT to be tested clinically is the TFV/LNG IVR, which is a segmented dual-reservoir design that delivers both drugs for up to 90 days. Other MPT IVR are under investigation for delivery of similar combinations, including DPV/LNG (also for HIV prevention and contraception) and a pod-insert style TFV/acyclovir ring (for HIV and HSV-2 prevention). Antiviral drugs, such as DPV and TFV, have been added to barrier contraceptives to form MPT, such as the SILCS diaphragm variants discussed earlier. Antivirals can also be added to an EC to form an on-demand MPT, which could theoretically be administered before or after intercourse. These on-demand MPT could be formulated as any of the semisolid or dissolvable solid dosage forms already discussed.

The Population Council has investigated MPT gels based on the carrageenan family of sulfated polysaccharides. The Carraguard gel was one of the first microbicide products tested clinically and was not effective in preventing HIV infection. However, carrageenans have shown activity against HSV-2 and HPV, which, along with their usefulness as gel-forming polymers, has led to their investigation as MPT. Zinc acetate, which is thought to have activity against HIV and HSV-2, has been formulated in a carrageenan gel, along with the HIV nonnucleoside inhibitor MIV-150, in the hope of an on-demand MPT capable of HIV, HSV-2, and HPV prevention. A MIV-150/zinc acetate/carrageenan (MZC) gel formulation was completely protective against SHIV transmission in macaques, and although the relative contributions to efficacy of MIV-150 and zinc acetate is not clear, a variant without MIV-150 was partially protective, confirming in vivo activity of the zinc salt. In mouse models, zinc acetate/carrageenan gels have demonstrated efficacy against HSV-2 challenge, and carrageenan-only gels have protected against HPV challenge. A MZC IVR is also in development, with the eventual goal of including a contraceptive (e.g., LNG). If such a device was feasible to manufacture and regulatory approval was possible, it could potentially offer long-acting protection against three STIs and pregnancy.

12.7.2 MUCUS-PENETRATING PARTICLES

The efficient trapping of particles by mucus through steric and adhesive interactions is a barrier to delivery to the vagina using nanoparticles. Densely coated PEG-covered PLGA and PS nanoparticles have been demonstrated to have improved vaginal distribution as compared to uncoated nanoparticles. When these mucus-penetrating particles were formulated with an acyclovir monophosphate (ACVp) core in mice, they showed improved protection against HSV-2 challenge (46.7% infected) as compared to soluble ACVp (84.0% infected).

12.7.3 SMALL-INTERFERING RNA AND GENE SILENCING

Small-interfering RNAs (siRNA) are short (21–23 nucleotide) strands capable of inducing RNA interference and gene silencing. Vaginally delivered siRNA-based therapeutics has been delivered using PLGA nanoparticles (100–300 nm) in mice by Woodrow et al. The small size of the nanoparticles was thought to allow for effective tissue penetration and cell entry, and gene silencing was found throughout the uterine horns, cervix, and vagina.

12.7.4 ANTIBODY DELIVERY

Challenges in vaccine delivery to the vaginal environment include the need to overcome degradative enzymatic activity, continual mucus turnover, and a fluctuating microenvironment due to cyclic changes in the menstrual cycle. In order to address these changes through sustained release of antigen, Kuo-Haller et al. created PLGA microparticles and 4 mm diameter discs of EVA, for introduction of the model antigen ovalbumin into mouse reproductive tracts, and obtained equivalent antibody response as compared to vaccines delivered via a different mucosal site.

12.7.5 ELECTROSPUN FIBER MATS

Electrospun fibers have also been studied for vaginal drug delivery. Electrospun fiber meshes can be made out of a variety of synthetic and biological polymers (PLLA/PEO, PEG-PLLA, PVA/PVAc, PLGA, PCL/PGC-C18) and encapsulate a diverse array of drugs, including azidothymidine, ACV, and maraviroc. Erodible solid dosage forms constructed from this platform could circumvent the leakage and messiness associated with vaginal gels, but the advantage over films and tablets is yet unproven.

12.7.6 SEMEN-TRIGGERED SYSTEMS

A number of strategies have been developed for semen-triggered vaginal delivery systems, primarily to prevent infection with HIV. Through triggered delivery of large amounts of anti-HIV drugs, these systems have the potential to deactivate virus before an infection event can occur. Three systems are in development: vaginal meshes, osmotic pump tablets, and microgel particles.

Vaginal meshes represent another novel approach to vaginal drug delivery. Huang et al. developed electrospun CAP fibers loaded with either TMC-125 or TDF. Because the CAP fibers have a pH-dependent solubility, they remain intact in the normally acidic pH environment of the vagina. Upon introduction of semen, the fibers dissolve, releasing the anti-HIV drugs.

Rastogi et al. recently reported a vaginal osmotic pump tablet formed from a hydroxypropylcellulose matrix and a pH-sensitive CAP coating. These tablets function as a reservoir at “normal” vaginal pH (≈4–5), providing sustained drug release to the proximal vagina for up to 10 days, but the outer membrane rapidly dissolves at neutral pH, thus potentially providing a semen-triggered bolus of anti-HIV drugs.

An enzymatically triggered microgel system containing the HIV-1 entry inhibitor sodium poly(styrene-4-sulfonate) (PSS) has also been presented. Microgel particles containing PSA peptide substrates and PSS were created; upon exposure to seminal plasma, the microgel degraded and released PSS. A pH- and thermosensitive microgel composed of N-isopropylacrylamide, butyl methacrylate, and acrylic acid has also been developed, capable of burst release upon exposure to semen; initial tests for release of acid orange dye and fluorescein isothiocyanate–dextran proved promising (see also Chapter 14, Section 14.5.1).

12.7.7 PERMEABILITY ENHANCERS AND ENZYMATIC INHIBITORS

There are several instances in literature where permeability enhancers have been investigated in vitro and in vivo to enhance the vaginal absorption of large molecular therapeutics, such as peptides and proteins. These include simple organic acids (e.g., citric acid), cationic surfactants (e.g., benzalkonium chloride), cyclodextrins, and even hydrogen peroxide. However, any permeability enhancers, however safe or biocompatible, will likely weaken the epithelial barrier to infection, thus lessening the usefulness of such an approach.

Other studies are based on the use of aminopeptidase inhibitors, including bestatin, leupeptin, pepstatin, sodium glycocholate, aprotinin, and p-chloromercuribenzenesulfonic acid, to enhance the stability of proteins when delivered vaginally. An interesting approach demonstrated that a thiolated variant of carbopol 974P (a common vaginal gel polymer) inhibited aminopeptidase activity against LHRH in vitro. Again, the effect of any such inhibition on the effectiveness of the epithelial barrier must be considered in a drug delivery strategy.

12.7.8 ADDITIONAL FUTURE TECHNOLOGIES

In addition to those described in detail earlier, the following technologies are under investigation at the cutting edge of FRT drug delivery:

• Genetically modified lactobacillus that excrete antigen-binding fragments against STD

• IVR systems that deliver lactobacillus to modify commensal bacteria populations

• IVR that deliver moisturizers to treat vaginal dryness

• Mucosal vaccines that seek to elicit a focused immune response in the genital mucosa

12.8 OTHER SYSTEMS WITH THE FRT AS THE PRIMARY SITE OF ACTION

There are several non-FRT-contacting dosage forms that have the FRT as the primary site of action. One example is the oral PrEP (see earlier), which includes products such as Truvada^®, a pill that combines 300 mg TDF and 200 mg emtricitabine. Other examples include HPV vaccination (Gardasil^®) and antibiotics for treatment of infections including syphilis, cervicitis, gonorrhea infection, chancroid, chlamydia, salpingitis, and genital tuberculosis. Antivirals for HSV-2 (acyclovir, famciclovir, and valacyclovir) are given systemically as well.

Combined oral contraceptives (COC), containing both estrogen and progestin, are a common form of oral contraceptive, with effectiveness based on adherence of scheduled use. COCs work by suppression of both the follicle-stimulating hormone and luteinizing hormone as well as thickening of the cervical mucus and possibly altering tubal transport. Oral progestin-only contraceptive pills work through affecting the cervical mucus and endometrium and are also approximately as effective as COCs. Low-dose COCs and progestin-only contraceptives (oral, injectable, and intrauterine form) may also be used in the treatment of abnormal uterine bleeding.

Women experiencing polycystic ovary syndrome (PCOS) desiring to become pregnant may be treated with the selective estrogen receptor modulator, clomiphene citrate. PCOS may also be treated with injectable human gonadotropins. Abnormal uterine bleeding can also be managed with one of low-dose combination oral contraceptives, cyclic oral progestin, injectable progestin (Depo-Provera^®), parenteral estrogen, and gonadotropin-releasing hormone agonists.

12.9 IN VIVO AND EX VIVO MODELS FOR FRT DRUG DELIVERY

Several animal models are used in the pharmacokinetic, safety, and efficacy testing of FRT dosage forms. However, anatomical and physiological differences between the human FRT and that of these animals should be considered when interpreting any results obtained. A key difference is the vaginal pH, which is neutral or near neutral in all animal models presented here (between 6 and 8), in contrast to the typical acidic pH (discussed earlier) in premenopausal women.

12.9.1 RABBIT

The New Zealand white rabbit (NZWR) model has been used extensively in the preclinical testing of vaginal DDS. Pharmacokinetic evaluations of vaginal dosage forms in NZWR date back to testing of early silicone IVR segments by Chien et al. in the 1970s. These studies also demonstrated a method for establishing empirical correlations between drug release rates from topical devices and blood–plasma concentrations. This practice continues today with the testing of IVR segments for HIV PrEP and contraception. Although such studies can provide valuable insight into the performance of vaginal formulations, care should be taken when interpreting results due to key differences between human and rabbit vaginal anatomy. First, two distinct epithelial types are found in the rabbit vagina. The lower third of the tract contains a stratified squamous epithelium similar to the human tract but the upper two-thirds of the tract, proximal to the urethral opening, are lined with a single layer of ciliated columnar cells, similar to those found in the human (and rabbit) endocervix. This difference is likely to affect the vaginal absorption and distribution of drugs, but such an effect has not been quantified.

NZWR are also used as a standard model for testing the safety of vaginally applied products, including drug formulations/drug-containing devices. A rabbit vaginal irritation test is specified in the ISO 10993 (“Biological Evaluation of Medical Devices”). The irritation test score is a composite of four histological examinations of necropsied vaginal tissue: epithelial integrity, leukocyte infiltration, vascular congestion, and edema.

NZWR have also been used to demonstrate the efficacy of vaginally administered contraceptives. Another interesting application of NZWR is the use of cutaneous transmission of cottontail rabbit papillomavirus as a model for vaginal HPV transmission.

12.9.2 NONHUMAN PRIMATES

NHP, most commonly pig-tailed and rhesus macaques, are used as a model for PK and efficacy testing for HIV-prevention formulations and devices. As HIV only infects humans, there is no way to demonstrate the prevention of vaginal transmission of HIV in an animal model. However, in many cases, antiretroviral drugs effective against HIV are also active against some strains of simian immunodeficiency virus (SIV) and/or SHIV. The potential efficacy of HIV PrEP products can be evaluated through vaginal SIV and SHIV challenge in macaques during or after dosing. This practice effectively proved the concept of PrEP through challenge studies in several vaginal gels and IVR, even in the face of several clinical failures now attributed to poor product adherence. For NHP, a reduced-diameter variant of IVR (approximately half of that of a human-sized IVR) must be manufactured, due to the smaller size of the macaque vagina.

12.9.3 SHEEP

The sheep model has become increasingly popular for PK and safety testing of vaginal products. The sheep vaginal tract is similar in size to the human tract allowing for the testing of full-sized nonerodible devices, especially for IVR. Interestingly, the rabbit vaginal irritation evaluation and score, described earlier, have been adapted for safety evaluations of ARV-releasing IVR in sheep.

12.9.4 MICE

Mice are susceptible to multiple STI, including HSV, allowing for the early assessment of the potential efficacy of PrEP products. However, the physiological nature of the murine vaginal mucosa confounds these assessments. The murine estrous cycle is rapid (4–5 days), necessitating hormonal synchronization of the mucosa. The epithelium is keratinized in the estrus phase, making mice less susceptible to infection, but thins in diestrus to more closely resemble the human epithelium. Similar to NHP, dosing mice with progestin can induce diestrus allowing for transmission events to be studied in the context of topical prevention. However, dosing mice with estradiol can induce estrus, wherein the mouse CVF more closely resembles human CVF and becomes useful in the study of drug and particle transport in mucus. Humanized mouse models can be susceptible to HIV infection and have been used in challenge studies to demonstrate potential efficacy.

Human endometrial grafts have also been used to create a model that allows for study of the effects of local progestin delivery, such as from the Mirena^® and Skyla^® IUS, in immunocompromised mice.

A major limitation for any mouse model is the size difference between the human and mouse vagina, especially for large devices such as IVR.

12.9.5 IN VITRO AND EX VIVO MODELS FOR FRT DELIVERY

Several ex vivo and in vitro models are used to evaluate permeability of drugs in the vaginal mucosa. The full porcine vaginal mucosa is often used to assess the permeability of vaginal drug delivery candidates. Drug diffusivity has been observed to be similar, but not always identical between the two tissues, meaning that data obtained can be a good first approximation. However, as with all explanted tissues, the exact thickness of tissue samples can be difficult to determine, hampering the quantitative evaluation of drug permeability. The EpiVaginal^™ series of models are a multilayered in vitro culture of human-derived vaginal/ectocervical cells and dendritic cells that approximates the human vaginal mucosa. EpiVaginal^™ tissues can be grown with a more reproducible thickness in transwells for comparative assessments of drug permeability and formulation toxicity.

To assess the potential efficacy of antiviral formulations ex vivo, macaque tissues can be biopsied following dosing and challenged with SIV or SHIV.

12.10 MAJOR QUESTIONS FOR FUTURE RESEARCH

A number of major questions in the field of FRT drug delivery have yet to be addressed thoroughly and will play an important role moving forward with systems designed to impact women’s health. We include here several areas of investigation that, if addressed, would advance the field:

• The role of CVF in drug delivery

• The importance of mucus in the FRT as an innate immune barrier and a barrier to drug delivery

• The role of epithelial structure in the lower human FRT and how cyclical changes in the epithelium affect drug absorption

• • Development of organ-wide transport models that describe how drugs are transported throughout the FRT

• Development of better tools to understand user perceptions, biases, and desires around FRT-targeted products, to inform the design of better systems with larger impact on women’s health

• Development of more predictive in vitro and animal models, to evaluate the many systems being studied

12.11 CONCLUSIONS

Drug delivery to the FRT has proven useful for several applications, most notably for contraception, hormone replacement, fertility maintenance, and topical treatment and prophylaxis of infection. These interventions have been achieved through a wide variety of dosage forms, including tablets, gels, creams, films, and innovative long-acting devices, such as IVR and IUD/IUS, which have been a major driving force in the evolution of controlled-release science. The major advantages to delivery by this route are the reduction of unnecessarily high drug concentrations when a topical effect is desired, the avoidance of first-pass hepatic metabolism when a systemic effect is desired, and the potential for long-acting prophylaxis and therapy to increase adherence and compliance. Delivery by the FRT also has several drawbacks, most notably the intra- and interuser/patient variability in the vaginal environment, which can induce high variability in PK, and thus confounding accurate assessments of PK, PD, and efficacy. Naturally, it is most logical to investigate this route for indications pertaining to women’s health, regardless of the potential feasibility for other applications.

FURTHER READING

Alexander, N.J., E. Baker, and M. Kaptein. 2004. Why consider vaginal drug administration? Fertil Steril 82(1):1–12.

Chien, Y.W., S.E. Mares, J. Berg et al. 1975. Controlled drug release from polymeric delivery devices. III: In vitro-in vivo correlation for intravaginal release of ethynodiol diacetate from silicone devices in rabbits. J Pharm Sci 64(11):1776–1781.

Clark, J.T., M.R. Clark, N.B. Shelke et al. Engineering a segmented dual-reservoir polyurethane intravaginal ring for simultaneous prevention of HIV transmission and unwanted pregnancy. PLOS ONE 9(3):e88509.

das Neves, J. 2014. Vaginal delivery of biopharmaceuticals. In Mucosal Delivery of Biopharmaceuticals: Biology, Challenges and Strategies, J. das Neves (ed.), pp. 261–280. New York: Springer.

das Neves, J. and M.F. Bahia. 2006. Gels as vaginal drug delivery systems. Int J Pharm 318(1–2):1–14.

Friend, D.R., J.T. Clark, P.F. Kiser et al. 2013. Multipurpose prevention technologies: Products in development. Antiviral Res 100(S):S39–S47.

Garg, S., D. Goldman, M. Krumme et al. 2010. Advances in development, scale-up and manufacturing of microbicide gels, films, and tablets. Antiviral Res 88(S1):S19–S29.

Geonnotti, A.R. and D.F. Katz. 2010. Compartmental transport model of microbicide delivery by an intravaginal ring. J Pharm Sci 99(8):3514–3521.

Gibbs, R.S., B. Karlan, A. Haney et al. 2008. Danforth’s Obstetrics and Gynecology. Philadelphia, PA: Lippincott Williams & Wilkins.

Harwood, B. and D.R. Mishell Jr. 2001. Contraceptive vaginal rings. Semin Reprod Med 19(4):381–390.

Hoffman, A.S. 2008. The origins and evolution of “controlled” drug delivery systems. J Control Release 132(3):153–163.

Kiser, P.F., T.J. Johnson, and J.T. Clark. 2012. State of the art in intravaginal ring technology for topical prophylaxis of HIV infection. AIDS Rev 14(1):62–77.

Kuo, P.Y., J.K. Sherwood, and W.M. Saltzman. 1998. Topical antibody delivery systems produce sustained levels in mucosal tissue and blood. Nat Biotechnol 16(2):163–167.

Malcolm, R.K. 2008. Vaginal rings for controlled-release drug delivery. In Modified-Release Drug Delivery Technology, M.J. Rathbone (ed.), pp. 499–510. New York: Informa Healthcare.

Mishell, D.R. Jr. and M.E. Lumkin. 1970. Contraceptive effect of varying dosages of progestogen in silastic vaginal rings. Fertil Steril 21(2):99–103.

Nel, A., S. Smythe, K. Young et al. 2009. Safety and pharmacokinetics of dapivirine delivery from matrix and reservoir intravaginal rings to HIV-negative women. J Acquir Immune Defic Syndr 51(4):416–423.

Okada, H. and A.M. Hillery. Vaginal drug delivery. In Drug Delivery and Targeting: For Pharmacists and Pharmaceutical Scientists, 1st ed., A.M. Hillery (ed.), pp. 301–328. Boca Raton, FL: CRC Press.

Rivera, R., I. Yacobson, and D. Grimes. 1999. The mechanism of action of hormonal contraceptives and intrauterine contraceptive devices. Am J Obstet Gynecol 181(5):1263–1269.

Saltzman, W.M., J.K. Sherwood, and D.R. Adams. 2000. Long-term vaginal antibody delivery: Delivery systems and biodistribution. Biotechnol Bioeng 67(3):253–264.

Sloane, E. 2002. Biology of Women, 4th ed. New York: Delmar.

Smith, J.M., R. Rastogi, R.S. Teller et al. 2013. Intravaginal ring eluting tenofovir disoproxil fumarate completely protects macaques from multiple vaginal simian-HIV challenges. Proc Natl Acad Sci USA 110(40):16145–16150.

Tortora, G.J. and B.H. Derrickson (eds.). 2013. Principles of Anatomy and Physiology, 14th ed. Hoboken, NJ: Wiley.

van den Heuvel, M.W., A.J.M. van Bragt, A.K. Mohammed et al. 2005. Comparison of ethinylestradiol pharmacokinetics in three hormonal contraceptive formulations: The vaginal ring, the transdermal patch and an oral contraceptive. Contraception 72:168–174.