Key points are not available for this paper at this time.
Introduction The need for woman-controlled barrier contraceptives that protect against both bacterial and viral sexually transmitted pathogens is widely recognized 1–7. In the absence of an effective vaccine or treatment, contraceptive methods capable of preventing sexual transmission of HIV as well as other sexually transmitted diseases (STD) are vital for protecting the health of women. Moreover, widespread violence against women, double standards of sexual behavior, and the imbalance of power in many sexual partnerships make methods initiated and controlled by women critically important. These issues may severely limit existing options for protection among women who cannot negotiate sex with their male partners without being accused of cheating, of being ‘loose’ women, or of accusing their partners of infidelity 3,4. Vaginal microbicides (topical chemical barriers that protect against acquisition of a variety of STD pathogens, including HIV) may provide such alternative woman-controlled methods. Compared to male and female condoms, microbicides are expected to interfere less with intimacy and sexual pleasure, and be more discrete. Because detergents like nonoxynol-9 (N9) are microbicidal as well as spermicidal, several existing N9 contraceptives have been tested in observational and controlled trials as microbicides for HIV/STD prevention 8–17. Modest protection against Chlamydia and gonorrhea has been shown 8–12, but HIV prevention studies 12–17 have yielded mixed results and overall, the protective effect for HIV appears doubtful. In fact, the most recently completed trial 17 reported greater HIV transmission in the women using N9 compared to those using a placebo gel, possibly due to detergent-induced compromise of the epithelial barrier after intensive use. New microbicides (only some of which are spermicidal) are being developed for vaginal protection, in an effort to improve efficacy, safety, and acceptability compared to existing detergent-based products such as N9. Although many of these new microbicides show robust activity against HIV and other STD pathogens, and some also appear to be less toxic than N9, achieving reliable protection with microbicides remains a significant challenge. We contend that the likelihood of success of such products could be greatly increased by an alternative prevention approach, namely the combination of a microbicide and an internal barrier device that protects the cervix. Like condoms, these devices (diaphragms, caps, and other novel designs) create a physical barrier that covers the cervix. Yet because they are worn completely inside the vagina, they avoid the obtrusiveness that limits the acceptability of male and female condoms. With microbicide applied on both the cervical and vaginal sides of these devices (as is commonly recommended for contraception in the UK, but not in the USA or other countries), they should offer all the benefits of the microbicide, with additional benefits provided by physical protection of the cervix. Although internal barrier devices cover the cervix, they do not provide a barrier for most of the vaginal epithelium. Thus, if transmission susceptibility were distributed equally across all epithelial surfaces, internal barrier devices might add only modestly to the protection given by the microbicides with which they were used. However, substantial epidemiological and biological evidence suggests that susceptibility is not evenly distributed, but that the cervix is a site of particularly high susceptibility to HIV and STD transmission. Thus, internal barrier devices that cover the cervix may enhance significantly the protection against HIV and STD that may be provided by microbicides alone. In addition, applying the microbicide to the vaginal side of the barrier may confer vaginal protection as well. Currently the traditional diaphragm and cervical cap are the only tested and approved internal devices that provide physical protection of the cervix. However, several new barrier methods are under development or at various stages of testing. These method include the Leah's shield (similar to a loose fitting cervical cap made of rubber with a loop for easy removal), the Femcap (also similar to the cervical cap but with a brim designed to fit into the vaginal fornices), the SILCS diaphragm a new single-size design (SILCS Inc., Tinton Falls, New Jersey, USA) expected to be easier to insert and remove, and disposable diaphragms (some of which may be provided with microbicide preapplied). Evidence for the importance of the cervix in acquisition of STD and HIV Cervical infection with bacterial STD STD facilitate transmission of HIV both by increasing susceptibility among the uninfected, and by increasing the infectiousness among those infected with HIV 18. Thus, the role of the cervix in STD susceptibility, and the importance of cervical protection in blocking acquisition of STD, are both relevant to preventing transmission of HIV. Cervicitis is the classic and predominant manifestation of Chlamydia and gonorrhea infection. Both organisms replicate in the cervical columnar epithelium, and are incapable of replication in the squamous epithelium of the vagina. Although these pathogens also infect Bartholin's glands and the urethra, infections at these sites are generally believed to be complications of established cervicitis 19,20. Epidemiological evidence of the protective effect of physical coverage of the cervix supports these observations. Although to date, there have been no experimental studies (i.e., controlled trials) to evaluate the effect of diaphragm use and STD acquisition, there have been several observational studies (case–control or cross-sectional designs) that report a protective effect of diaphragms in decreasing susceptibility to STD and associated long-term sequelae. All of the studies compared diaphragm users to non-users, and all used some type of multivariate analysis to control for known co-factors or confounders such as socioeconomic status or age. Although not always specified, in most studies women who used diaphragms used them together with spermicides. Thus, although we cannot separate the protective effect of diaphragms from that conferred by spermicides used alone, this limitation does not affect our fundamental argument that diaphragms used together with microbicides may offer significant protection. Table 1 summarizes these results. Because the majority of these studies were not designed to test the efficacy of the diaphragm as their primary objective, and (as stated above), because they are all observational studies and thus subject to biases inherent in that design, results in the table must be seen as suggestive rather than definitive.Table 1: Observational studies reporting the association between diaphragm use and STD (Diaphragm use versus all other methods). Two cross-sectional studies confirmed a protective effect for diaphragm use among women seen at a STD clinic. Magder et al.9 reported than none of 77 diaphragm users had gonorrhea compared to 20% among non-users odds ratio (OR), 0.8; Rosenberg et al.10 also reported a reduced risk of gonorrhea among diaphragm users OR, 0.32; 95% confidence interval (CI), 0.16–0.05. These results were confirmed in a case–control study conducted by Austin et al.11 comparing women with gonorrhea to other STD clinic attendees without a current infection. Compared to non-users, women who used diaphragms were 55% less likely to have gonorrhea (OR, 0.45; 95% CI, 0.15–1.3). Data are more limited regarding the protective effect of diaphragm use against other STD pathogens. In the cross-sectional study mentioned above, Rosenberg et al.10 were also able to consider other STD and detected an OR for diaphragm users of 0.24 (95% CI, 0.12–0.48) for trichomoniasis and 0.25 (95% CI, 0.05–1.36) for Chlamydia. Human papillomaviruses can infect the cervix, vagina, and vulva, but cancer risk is predominantly a result of cervical infection 21. Becker et al.22 in a case–control study, detected a decreased risk for cervical intraepithelial neoplasia (CIN) II or CIN III among users of diaphragms (OR, 0.3; 95% CI, 0.1–0.8). Two case–control studies have examined sequelae of STD, specifically pelvic inflammatory disease (PID), as an outcome. Wolner-Hanssen et al.23, comparing women with PID to uninfected women seen at an STD clinic, reported an OR of 0.3 (95% CI, 0.09–0.75) for diaphragm users, and Keleghan et al.24, studying women hospitalized for PID compared to women hospitalized for other reasons, detected a 60% decrease in risk among diaphragm users (OR, 0.4; 95% CI, 0.2–0.7) controlling for potentially confounding variables. The susceptibility of the cervix to HIV To date, no studies have examined the protective effect of physical coverage of the cervix and HIV acquisition. However, because of its fragility, frequent compromise by classical STD, and the presence of HIV receptor sites (all of which are discussed below), the cervix is probably more susceptible to HIV than is the vaginal tissue. The importance of the cervix in acquisition of HIV infection is suggested by a recent experiment in which rhesus macaques were infected vaginally with SIV 25. Using in situ hybridization to detect SIV-infected cells, the first cellular targets were found to be located in the lamina propria of the columnar endocervical epithelium. These cervical cells were detectably infected by day 3, whereas the vaginal mucosa was not infected until day 12, a time when virus was systemically disseminated. Thus, the cervix appeared to be the site of initial infectious entry. The cervix may also serve as a portal allowing pathogen access to the upper genital tract. Human cervical tissue section explants are easily infectable with HIV, as are uterine and fallopian tube sections 26. This suggests that upper tract access may be followed by infectious entry of HIV. As is apparent from these results and those reported for STD above, overall, there is a consistent indication that the cervix is an important infection site for STD and HIV. Below we review biological mechanisms that may account for these observations. Biological plausibility: mechanisms for cervical susceptibility to STD and HIV Cervical infection with viral STD (HSV, SIV, and HIV) Herpes simplex virus (HSV) is known to affect the cervix, vagina, and external genital skin 27. But whereas HSV infects surface columnar epithelial cells of the cervix directly, the susceptible cells of the vagina and skin reside in the deeper parabasilar and intermediate epithelial layers 27. Thus, microtrauma may be required to provide access to squamous but not to cervical columnar target cells. In this sense the cervix may be a site of higher susceptibility than vaginal or genital skin, and on theoretical grounds, a cervical barrier might be expected to be protective. However, no studies have been published examining the effect of diaphragms or caps on HSV transmission. Experiments with the SIV/macaque vaginal transmission model show that the normal genital tract is a substantial barrier to transmission. Ten thousand times more SIV must be inoculated vaginally to achieve reliable transmission than the amount required when the virus is injected into the blood stream 28. This suggests that intact genital epithelium is a potent barrier to transmission, particularly as semen from HIV-infected men contains far less infectious virus than used in these experiments 29,30. These facts are consistent with clinical and epidemiological observations that disruption of the epithelium is associated with enhanced HIV acquisition. For example, both genital ulcer disease 31,32 and trauma (marked by post-coital bleeding) 33,34 have been associated with increased HIV risk. Likewise, non-ulcerative, but inflammatory STD gonorrhea, Chlamydia, trichomoniasis (as discussed above), and bacterial vaginosis have also been associated with increased HIV susceptibility. (Bacterial vaginosis, although historically considered a non- inflammatory disease, is characterized by increased cytokine levels in cervicovaginal secretions 35.) These often untreated and highly prevalent disruptions of natural epithelial protective mechanisms are thought, in part, to be responsible for the disastrous pace of the AIDS epidemic in regions such as sub-Saharan Africa 36. Although many of these conditions could affect both the cervical and vaginal epithelia, the cervix may be more susceptible than the vagina for reasons reviewed below. The cervical epithelium is an easily compromised barrier The relative fragility of the cervix as compared to the vagina provides evidence that the cervix is a likely site of entry for STD pathogens and HIV. The cervical columnar epithelium extending from the endocervical canal out to the transitional zone is much thinner than vaginal epithelium. Throughout this region the epithelium consists of only a single layer of columnar cells. It is consequently more easily damaged than the thicker (30–45 cells thick) stratified squamous epithelium of the vagina 37. This cervical fragility is evident during pelvic examination, where the cervix may be friable (bleeding easily after gentle contact with a cotton swab), particularly when cervical ectopy is present. Ectopy is a common physiological condition in which the columnar epithelium extends well out onto the face of the cervix, and is thereby exposed to trauma from intercourse and contacts pathogens in semen. Ectopy has been reported to be strongly associated with HIV infection (OR, 5.0; P = 0.007) 38. Ectopy is particularly common in adolescents. This biological phenomenon, in addition to other behavioral factors, probably contributes to increased risk for HIV and other STD 39 among this age group. Oral contraceptives have also been associated with ectopy, and are also associated with erythema and edema of the zone of ectopy, thereby increasing mucosal fragility further and thus probably increasing susceptibility to a range of pathogens 40. Bleeding ectopy (friability) is very common in some populations, for example, a prevalence of 26% was reported in a study of 257 consecutive women seen at a maternal and child health clinic in India 41. In contrast, vaginal epithelium is rarely friable except in pathological inflammatory conditions, such as frankly ulcerative STD. Thus, damage to mucosal epithelia, whether traumatic or infectious, may enhance STD/HIV transmission by compromising what is otherwise a significant barrier against infection and providing access to deeper cells. Because the cervix is likely more susceptible to damage, a high priority should be placed on protecting it both from traumatic damage and from direct exposure to pathogens. Uterine peristalsis rapidly exposes the upper reproductive tract by aspirating vaginal fluids The tissues of the upper genital tract are susceptible to classical STD 42 and have recently been shown to be susceptible to HIV also 26. Although the endocervical mucus ‘plug’ has generally been assumed to be a substantial barrier to the uptake of vaginal fluids into the upper genital tract, this view is no longer tenable. Many investigators have reported that sperm and immotile particles deposited in the vagina are rapidly transported to the fallopian tubes and peritoneal cavity 43,44. Over the past decade extensive sonographic evidence generated by multiple investigators has shown that the uterus continuously predominantly from cervix to and in and at observations show that these are aspirating fluids out of the vagina after vaginal of a a of being placed in the vagina is into the fallopian tube that the side on which had that uterine of vaginal is a highly physiological experiments have confirmed these by sonographic of uterine of deposited sonographic Moreover, peristalsis has been to fluids all the to the which is by and targets for HIV. This of deposited in the vagina may be an important not only in sperm to enhance but also in STD pathogens such as Chlamydia, and HIV to the upper genital tract, and by the uterus the time for a microbicide to with and pathogens in semen or against the cervix. Thus, protecting the cervix and upper genital tract may be using microbicides at blocking pathogens only in the vagina. receptor sites are on the surface of the cervix The columnar cervical epithelium also in HIV acquisition. cells are susceptible to HIV, and may also as cellular show that cells are rarely found in the vaginal but are easily detected in the endocervical and are also on the surface of the a receptor that as a for HIV is on cells in the female genital tract. of is much higher in the cervix than in the vagina in with infection or was in the epithelial layer of the vagina, it was found only in tissue In contrast, is the epithelium of the and zone have been to a role in HIV entry into cells that these cells. receptor cervical strongly at the transitional zone columnar and squamous evidence of the importance of the cervix in acquisition of HIV The all that the cervix may be a site for HIV transmission. However, there are also that do not the importance of the cervix in HIV. In the SIV/macaque of the cervix vaginal not decrease the of transmission, that transmission can across the vaginal epithelium Moreover, the of et above, et found SIV-infected cells after infection not only in the cervix, but also in the stratified squamous epithelium of the vagina. cells in the vaginal epithelium are to be important in SIV uptake and transmission in this in with the HIV has been vaginally by women who have had these observations have been to that the cervix must not be a site of increased susceptibility compared to the vaginal epithelium. Yet these studies do not model the increased susceptibility of the cervix that is a likely result of STD or trauma in the but in experiments with HIV in provide In the an a very of infectious HIV-infected transmitted infection when inoculated into the endocervical canal The of these endocervical has been on the that semen not have access to this site However, as reviewed there is evidence that vaginal fluids are to the upper tract. Likewise, cervical ectopy commonly exposes cervical columnar epithelium to the vaginal Thus, it is that HIV transmission probably can in the vagina probably also at the the majority of transmission may at the cervix and upper tract. the evidence that the cervix is not for transmission, it does not the that the cervix is a site of susceptibility in women. for STD/HIV from contraceptive The contraceptive provides evidence that in cervical barrier devices significantly the contraceptive efficacy of spermicides. The most recent published show that the risk of in women using is 60% higher than the using with a when for and associated with contraceptive and also with likelihood of using Although contraceptive of spermicides used are often due to of consistent and some are to the The spermicides used in most trials more than times the and to all and sperm in an can spermicides of this to It is likely that are by contact times between semen and of the contact between the and cervical mucus that is not with by of semen into the upper genital tract. All of these contraceptive be reduced significantly or by a cervical barrier device that also distributed, and the more The may also be relevant to microbicide to pathogens. and contact time may be and the upper tract may be we use of cervical barriers with microbicides probably result in similar in method for STD/HIV prevention to that well for contraceptive spermicides. All microbicides must be to and distributed in the vagina. Thus, they the use of some of method of is which is with and This is but not because and can the without the to or to other vaginal products the use of an that the is deposited well inside the vagina, the cervix. and caps can be the and of a microbicide the cervix. Moreover, barrier devices can provide the benefits of to both the cervical and vaginal epithelium, and enhanced microbicide We that microbicide higher efficacy for both contraception and disease prevention if it is used with an internal barrier that protects the cervix. The microbicide should not only be applied on the cervical side as has been traditional for contraceptive but also on the vaginal side of the device to with semen and protect the vaginal epithelium. in of its efficacy, as with new method of its efficacy only be as as its which is by of the cervical and female and caps are by some as acceptability because they are used by a very of women. were much more widely but were by contraceptives that higher contraceptive and the of a To date, the studies that have the acceptability of the diaphragm among women to from a range of products from of women the diaphragm on the However, the used to the diaphragm was not and due to of regarding STD and HIV prevention could not be In women who the diaphragm reported a higher of intercourse than women who other methods and and that use was under their control and not male sex as being significantly associated with their studies have been acceptability of internal barrier devices among male However, in a recent contraceptive trial only out of using diaphragms that their could the device during This suggests that acceptability by men probably be and that diaphragms are in use. These well for the of the diaphragm among women. study found that women who diaphragms were and than women who the and were more likely than device users to for as to limited those who of women use after of although that this was in due to the and diaphragm use given by It likely that in the presence of the HIV and STD the of caps, and other methods that may be easier to use if clinical trials that they to protect against The of these devices as methods of control may also to their In addition, the acceptability of the female a device more and than the suggests that internal barrier devices be more if their efficacy were the of or other similar probably be than the of female condoms, and not much more than male as devices for with of As far as we are no published on diaphragm acceptability are from those by However, studies that should provide additional on diaphragm acceptability and use are in and in To our no studies of the HIV of internal barrier devices have been are or are and epidemiological and biological evidence strongly the that a microbicide with such a barrier enhance protection. of this with controlled trials are well and should be a high The for the
Moench et al. (Sat,) studied this question.