Secondary Logo

Journal Logo

Comparison of Microbicides for Efficacy in Protecting Mice Against Vaginal Challenge With Herpes Simplex Virus Type 2, Cytotoxicity, Antibacterial Properties, and Sperm Immobilization

MAGUIRE, ROBIN A. MA; BERGMAN, NAOMI BS; DAVID M. PHILLIPS, AND PhD

Articles
Free

Background Currently, a number of different over-the-counter spermicides and potential microbicides under development are in various phases of clinical trials. It is difficult to know how the various formulations would compare with each other or how efficacious they would be because no existing microbicides are commercially available.

Goal To evaluate, in a standardized manner, various parameters of potential microbicides.

Study Design In an effort to make a comprehensive comparison, several potential microbicides and over-the-counter vaginal products were assayed for their efficacy in protecting mice from infection by herpes simplex virus type 2 (HSV-2), for their cytotoxicity to human vaginal epithelial cells, for their effect on the growth rate of L acidophilus, and for their spermicidal activity. Test formulations were K-Y Plus, Gynol II, Advantage S, Replens, BufferGel, No Fertil, Carrageenan, and PC-550. Additionally, several formulations were evaluated for their use as a possible placebo in microbicide clinical trials.

Results The formulations tested fell into three categories of efficacy in protecting mice from HSV-2 infection. The most efficacious were Carraguard and PC-550. All the other test formulations except methyl cellulose afforded varying degrees of protection against herpes simplex virus-2 infection. It was found that formulations containing the surfactant N9 had a cytotoxic effect on human vaginal cells, inhibited the growth rate of L acidophilus, and exhibited spermicidal activity. In addition, it was found that Replens, BufferGel, No Fertil, and the Carbopol formulation might have some effect on sperm motility. Also, K-Y Jelly significantly inhibited the growth rate of L acidophilus.

Conclusion Evaluating formulations under the same testing conditions can help to distinguish among potential formulations that are likely to show promise as safe and effective microbicides.

From the Population Council, New York, New York

BufferGel was supplied by Dr. Kevin Whaley, No Fertil was supplied by Dr. Sheldon Segal, and VK2 was supplied by Dr. Deborah Anderson.

The authors have no proprietary interest in any of the above formulations.

Correspondence: David M. Phillips, PhD, Population Council, 1230 York Avenue, New York, NY 10021. E-mail: dphillips@popcouncil.org

Received for publication May 25, 2000, revised August 17, 2000, and accepted September 13, 2000.

THE CONCEPT OF USING VAGINAL SPERMICIDES as vaginal agents against sexually transmitted pathogens has been under development for nearly 30 years. Since the discovery of HIV, the effort has intensified, particularly because the active ingredient in most over-the-counter (OTC) spermicides is the surfactant, nonoxynol 9 (N9), which has been shown to inactivate HIV. 1,2 The term “microbicide” was coined early on because the active ingredient has a “cidal” effect on microorganisms. Microbicide is the term now applied to all topical products that would be used to prevent sexually transmitted infections (STIs), although the active ingredient may prevent infection by a mechanism other than inactivating or killing the pathogen. In fact, a number of potential microbicides currently under development have an active ingredient that does not kill pathogens, but rather prevents pathogens from adhering to target cells or a blocking postadherence process such as HIV replication.

Because there are no existing microbicide products on the market, determining the relative effectiveness of a potential microbicide formulation has been problematic, particularly evaluating its efficacy against HIV infection. Unfortunately, no standard screening assays exist to test the efficacy of a potential microbicide. Nor are there commercial laboratories that carry out microbicide screening assays. Rather, each group working on the development of a microbicide uses different types of in vitro assays and in vivo animal systems. Considerable variation exists among the assays, and there are methodological differences among similar assays when used by different laboratories. Therefore, it is difficult to compare products tested under different conditions and by different laboratories.

One approach to evaluating the relative effectiveness of potential microbicides is to test the various candidate formulations in the same laboratory assays under the same testing conditions by the same personnel and for the same variety of parameters. The authors have used a herpes simplex virus type 2 (HSV-2)–mouse system to compare various formulations for their relative effectiveness in blocking HSV-2 viral infection. It has been shown that laboratory mice are susceptible to human HSV-2 via the vaginal route. 3 The HSV-2 mouse assay has a number of attractive features. Mice are the least expensive laboratory mammals. When infected by HSV-2, mice experience redness, inflammation, hair loss and lesions in the genital region, 95% of which develop within the first 2 weeks after the challenge, making infection easy to determine in a relative short period. In addition, results may be relevant to HIV infection because HSV-2 and HIV are enveloped viruses. Many of the active ingredients in potential microbicides, including surfactants and sulfated polysaccharides, act nonspecifically on the viral envelope.

This article reports a comparison of various potential microbicides, including several formulations currently in clinical studies, for efficacy in protecting mice against HSV-2 infection. Because vaginal irritation is a major concern because of its potential to increase the incidence of STIs, 4 the formulations have been evaluated for their level of cytotoxicity to human vaginal epithelial cells. In addition, the formulations have been evaluated for their impact on the normal vaginal environment and their effect on sperm motility.

Back to Top | Article Outline

Materials and Methods

Test Formulations

The formulations tested included four vaginal products purchased at a local pharmacy (K-Y Plus, an OTC vaginal spermicide that contains 2.2% N9; Gynol II, an OTC vaginal spermicide that contains 2% N9; Replens, a Carbopol-based OTC vaginal lubricant; Advantage S, an OTC vaginal spermicide that uses Replens as a vehicle base with the addition of 3.5% N9); four potential microbicides (BufferGel, a strongly buffered pH 4 Carbopol formulation currently in clinical studies; No Fertil, a gossypol formulation currently in clinical studies for evaluation as a potential contraceptive with possible microbicidal activity; Carraguard, a sulfated polysaccharide formulation containing 3% PDR98 to 15 λ-carrageenan [FMC, Rockville, ME] currently in clinical studies; PC-550 Carraguard containing 2% N9 [Tergitol NP-9, Sigma, St. Louis, MO]); and three potential placebo formulations (K-Y Jelly, an OTC product recommended for use as a vaginal lubricant [purchased at a local pharmacy]; a 2.5% Carbopol formulation [Carbopol 971P-NF; BF Goodrich, Cleveland, OH], a bioadhesive polymer reported to facilitate retention of active ingredients on mucosal surfaces and prepared by slow dispersion of 2.5 g of polymer through a Nytex mesh into 100 ml of distilled water vigorously stirred at room temperature for 3 hours; a 2.5% methyl cellulose formulation, previously described 5).

Back to Top | Article Outline

Propagating Virus

Methods for propagating and titering HSV-2 have been previously described. 6

Back to Top | Article Outline

Viral Challenge

At the Rockefeller University Laboratory Animal Research Center, 6- to 8-week-old female BALB/c mice were maintained on a 12-hour light cycle. The mice were injected subcutaneously with 0.1 ml of medroxyprogesterone acetate (Depo-Provera, Upjohn, Kalamazoo, MI), 25 mg/ml in phosphate-buffered saline (PBS). Then, 5 days later, 10 μl of a test formulation was carefully instilled into the vagina using a P20 “Pipetman” (Rainin, Woburn, MA). The end of the pipette tip was clipped off to a diameter of approximately 1 mm for vaginal delivery of viscous formulations.

Approximately 10 minutes after pretreatment with the test formulation, the mice were challenged intravaginally with 10 μl of HSV-2. The mice were monitored for 2 weeks after the challenge. They were scored as uninfected or infected depending on the presence or absence of redness, inflammation, hair loss, or lesions in the genital region. At the first sign of infection, the mice were killed. All test formulations were evaluated in a group of 20 mice.

Back to Top | Article Outline

Cytotoxicity Assay

The cytotoxicity assay was performed as described previously, 7 with the following modifications. Instead of ME 180 cells, the immortalized cell line, VK2 derived from human vaginal cells were used. The VK2 cells were grown in keratinocyte basal medium (Biowhittaker, Walkersville, MD) according to the methods described in Fichorova et al. 8 and seeded at 5 × 104 cells per well on 96-well flat-bottom culture plates (Becton Dickinson, Rutherford, NJ). Confluent cells were treated with twofold serial dilutions (1–0.03 mg/ml) of test formulations, and with similar dilutions of 0.1% N9 solutions (in physiologic saline) used as a positive control.

Back to Top | Article Outline

Lactobacillus Assay

Lactobacillus acidophilus ATCC#4356 (American Type Culture Collection, Rockville, MD) bacteria were cultured in MRS Lactobacilli Broth (DIFCO Laboratories, Detroit, MI) in 16-ml screw-cap test tubes maintained at 37 °C and 5% CO2, where they were passaged and incubated for 24 hours before each experiment. Two to four 1.5-ml aliquots were removed from stock cultures and centrifuged at 5000 g for 10 minutes. The broth was aspirated. The pellets, resuspended in sterile PBS and centrifuged at 5000 g for 10 minutes (three times), were pooled together in sterile PBS. Their optical density (OD) value was measured at 540 nm. The suspensions were diluted with PBS to obtain an OD of 1, which represents a titer of 2.86 × 108 cfu.

In 16-ml screw-cap test tubes, 5-ml dilutions of the different test formulations were prepared in minimal medium: 40% Lactobacilli MRS Broth and 60% Minimal Broth Davis without dextrose (DIFCO Laboratories, Detroit, MI). Thereafter, 75 μl of the bacterial suspension were added and mixed with each dilution. The test tubes were incubated for 20 hours at 37°C and 5% CO2. The OD measurements were taken at 540 nm immediately after the bacteria suspension was added and mixed with the test formulation dilutions and then after 20 hours of incubation. The OD values are the mean of six wells ± SD.

Back to Top | Article Outline

Spermicidal Activity

Spermicidal activity was performed essentially as described by Sander and Cramer. 9 The formulations were evaluated using serial twofold dilutions in physiologic saline and a normal semen sample. The end point was the total immobilization of spermatozoa after a 20-second incubation. The results are expressed as minimum effective concentration (MEC), as defined by Sander and Cramer. 9 The MEC for a given compound is obtained by the highest dilution that displays complete immobilization. Compounds that present a MEC of 2 mg/ml or more are considered to have significant spermatozoa immobilizing activity and regarded as a spermicidal agent. In addition to the aforementioned test formulations, an additional dilute formulation of 0.5% Carbopol was evaluated because of the unusual results observed in formulations containing Carbopol. Serial dilutions of a 1% N9 in physiologic saline solution were used as a positive control. Experiments were repeated twice.

Back to Top | Article Outline

Results

Efficacy in Protecting Mice From Herpes Simplex Virus Type 2 Infection

All the formulations were evaluated by using four different viral doses: 103, 104, 105, 106 pfu. In control animals, a viral dose of 103 pfu resulted in a 50% infection rate (Figure 1) and a dose of 104 in a 95% infection rate (Figure 2) when PBS was used before the viral challenge.

Fig. 1

Fig. 1

Fig. 2

Fig. 2

In terms of protecting animals from infection, the formulations tested fell into one of three categories. The formulations in the first category were K-Y Jelly, 2.5% Carbopol, and 2.5% methyl cellulose (Figure 1). In this category, none of the formulations showed significant efficacy in protecting against HSV-2 infection, as compared with the use of PBS. However, the K-Y Jelly and 2.5% Carbopol formulations appeared to afford slight protection at a viral dose of 103 pfu, whereas methyl cellulose was the same as PBS.

The second category included the OTC spermicides K-Y Plus, Gynol II and Advantage S, the OTC vaginal lubricant Replens, and the potential microbicides BufferGel and No Fertil. This group exhibited approximately an 80% protection rate against HSV-2 infection when the animals were challenged with 104 pfu, a viral dose that would infect all of the PBS (control) animals (Figure 2). At a viral dose of 105 pfu (10 × 100% infection dose), the protection rate dropped, approximating 15% to 45% (Figure 3), and at a viral dose of 106 pfu (100 × 100% infection dose), no animals were protected (Figure 4).

Fig. 3

Fig. 3

Fig. 4

Fig. 4

The carrageenan formulations, Carraguard and PC-550, comprised the most protective category. Both formulations exhibited a 100% protection rate against HSV-2 infection when the animals were challenged with a viral dose that would infect all the animals (Figure 2). At 105 pfu (10 × 100% infection dose), there was only a slight reduction in the protection to 85% for Carraguard and 90% for PC-550;Figure 3). At 106 pfu (100 x 100% infection dose), Carraguard protected at 35% and PC-550 at 50% (Figure 4).

Back to Top | Article Outline

Cytotoxicity Assay

Serial dilutions of test formulations were evaluated to determine a dose response (Figure 5). The concentration of the formulation that caused a 50% reduction in BDECF-AM fluorescence was determined from the dose–response curve. The test formulations Replens, BufferGel, Carraguard, K-Y Jelly, Carbopol, and methyl cellulose did not exhibit a cytotoxic effect over the concentration range. The remaining test formulations, K-Y Plus, Gynol II, Advantage S, No Fertil, and PC-550, exhibited a 50% cytotoxic effect in the concentration range of 250 to 600 μg/ml (Table 1). All of these formulations, with the exception of No Fertil, contain varying concentrations of N9. The N9 control dilutions exhibited more than a 50% cytotoxicity over the concentration range.

Fig. 5

Fig. 5

Table 1

Table 1

Back to Top | Article Outline

Effect of Formulations on the Growth Rate of Lactobacillus Acidophilus

Lactobacillus acidophilus helps to maintain a healthy vaginal environment by producing lactic acid and hydrogen peroxide. 10,11 To evaluate whether any of the formulations would have an inhibitory or enhancing effect on the growth rate of L acidophilus, bacterial cultures were grown in minimal media in the presence of test formulations. In minimal media, the growth rate of lactobacilli is slower than in 100% MRS Broth. Therefore, both enhancement and inhibition of growth can be measured.

Onefold serial dilutions of the formulations were evaluated to determine a dose effect (Figure 6). The formulations fell into two categories. The first category included Replens, No Fertil, Carraguard, and methyl cellulose. In this group, the formulations had neither an inhibitory nor an enhancing effect on the growth rate of Lactobacillus (Table 2).

Fig. 6

Fig. 6

Table 2

Table 2

The second category consisted of K-Y Plus, Gynol II, Advantage S BufferGel, PC-550, K-Y Jelly, and Carbopol. This group exhibited a 50% inhibition of the growth rate at formulation concentrations of 1% to 25%, and a 100% inhibition of the growth rate over a wider range of formulation concentrations: 5% to 50%. K-Y Jelly had the most significant impact on bacterium growth, with total inhibition at a concentration of 5% (Table 2). None of the formulations studied, enhanced the growth of L acidophilus.

Back to Top | Article Outline

Effect on Sperm Motility

The effects that formulations may have on sperm motility were evaluated by the Sander–Cramer assay. 9 The formulations fell into one of three categories. The first category included Carraguard, K-Y Jelly, and methyl cellulose. The formulations in this group had no effect on the immobilization of spermatozoa.

The second category included K-Y Plus, Gynol II, Advantage S, No Fertil, PC-550, and N9 (control) solutions. All of these formulations effectively impeded sperm motility. However, No Fertil had an MEC value of 2.5 mg/ml, which is slightly higher than the acceptable cutoff for consideration as a spermicidal agent according to Sander and Cramer 9 (Table 3).

Table 3

Table 3

The third category included the Carbopol-containing formulations Replens, BufferGel, and 2.5% Carbopol. Although these formulations did not kill spermatozoa, they hindered progressive mobility. Sperm tail movement was observed, but the spermatozoa remained in place. Because these results appeared to be a phenomenon of Carbopol, the common ingredient in the four formulations, a 0.5% Carbopol formulation also was evaluated and found to produce the same phenomenon at a dilution of 0.015 mg/ml, the lowest concentration evaluated.

Back to Top | Article Outline

Discussion

In the effort to evaluate the basic properties of a potential microbicide in a comparative manner, 11 formulations were evaluated for their efficacy in protecting mice from infection by HSV-2, for any cytotoxic effect they may have on human vaginal cells, for any impact on they might have on the growth rate of L acidophilus, and for their effect on sperm motility.

Although it may not be reasonable to equate the degree of efficacy in blocking HSV-2 infection in mice with that of blocking HIV infection in people, the authors suggest that a reasonable correlation between the two is likely. Both HSV-2 and HIV are enveloped viruses. It is known that the anionic surfactant N9 and sulfated polysaccharides (e.g., carrageenan) work by acting on the envelope of viruses. 12–15 This observation might suggest that they act nonspecifically.

To compare formulations and gain a better understanding concerning their degree of efficacy in blocking HSV-2 infection in mice, 10-fold increasing amounts of viral challenge were used. The HSV-2 mouse assays performed by other laboratories have been limited to viral infection doses of 50% and 100%. 3 At these challenge doses, many formulations have been shown to have significant effect in protecting mice from infection. However, this does not indicate their true strength or limit of efficacy. Thus, there is a need to increase the viral dose to establish a level at which formulations can be compared.

With the exception of K-Y Jelly, Carbopol, and methyl cellulose, all the formulations significantly blocked infection when the 100% infection dose was used. However, when the viral dose was increased 10-fold, the rate of protection dropped off dramatically, except for that of Carraguard and PC-550, which protected most of the mice at this dose. At a viral dose 100 times the 100% infection rate, the only formulations that showed protection were Carraguard and PC-550, with PC-550 proving to be more efficacious, as would be anticipated if both carrageenan and N9 were active ingredients and working in an additive manner.

It was impressive that the non–N9-containing formulations, Replens, BufferGel, and Carraguard, were as effective or more effective than the formulations containing N9 and were found not to be cytotoxic. Replens and BufferGel are Carbopol-containing formulations. It is interesting also that the OTC spermicide Advantage S is formulated by the addition of 3.5% N9 to the vaginal lubricant Replens, and that Advantage S exhibited no increase in efficacy against HSV-2 infection. This observation may confound the results of the current Joint United Nations Program on AIDS (UNAIDS)-supported COL-1492 efficacy studies, in which Replens is being used as the “placebo.”

Because healthy women conceivably will use a microbicide frequently over long periods of time during their sexually active years, it is critical that a product not be harmful to the reproductive tract epithelium. It has been shown that the use of N9-containing products increases the incidence of urinary tract infections, epithelium irritation, and Candida.16 The non-N9 formulations, Replens, BufferGel, and Carraguard, afforded protection against HSV-2 infection without cytotoxic effects on the human vaginal VK2 cells. An interesting finding in the cytotoxicity assay was that formulations containing N9 were less toxic than equivalent concentrations of N9 mixed in physiologic saline, suggesting that a “carrier” ingredient in the formulation may mediate the cytotoxicity of N9.

Another critical parameter of a potential microbicide is not to alter the normal vaginal flora. Any enhancement or inhibition of the growth rate for L acidophilus was assessed because it is the most common bacterium to inhabit the vaginal tract, and its presence is needed to maintain a low vaginal pH, which is necessary for a healthy vaginal environment. 10,11 The test formulations that contained N9 and K-Y Jelly inhibited the growth rate of lactobacilli. The inhibitory effect of K-Y Jelly most likely results from the effect of the preservative, chlorhexidine gluconate. (This preservative is not in K-Y Plus.) Because it has been shown that the effect of K-Y Jelly on the growth rate of lactobacillus is considerably more inhibitory than the N9 test formulations, it is possible that regular use of this product as a vaginal lubricant could have a deleterious effect on the vaginal environment.

Ideally, microbicides should be available in two forms: those that also are contraceptives and those that are not. Women may wish to conceive but yet desire to be protected against HIV and other sexually transmitted infections. Test formulations that result in an MEC value greater than 2 mg/ml are considered, under the criteria of the Sander-Cramer assay, not to have sufficient sperm immobilizing activity, and therefore cannot be regarded as a spermicidal agent. The N9-containing formulations had MEC values in the range of 0.4 to 0.6 mg/ml, which is well within the range that allows them to be considered spermicidal agents. No Fertil had a MEC value of 2.5 mg/ml, which indicates poor immobilizing activity and suggests debatable contraceptive efficacy. The finding that Replens, BufferGel, and Carbopol did not totally immobilize sperm but hindered forward progression may suggest some contraceptive activity. However, these formulations would be significantly less effective than OTC spermicides. The degree to which they would afford contraceptive properties would need to be explored further.

Because the active ingredient, carrageenan, in Carraguard and PC-550 is also the formulation’s vehicle, it was necessary to evaluate other formulations for their possible use as a placebo in microbicide clinical trials. Of the formulations tested, only the inert compound methyl cellulose did not afford any protection against HSV-2 infection. Additionally, it was not cytotoxic to vaginal cells, did not inhibit or enhance the growth of L acidophilus as observed by KY Jelly and Carbopol, and did not immobilize or hinder forward progression of spermatozoa as observed with Carbopol.

The two carrageenan formulations proved to be the most efficacious in protecting mice against HSV-2 infection, even showing efficacy at a viral challenge of 100 times a 100% infection dose. Additionally, Carraguard was not cytotoxic to vaginal cells, did not inhibit or enhance the growth rate of L acidophilus, and had no impact on sperm motility. Although PC-550 showed cytotoxicity to vaginal cells and inhibited the growth rate of L acidophilus, it required higher concentrations than other N9-containing formulations. Also, PC-550 was shown to have sufficient sperm immobilizing activity. Therefore, microbicides formulated with carrageenan may prove to be more efficacious while being safer.

To make a reliable and comprehensive comparison of microbicide formulations, the authors believe that it is important to evaluate them in a standardized manner. There also is a need to expand this comparison to evaluate other parameters such as the length of time a formulation remains active, the absorption and elimination of the product from the body, and the effect of the formulation on other sexually transmitted pathogens.

Back to Top | Article Outline

References

1. Bird KD. The use of spermicide containing nonoxynol-9 in the prevention of HIV infection. AIDS 1991; 5: 791–796.
2. Doncel GF. Chemical vaginal contraceptives: preclinical evaluation. In: Mauck CK, Cordero M, Gabelnick HL, Spieler JM, Rivera R, eds. Barrier Contraceptives: Current Status and Future Prospects. New York: Wiley-Liss, 1994: 147–162.
3. Whaley KJ, Barratt RA, Zeitlin L, Hoen TE, Cone RA. Nonoxynol-9 protects mice against vaginal transmission of genital herpes infections. J Infect Dis 1993; 168: 1009–1011.
4. Martin HL, Richardson BA, Nyange PM, et al. Vaginal lactobacilli, microbial flora, and risk of human immunodeficiency virus type 1 and sexually transmitted disease acquisition. J Infect Dis 1999; 6: 1863–1868.
5. Maguire RA, Zacharopoulos VR, Phillips DM. Carrageenan-N9 spermicides for preventing pregnancy and sexually transmitted infections. Sex Transm Dis 1998; 25: 494–500.
6. Zacharopoulos VR, Phillips DM. Vaginal formulations of carrageenan protect mice from herpes simplex virus infection. Clin Diag Lab Immunol 1997; 4: 465–468.
7. Pearce-Pratt R, Phillips DM. Studies of adhesion of lymphocytic cells: implications for sexual transmission of human immunodeficiency virus. Biol Reprod 1993; 48: 431–445.
8. Fichorova RN, Rheinwald JG, Anderson DJ. Generation of papillomavirus-immortalized cell lines from normal human ectocervical, endocervical, and vaginal epithelium that maintain expression of tissue-specific differentiation proteins. Biol Reprod 1997; 57: 847–855.
9. Sander FV, Cramer SD. A practical method for testing the spermicidal action of chemical contraceptives. Hum Fertil 1941; 6: 134–153.
10. Hillier SL, Krohn MA, Klebanoff SJ, Eschenbach DA. The relationship of hydrogen peroxide–producing lactobacilli to bacterial vaginosis and genital microflora in pregnant women. Obstet Gynecol 1992; 79: 369–373.
11. Klebanoff SJ, Hillier SL, Eschenbach DA, Waltersdorph AM. Control of the microbial flora of the vagina by H2O2-generating lactobacilli. J Infect Dis 1997; 164: 94–100.
12. Mitsuya H, Looney DJ, Kuno S, Ueno R, Wong-Staal F, Broder S. Dextran sulfate suppression of viruses in the HIV family: inhibition of virion binding to CD4+ cells. Science 1988; 240: 646–649.
13. Schols D, Baba M, Pauwels R, Desmyter J, De Clercq E. Specific interaction of aurintricarboxylic acid with the human immunodeficiency virus/CD4 cell receptor. Proc Natl Acad Sci U S A 1989; 86: 3322–3326.
14. Schols D, Pauwels R, Desmyter J, De Clercq E. Dextran sulfate and other polyanionic anti-HIV compounds specifically interact with the viral gp120 glycoprotein expressed by the T cell persistently infected with HIV-1. Virology 1990; 175: 556–561.
15. Looney DJ, Hayashi S, Nicklas H, et al. Differences in the interaction of HIV-1 and HIV-2 with CD4. J AIDS 1990; 3: 649–657.
16. Roddy RE, Cordero M, Cordero C, Fortney JA. A dosing study of nonoxynol-9 and genital irritation. Int J STD AIDS 1993; 4: 165–170.
© Copyright 2001 American Sexually Transmitted Diseases Association