Dhondt, Marijke M. M. Pharm; Adriaens, Els Sc, PhD; Remon, Jean-Paul Pharm, PhD
TO DATE, RESEARCH ON novel vaginal formulations is focused on local or systemic delivery of microbicides, spermicides, hormones, antiinfectious agents, proteins, and peptides. Vaginal formulations can be applied frequently over a period of months or years. Frequent use of some vaginal formulations can induce mucosal irritation and damage of the vaginal epithelium, one of the natural protective barriers to disease. 1,2 Vaginal inflammation and ulceration might increase the susceptibility to sexually transmitted pathogens during sexual intercourse. 3–5 Therefore, the safety and vaginal tolerance of newly developed pharmaceutical products are important issues to be addressed by manufacturers and drug-licensing companies. Efficacy studies should then only be conducted with products that have been evaluated for safety and appear to be non-toxic. 6
The importance of an adequate safety assessment can be illustrated by the history of the non-ionic surfactant nonoxynol-9 (N-9). N-9 has been widely available as a contraceptive for many years and has been shown to be effective against a number of sexually transmitted pathogens in laboratory studies. 7,8 Unfortunately, the repeated vaginal application of N-9 in humans has been associated with irritation or disruption of the vaginal and cervical epithelia 1,2 and with induction of an inflammatory response. 2 Serious public health concerns were raised by a recent phase II/III clinical trial that showed that a gel containing 3.5% N-9 (Advantage S) may increase HIV transmission by causing lesions. 5
The evaluation of the local tolerance of newly developed vaginal formulations has to be performed in laboratory experiments before human exposure to the product. 6,9 It is generally recommended that both the active agent and the clinical formulation of the product are tested in a rabbit vaginal irritation model (standard 10-day application) early in the development process. 6 Various other in vivo and in vitro models have been used to investigate the safety of vaginal products. The effects on the vaginal epithelium were studied in mice, 10,11 rats, 12 and non-human primates. 13 In vitro assays used cells of primary human origin or immortalized cervical and vaginal epithelial cell lines to evaluate the cytotoxicity of vaginal formulations. 4,14
In general, the use of vertebrates to evaluate the safety has been widely criticized based on scientific, ethical, and economic grounds. The concept of the 3 R’s (refinement, replacement, and reduction) stimulates the development of alternative methods such as in vitro methods and the use of “lower” organisms (invertebrates, plants, and microorganisms) as test organisms. 15
Adriaens and Remon developed an alternative test using invertebrates as a model organism. 16 The terrestrial slug Arion lusitanicus was selected as test organism. The body wall of slugs consists of a single-layered epithelium composed of epithelial cells and mucous gland cells overlying connective tissue. Slugs exposed to irritating substances produce mucus to protect the body wall. The amount of mucus produced by the slugs during a repeated contact period is a measure for irritation. Membrane damage can be estimated from the release of proteins, lactate dehydrogenase (LDH), and alkaline phosphatase (ALP) from the body wall of the slugs. The mucosal irritation test has been validated with reference molecules as an alternative test for screening the eye irritation potential of chemicals. 17 Furthermore, studies showed that the irritation potential of bioadhesive powder formulations could be estimated with this test. 18,19
In the current study, the mucosal irritation test was optimized for the evaluation of the local tolerance of vaginal gels. It was determined if a repeated treatment on 5 successive days was necessary. The influence of the amount of gel on the end points of the test was evaluated. Indeed, previous studies with powder formulations have shown that the amount of formulation influenced the mucus production of the slugs. Besides, the mucus production is the most important end point to discriminate between non-irritating and mildly irritating formulations. For this purpose, a gel containing 4.0% N-9 (Conceptrol) was used as positive control like it is recommended by the International Working Group on Vaginal Microbicides. 6 Today, there is no generally accepted negative control gel available, hence 2 gels that possibly could serve as negative controls were selected. On the one hand, hydroxyethyl cellulose gel was chosen, because Ballagh et al. used this gel as a negative control product in a phase I randomized, double-blind study. 20 On the other hand, Replens was selected, because it was used as a placebo in randomized, placebo-controlled, triple-blinded trials. 5,21 Then the relevance of the optimized procedure was investigated. Because no list of reference standards for screening the local tolerance of vaginal formulations is available, a few over-the-counter (OTC) formulations that have been evaluated in animals and/or humans were used in this study. The results of the mucosal irritation test were compared with available in vivo data.
Materials and Methods
All the vaginal formulations except the hydroxyethyl cellulose (HEC) gel were commercially available gels. HEC gel consisted of 5.0% (w/w) HEC (377 mPa.S), 10% (w/w) glycerin, 0.68% (w/w) methylparaben, 0.02% (w/w) propylparaben, and water. The OTC vaginal lubricant Replens (Columbia Laboratories, Hitchin, UK) contains carbomer, polycarbophil, glycerin, paraffin, hydrogenated palm oil glyceride, sorbic acid, sodium hydroxide, and purified water. K-Y jelly (Johnson & Johnson Inc., Montreal, Canada) is an OTC vaginal lubricant that is composed of chlorhexidine gluconate, hydroxyethyl cellulose, glucono delta lactone, glycerin, methylparaben, sodium hydroxide, and purified water. The OTC vaginal contraceptive gel Advantage S (Columbia Laboratories) uses Replens as a vehicle base and contains 3.5% N-9, carbomer 934P, polycarbophil, glycerin, mineral oil, hydrogenated palm oil glyceride, methylparaben, sorbic acid, sodium hydroxide, and purified water. Protectaid (Pirri Pharma Canada Inc., Montreal, Canada) is commercially available as a contraceptive sponge that is impregnated with F-5 gel containing 0.125% N-9, 0.125% benzalkonium chloride, 0.5% sodium cholate, hydroxypropyl methylcellulose, glycerin, dimethicone, and purified water. Conceptrol (Ortho-McNeil Pharmaceutical Inc., Raritan, NJ) is an OTC vaginal contraceptive gel preparation containing 4.0% N-9, sodium carboxymethylcellulose, propylene glycol, methylparaben, povidone, sorbic acid, sorbitol solution, lactic acid, and purified water.
Mucosal Irritation Test
The mucosal irritation test procedure described by Adriaens and Remon 16 was modified for the evaluation of vaginal gels and is described briefly. Slugs weighing between 3 g and 6 g were isolated from the culture 2 days before the start of an experiment and were placed in a vented plastic box lined with a paper towel (moistened with phosphate-buffered saline [PBS, pH 7.4]) at 18 to 22°C.
The slugs were placed daily during 30 minutes on the formulation in a Petri dish. For each formulation, 5 slugs were used. The amount of mucus produced during the contact period was measured by weighing the Petri dishes containing the formulation before and after the contact period. The mucus production was expressed as percentage (w/w) of the body weight. After the contact period, the slugs were transferred to a fresh Petri dish containing 1 mL PBS. The PBS samples were collected with a micropipette after 60 minutes. Then the slugs were placed in a fresh Petri dish and again 1 mL PBS was added. After 60 minutes, the PBS samples were removed. The samples were immediately analyzed for the presence of proteins, LDH, and ALP released from the body wall of the slugs. The slugs were placed in a Petri dish on a membrane filter (0.45 μm cellulose acetate; Sartorius AG, Goettingen, Germany) moistened with 2 mL PBS until the next contact period. This procedure was repeated during 5 successive days.
The protein concentration in the samples was determined with a NanoOrange protein quantitation kit (Molecular Probes, Leiden, The Netherlands) and expressed as micrograms per milliliter per gram of body weight. The NanoOrange reagent allows accurate detection of proteins in solution at concentrations between 10 ng/mL and 10 μg/mL. The fluorescence measurements were carried out on a fluorometer (Wallac 1420 multilabel counter; Perkin Elmer, Turku, Finland) using excitation/emission wavelengths of 485/590 nm. Bovine serum albumin was used as a standard.
Lactate Dehydrogenase Determination
The lactate dehydrogenase activity (LDH, EC 126.96.36.199) was measured with an enzyme kit (DG 1340-UV; Sigma Diagnostica, Bornem, Belgium) and expressed as units per liter per gram of body weight. The LDH reagents measure the enzyme activity based on the optimized standard method recommended by the German Society for Clinical Chemistry. 22 The LDH activity measurements were conducted on a Cobas Plus analyser (ABX, Brussels, Belgium) at 37°C.
Alkaline Phosphatase Determination
The alkaline phosphatase activity (ALP, EC 188.8.131.52) was measured with an enzyme kit (DG 1245-UV; Sigma Diagnostica) and expressed as units per liter per gram of body weight. The ALP reagents measure the enzyme activity based on the optimized standard method recommended by the German Society for Clinical Chemistry. 22 The ALP activity measurements were conducted on a Cobas Plus analyzer (ABX, Brussels, Belgium) at 37°C.
For each slug, the total mucus production, the mean protein release, the mean LDH release, and the mean ALP release were calculated and these data were used for the statistical analyses. Statistically significant differences between different amounts of gel or repeated experiments or different treatments were determined using a 1-way analysis of variance (ANOVA). The data were tested for normal distribution with a Kolmogorov-Smirnov test. The homogeneity of variances was tested with the Levene’s test. If the variances were found to be not equal, the data were transformed to their logarithm. To further compare the effects of the different amounts or treatments, a multiple comparison among pairs of means was performed using a Scheffé test that regarded P <0.05 as statistically significant. For all the statistical analyses, the computer program SPSS (version 11.0; SPSS, Chicago, IL) was used.
Optimization of the Procedure
The procedure was optimized for the evaluation of vaginal gels by testing several amounts (50, 100, 150, and 200 mg) of 2 gels that possibly could serve as negative control (HEC gel and Replens) and a positive control gel containing 4.0% N-9 (Conceptrol). All the slugs treated with the different amounts of HEC gel and Replens survived a repeated treatment on 5 successive days. A daily treatment with 50 mg and 100 mg of Conceptrol led to 20% mortality by day 5 (before the fifth treatment). Sixty percent of the slugs treated with 150 mg and 200 mg Conceptrol were dead by day 4 and eventually all the slugs were dead by day 5.
First, the data were analyzed to find out if a repeated treatment on 5 successive days was necessary. The mucus production profiles of the slugs treated with 50 mg of each gel were comparable to those of the slugs treated with 100 mg of the same gel; hence, only the latter are shown in Figure 1. The mucus production profiles after treatment with 200 mg gel were similar to those of 150 mg gel and are presented in Figure 2. To classify the formulations into different irritation classes, it is important that the range between non-irritating and irritating formulations is as large as possible. Figures 1 and 2 illustrate that one treatment did not suffice to make a distinction between the 3 gels. For the total mucus production, the range between the 3 gels increased with a repeated treatment resulting in the largest range on day 5.
The different amounts of HEC gel and Replens did not affect the protein release. For Conceptrol, the protein release increased with an increasing amount of gel. Only the protein release profiles after treatment with 100 mg gel are shown (Fig. 3). After the first treatment, all gels induced a high protein release. However, from the second day on, HEC gel and Replens resulted in low protein release levels comparable to the ones of the negative control slugs in previous studies. 18,19 The protein release by the Conceptrol slugs increased with a repeated treatment. The largest range of protein release between the different gels was obtained on day 5.
An increased release of the cytosolic enzyme LDH and the membrane-bound enzyme ALP from the mucosa of the slugs is an indication of severe membrane damage. Enzyme release was only detected for the slugs treated with Conceptrol. Fifty milligrams of Conceptrol induced LDH release after the third contact period. Amounts of 100 mg Conceptrol and more resulted in LDH release after the second contact period. ALP release appeared 1 day after LDH release except for 150 mg Conceptrol, which induced ALP release after the second contact period. Figure 4 illustrates that the LDH and ALP release increased with a repeated treatment.
With respect to the irritation potential, the best discrimination between the different gels could be made after a repeated treatment on 5 successive days. Prolongation of the experiment was not useful, because the mortality of the slugs treated with irritating gels would increase so that the end point range between non-irritating and irritating gels would not increase.
Next, the effect of the amount of gel on the end points of the test was evaluated to select the most suitable amount. The influence of the amount of gel on the total mucus production and on the mean protein, LDH, and ALP release is shown in Table 1. For each gel, the total amount of mucus increased with an increasing amount of gel. The slugs treated with HEC gel produced generally the smallest amount of mucus. The mucus of the HEC gel slugs was colorless, which is normal, whereas the mucus produced by the Replens slugs was slightly yellow. The 50, 100, and 200 mg amounts of Conceptrol induced a significantly increased mucus production compared with the analogous amounts of HEC gel and Replens (P <0.05) and the mucus color was yellow up to orange. The largest range of mucus production between the slugs exposed to Conceptrol and the other 2 gels was obtained when the slugs were placed on 100 mg gel.
Because all slugs released high protein amounts on the first day of treatment (Fig. 3), the mean protein concentrations were calculated without the data of day 1. Table 1 shows that HEC gel and Replens resulted in low protein release. The different amounts of Conceptrol showed a significantly increased protein release compared with the same amounts of HEC gel and Replens (P <0.05). The mean protein release increased with an increasing amount of Conceptrol, although the difference was not statistically significant (P >0.05). The effect of the different amounts of Conceptrol on the mean LDH and ALP release was comparable (P >0.05).
The best discrimination between the HEC gel, Replens, and Conceptrol was made based on the mucus production when the slugs were placed on 100 mg gel during 30 minutes for 5 successive days. Thus, 100 mg was selected as the test amount to evaluate the local tolerance of vaginal gels and the treatment was repeated for 5 successive days. Furthermore, HEC gel and Replens induced a low mucus production and protein release, and no enzyme release was detected. However, HEC gel induced a lower mucus production and protein release than Replens. For this reason, HEC gel was chosen as the negative control. Conceptrol was selected as the positive control in the mucosal irritation test, because the gel induced a high mucus production, a high protein release, and the release of enzymes.
Relevance of the Optimized Procedure
The effect of a repeated treatment with several vaginal gels on the end points of the mucosal irritation test is shown in Table 2. HEC gel, Replens, and Conceptrol were tested independently on 2 separate occasions. ANOVA testing resulted in no significant difference (P >0.05) for the total mucus production and the mean protein release between the repeated experiments with HEC gel and Replens. There was 1 exception for the repeated experiments with Conceptrol because the total mucus production of the first experiment (24.6 ± 5.3, n = 5) was significantly lower than the total mucus production of the second experiment (37.9 ± 5.7, n = 5). No significant difference (P >0.05) was detected for the mean protein release and the mean enzyme release between the repeated experiments with Conceptrol.
The total mucus production of the slugs treated with Replens was higher than that of the negative control slugs (treated with HEC gel); however, the difference was not significant (P >0.05). A repeated treatment with Replens resulted in a protein release comparable to the negative control slugs (P >0.05). No enzyme release and no mortality were detected. The total mucus production of the slugs treated with K-Y jelly was comparable to the positive control slugs (treated with Conceptrol) (P >0.05). The slugs produced slightly yellow mucus. The protein release induced by K-Y jelly was similar to the protein release of the negative control slugs (P >0.05). No enzyme release and no mortality were detected. Slugs treated with Advantage S produced an amount of yellow mucus that was comparable to the positive control slugs (P >0.05). Moreover, the slugs treated with this gel released more proteins than the negative control slugs, although the difference was not statistically significant (P >0.05). No enzyme release and no mortality were detected. Treatment with Protectaid induced a total amount of yellow mucus that was comparable to the positive control slugs (P >0.05). Protectaid resulted in a significantly higher protein release than the negative control slugs (P <0.05) and in the release of LDH and ALP. However, no mortality was detected.
The vaginal route offers a promising approach for the administration of, for instance, microbicides and spermicides. Frequent use of some vaginal formulations can damage the vaginal epithelium. 1,2 Mucosal irritation and damage might favor the entry of sexually transmitted pathogens. 3–5 Therefore, preclinical safety studies need to address the safety and local tolerance of newly developed vaginal formulations. The current preclinical model for assessment of vaginal irritation is the rabbit vaginal irritation model. 6 However, this model, as currently practiced and scored, failed to predict the vaginal toxicity caused by N-9 in humans. Moreover, there is a tendency to reduce, refine, and replace the use of vertebrates for preclinical safety studies. 15
The objective of this study was to optimize and evaluate an alternative in vivo test using slugs for local tolerance testing of vaginal gel formulations. The results of the mucosal irritation test using slugs were compared with available in vivo data on the safety of the tested gels. Predicting hazards to humans is the aim of most alternative tests, so human data should be the standard against which the relevance of an alternative test is compared. The major obstacle to interpreting the available clinical data was that the studies differed with respect to their purpose, frequency and duration of use, sample sizes, target populations, comparison products, rules regarding intercourse, the means and time points for assessing safety outcomes. Genital lesions were often assessed by means of colposcopy, which is a visual diagnostic tool with several limitations. 21
The mucosal irritation test procedure was optimized for local tolerance testing of vaginal gels. The study showed that a repeated treatment on 5 successive days was necessary to obtain a large range of mucus production, protein and enzyme release between the slugs exposed to HEC gel or Replens and those treated with Conceptrol. The results indicated that for the mucus production (which is the most important end point to discriminate between non-irritating and mildly irritating formulations), the largest range was achieved when the slugs were placed on 100 mg gel.
Because HEC gel induced the lowest mucus production and protein release, no enzyme release, and no mortality, HEC gel was classified as non-irritating and was selected as the negative control. Ballagh et al. concluded that colposcopic examination of 20 sexually inactive women who used HEC gel during 7 consecutive nights revealed no serious lesions not attributed to applicator injury. HEC gel was found to be acceptable in that study. 20
A repeated treatment of the slugs with Conceptrol resulted in a high mucus production, a high protein release, and the release of LDH and ALP, indicating severe damage of the slug mucosa. Because it is recommended that a gel containing 4.0% N-9 is used as a positive control in the rabbit vaginal irritation model, 6 Conceptrol was selected as the positive control in the mucosal irritation test. Several in vivo studies showed genital irritation after single or frequent use of Conceptrol. 10,13,23,24 A single application of Conceptrol resulted in the entry of inflammatory leukocytes into the vagina of mice within 4 hours. 10 When pig-tailed macaques were treated daily for 3 days, not treated for 2 days, and then treated for 3 additional days with Conceptrol, epithelial disruption and inflammatory response were noted. 13 Poindexter et al. reported that the use of Conceptrol by 31 sexually inactive women during 7 consecutive nights in a crossover trial caused cervical redness and/or ulceration in 26% of the subjects. 23
The slightly yellow (nonsignificantly) higher mucus production of the slugs treated with Replens compared with the negative control slugs is an indication for mild irritation; however, no increased protein release, no enzyme release, and no mortality were detected. On the one hand, Replens is considered to be safe and without adverse effects on the vaginal or cervical mucosa. After a daily administration of Replens for 10 consecutive days to 1 rabbit, histology was normal and there was no systematic increase of the number of inflammatory cells. 25 On the other hand, the application of Replens once daily for 14 consecutive days by 178 healthy sexually active women resulted in a significantly higher incidence of edema, reported vaginal discharge, and genital itching than the no-treatment control group. Notwithstanding these results, it was concluded that Replens was fairly innocuous. 26 The use of Replens as a placebo by female sex workers in a phase II/III trial revealed that frequent use caused lesions. 5
Treatment of the slugs with K-Y jelly induced a total mucus production comparable to the positive control slugs, a low protein release, no enzyme release, and no mortality. The increased mucus production indicates moderate irritation of the mucosal tissue of the slugs, which might be due to the presence of chlorhexidine gluconate, because a single 15-minute exposure to a 0.5% chlorhexidine digluconate gel damaged a major portion of the vaginal epithelium of mice. 11 Vaginal administration of K-Y jelly to rabbits for 10 consecutive days has also been shown to cause mild irritation to the vagina (Emilia Lonardo, principal scientist, Personal Products Co., personal communication, December 20, 2000). 24 A repeated application of K-Y jelly by 12 healthy sexually inactive women during 6 successive nights resulted in colposcopic findings, and the number of findings was lower than in the Conceptrol group. 24
In this study, Advantage S induced a total mucus production comparable to the positive control slugs and an increased protein release. It is interesting to note that Advantage S is formulated by the addition of 3.5% N-9 to the vaginal lubricant Replens and that it contains less carbomer than Replens. 5 So the irritation caused by Advantage S is probably the result of the presence of N-9. Various phases of clinical trials were completed before the adverse effects of Advantage S became clear. The application of Advantage S once daily for 14 consecutive days by 179 healthy sexually active women induced significantly more lesions than the Replens and the no-treatment control group. However, the use of Advantage S was not associated with a significant level of epithelia-disrupting lesions compared with the control groups. 26 A phase II trial involving the application of Advantage S multiple times a day by female sex workers revealed no difference between Advantage S and Replens with respect to the incidence of lesions. 21 A recent phase II/III trial in the same target population finally concluded that frequent use of Advantage S increased women’s susceptibility to HIV-1 infection by causing lesions (with or without epithelial breach). 5
A repeated treatment of the slugs with Protectaid resulted in a total mucus production comparable to the positive control slugs and caused severe membrane damage as was demonstrated by the release of LDH and ALP. The observed membrane damage can be the result of the presence of 0.125% N-9 and 0.125% benzalkonium chloride. Indeed, Patton et al. reported that a repeated application of N-9 plus benzalkonium chloride induced a more robust inflammatory response in pig-tailed macaques than repeated applications of either N-9 or benzalkonium chloride alone. 13 Our results contrast with recent studies involving the vaginal use of the Protectaid sponge by women. Creatsas et al. concluded that the use of the sponge by 15 women during intercourse during 1 year caused no colposcopic signs of vaginal lesions or irritation. 27 The apparent discrepancy in results could be explained by the fact that we tested only the gel that is impregnated in the sponge. A sponge releases smaller amounts of gel over a long period of time, possibly leading to less exposure to gel at any one time. 23
Based on the results of this study and on previous studies concerning the mucosal irritation test, a prediction model can be developed. A prediction model is important, because it defines exactly how to convert the results from an alternative method into predictions of in vivo toxicity. 28 Generally, non-irritating formulations induce a low mucus production, a low protein release, and no enzyme release, whereas irritating formulations result in an increased mucus production and an increased protein and enzyme release. 16,18,19,29 However, because the relevance of the mucosal irritation test needs to be evaluated with additional vaginal formulations, the proposed prediction model is susceptible to change. Vaginal gels that induce a low total mucus production (<15%), a low protein release, and no enzyme release will be classified as non-irritating (eg, HEC gel). Gels that cause no additional effect on the protein and enzyme release, but induce a mucus production of 15% to 20% or ≥20%, will respectively be predicted as mildly irritating (eg, Replens) or moderately irritating (eg, K-Y jelly) formulations. Vaginal gels that result in an increased mucus production (≥15%) and in an increased protein release (≥30 μg/mL.g) and/or enzyme release will be classified as severely irritating gels (eg, Advantage S, Protectaid, and Conceptrol). A classification prediction model simplifies the interpretation of the data. A disadvantage of a classification model is that formulations with a borderline irritation potential are not always classified correctly. So a careful analysis of the data are still recommended.
The mucosal irritation test seems to be able to classify the formulations into more irritation categories than commonly practiced in vitro cytotoxicity tests. Maguire et al. classified several vaginal formulations as non-irritating or irritating by using an immortalized cell line derived from human vaginal cells. This in vitro study concluded that Replens and K-Y jelly were not cytotoxic to human vaginal cells and that Advantage S had a cytotoxic effect on the cells. 14
The results of this study indicated that the mucosal irritation test using slugs could be used to evaluate the effects of a repeated treatment with vaginal gels and to classify the formulations into at least 4 irritation categories. Local tolerance data could be obtained accurately, quickly, and economically without using large amounts of compounds. The test seems to be useful to screen new vaginal formulations for local tolerance before preclinical studies in vertebrates and clinical studies in humans. The mucosal irritation test can be used together with other models that enable screening of new compounds for physical-chemical properties and efficacy early in the development process.
1. 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.
2. Stafford MK, Ward H, Flanagan A, et al. Safety study of nonoxynol-9 as a vaginal microbicide: Evidence of adverse effects. J Acquir Immun Defic Syndr Hum Retrovirol 1998; 17: 327–331.
3. Stephenson J. Widely used spermicide may increase, not decrease, risk of HIV transmission. JAMA 2000; 284: 949–949.
4. Fichorova RN, Tucker LD, Anderson DJ. The molecular basis of nonoxynol-9-induced vaginal inflammation and its possible relevance to human immunodeficiency virus type 1 transmission. J Infect Dis 2001; 184: 418–428.
5. Van Damme L, Ramjee G, Alary M, et al. Effectiveness of COL-1492, a nonoxynol-9 vaginal gel, on HIV-1 transmission in female sex workers: A randomised controlled trial. Lancet 2002; 360: 971–977.
6. The International Working Group on Vaginal Microbicides. Recommendations for the development of vaginal microbicides. AIDS 1996; 10: UNAIDS1–UNAIDS6.
7. Hicks DR, Martin LS, Getchell JP, et al. Inactivation of HTLV-III LAV-infected cultures of normal human-lymphocytes by nonoxynol-9 in vitro. Lancet 1985; 2: 1422–1423.
8. Harrison C, Chantler E. The effect of nonoxynol-9 and chlorhexidine on HIV and sperm in vitro. Int J STD AIDS 1998; 9: 92–97.
9. European Commission. Non-clinical Local Tolerance Testing of Medicinal Products (Directive 75/318/EEC). Eudralex 3BS12a; 1998: 107–116.
10. Milligan GN, Dudley KL, Bourne N, Reece A, Stanberry LR. Entry of inflammatory cells into the mouse vagina following application of candidate microbicides—Comparison of detergent-based and sulfated polymer-based agents. Sex Transm Dis 2002; 29: 597–605.
11. Achilles SL, Shete PB, Whaley KJ, Moench TR, Cone RA. Microbicide efficacy and toxicity tests in a mouse model for vaginal transmission of Chlamydia trachomatis
. Sex Transm Dis 2002; 29: 655–664.
12. Chvapil M, Droegemueller W, Owen JA, Eskelson CD, Betts K. Studies of nonoxynol-9. 1: The effect on the vaginas of rabbits and rats. Fertil Steril 1980; 33: 445–450.
13. Patton DL, Kidder GG, Sweeney YC, Rabe LK, Hillier SL. Effects of multiple applications of benzalkonium chloride and nonoxynol 9 on the vaginal epithelium in the pigtailed macaque (Macaca nemestrina). Am J Obstet Gynecol 1999; 180: 1080–1087.
14. Maguire RA, Bergman N, Phillips DM. Comparison of microbicides for efficacy in protecting mice against vaginal challenge with herpes simplex virus type 2, cytotoxicity, antibacterial properties, and sperm immobilization. Sex Transm Dis 2001; 28: 259–265.
15. Balls M, Goldberg AM, Fentem JH, et al. The three Rs: The way forward—The report and recommendations of ECVAM workshop 11. Altern Lab Anim 1995; 23: 838–866.
16. Adriaens E, Remon JP. Gastropods as an evaluation tool for screening the irritating potency of absorption enhancers and drugs. Pharm Res 1999; 16: 1240–1244.
17. Adriaens E, Remon JP. The evaluation of an alternative mucosal irritation test using slugs. Toxicol Appl Pharmacol 2002; 182: 169–175.
18. Callens C, Adriaens E, Dierckens K, Remon JP. Toxicological evaluation of a bioadhesive nasal powder containing a starch and Carbopol® 974 P on rabbit nasal mucosa and slug mucosa. J Control Release 2001; 76: 81–91.
19. Adriaens E, Ameye D, Dhondt MMM, Foreman P, Remon JP. Evaluation of the mucosal irritation potency of co-spray dried Amioca®/Poly(acrylic acid) and Amioca®/Carbopol® 974P mixtures. J Control Release 2003; 88: 393–399.
20. Ballagh SA, Baker JM, Henry DM, Archer DF. Safety of single daily use for one week of C31G HEC gel in women. Contraception 2002; 66: 369–375.
21. Van Damme L, Chandeying V, Ramjee G, et al. Safety of multiple daily applications of COL-1492, a nonoxynol-9 vaginal gel, among female sex workers. AIDS 2000; 14: 85–88.
22. Deutsche Gesellschaft für Klinische Chemie. Standardisierung von methoden zur bestimmung von enzymaktivitäten in biologischen flüssigkeiten. Z Klin Chem Klin Biochem 1972; 10: 182.
23. Poindexter AN, Levine H, Sangi-Haghpeykar H, et al. Comparison of spermicides on vulvar, vaginal, and cervical mucosa. Contraception 1996; 53: 147–153.
24. Mauck C, Weiner DH, Ballagh S, et al. Single and multiple exposure tolerance study of cellulose sulfate gel: A phase I safety and colposcopy study. Contraception 2001; 64: 383–391.
25. Balzarini J, Naesens L, Verbeken E, et al. Preclinical studies on thiocarboxanilide UC-781 as a virucidal agent. AIDS 1998; 12: 1129–1138.
26. Van Damme L, Niruthisard S, Atisook R, et al. Safety evaluation of nonoxynol-9 gel in women at low risk of HIV infection. AIDS 1998; 12: 433–437.
27. Creatsas G, Elsheikh A, Colin P. Safety and tolerability of the new contraceptive sponge Protectaid®. Eur J Contracept Reprod Health Care 2002; 7: 91–95.
28. Bruner LH, Carr GJ, Chamberlain M, Curren RD. Validation of alternative methods for toxicity testing. Toxicol In Vitro 1996; 10: 479–501.
29. Adriaens E, Dierckens K, Bauters TGM, et al. The mucosal toxicity of different benzalkonium chloride analogues evaluated with an alternative test using slugs. Pharm Res 2001; 18: 937–942.