A pilot clinical trial comparing an acid-buffering formulation (ACIDFORM gel) with metronidazole gel for the treatment of symptomatic bacterial vaginosis

Abstract

Aim

To compare the effectiveness of an acid-buffering formulation gel (ACIDFORM) with metronidazole gel in the treatment of symptomatic bacterial vaginosis (BV).

Methods

After a confirmed diagnosis of BV according to the criteria established by Nugent and Amsel, 30 nonpregnant women were enrolled in a randomized, double-blind clinical study. The women were randomly assigned to receive either 5 g ACIDFORM gel (n = 13) or 10% metronidazole gel (n = 17) intravaginally once daily for five consecutive days. Participants were evaluated in two follow-up visits (7–12 days and 28–35 days after treatment). Therapeutic success was defined as the presence of less than three of Amsel's criteria. If three or more criteria were present at first or second follow-up visit, the woman was excluded from the study and treated orally with metronidazole. Nugent scores were recorded at each visit but these were not used to define cure.

Results

At the first follow-up visit, 15 (88%) of the women in the metronidazole group were cured compared with only three (23%) in the ACIDFORM group (P < 0.001). The remaining 12 women (10 of the ACIDFORM group and two of the metronidazole group) were considered as failure and were treated orally with metronidazole. At the second follow-up visit, two of the ACIDFORM-treated women and six of the metronidazole-treated women presented recurrent BV. Four women in the ACIDFORM group and one in the metronidazole group reported occasional burning and itching during product use.

Conclusion

ACIDFORM gel was significantly less effective than high-dose metronidazole gel for the treatment of symptomatic BV.

Introduction

Bacterial vaginosis (BV) is the most common vaginal disorder among women of reproductive age. It is present in 15–25% of gynaecology patients and in up to 60% of women seen at sexually transmitted diseases (STD) clinics [1, 2]. BV is defined as a clinical syndrome in which the normal lactobacilli-predominant vaginal flora is replaced by an overgrowth of Gardnerella vaginalis and anaerobic microorganisms [1, 2]. Although perceived as a mild medical problem, BV has been associated with several gynaecological and obstetrical complications, and an increased risk of HIV acquisition and transmission [3–5].

Metronidazole is the drug of choice recommended by the Centers for Disease Control and Prevention (CDC) for the treatment of BV [6]. Cure rates 1 week after treatment regimen are approximately 80%[7]; however, the recurrence rate is relatively high. After 1 month, cure rates fall to 60–70%[8–11]. Prolonged treatment with metronidazole may be more effective in preventing the recurrence of BV [1]. However, such long-term treatment is unlikely to be acceptable and raises concerns regarding potential toxicity. In addition, high concentrations of metronidazole inhibit the growth of vaginal lactobacilli, at least in vitro[12].

A possible alternative treatment for BV is the acidification of the vagina because most BV-causing bacteria cannot grow at a pH < 4.5. Studies with an intravaginal lactate-gel showed it to be as effective as oral metronidazole [13, 14]. However, other studies have shown that vaginal acidification is an ineffective treatment for BV [15–17]. Recently, two new acid-buffering gel formulations have reached clinical trial: ACIDFORM (TOPCAD, Chicago, IL, USA; licensed to Instead, Coppell, TX, USA) and BufferGel (ReProtect LLC, Baltimore, MD, USA). Preliminary results obtained during a Phase I safety study with BufferGel suggest that vaginal application twice daily for 14 days was associated with a decrease in the prevalence of BV from 30% at enrolment to 6% at 1 week and 7% at 2 weeks [18]. However, no comparisons were performed with metronidazole, nor were observations performed 1 month after treatment, when relapses are more common.

ACIDFORM possesses high acid-buffering and bioadhesive properties in vitro; these bioadhesive properties should provide long-term vaginal retention [19]. A clinical trial with a small sample of six healthy volunteers showed that ACIDFORM caused no vaginal or cervical irritation and resulted in no patient complaints when the gel was applied intravaginally once daily for six consecutive days [20]. Furthermore, ACIDFORM showed in vitro inhibition of G. vaginalis and of Peptostreptococcus sp., Prevotella sp., and several other anaerobic bacteria associated with BV, but failed to inhibit Lactobacillus acidophilus (Citron, Zaneveld, Anderson, and Waller, unpublished).

The aim of this study was to compare the effectiveness of ACIDFORM gel with metronidazole gel in the treatment of symptomatic BV.

Materials and methods

The study was conducted at the Gynaecological Outpatient Clinic, Department of Obstetrics and Gynaecology, School of Medicine, Universidade Estadual de Campinas (UNICAMP), Campinas, Brazil. IRB approvals were obtained both from the University of Campinas and the Eastern Virginia Medical School, Norfolk, VA, host institution of the sponsor (CONRAD).

As this was a pilot, double-blind clinical trial, the sample size was not chosen on the basis of statistical considerations. Therefore the statistical tests have power to detect only very large differences between groups. Nonpregnant women with complaints of vaginal discharge and/or vaginal malodour were invited to participate in the study if they met the following inclusion criteria: a confirmed diagnosis of BV (by both Amsel's [21] and Nugent's [22] criteria); not currently breastfeeding; wet mount and Gram stain negative for yeast and Trichomonas vaginalis; using an effective modern contraceptive method; not allergic to metronidazole; not having used antibiotics in the past 14 days; not having a history of STDs, including HIV and human papillomavirus, in the previous 6 months. The presence of a positive test for Neisseria gonorrhoea and/or Chlamydia trachomatis, as well as antibiotic use during the study, were postadmission withdrawal criteria.

A pelvic examination was performed and the presence or absence of homogeneous discharge was noted. Vaginal pH measurement with indicator strips and whiff test were performed. Vaginal discharge was examined by wet mount microscopy for the presence of clue cells. Amsel et al.[21] recommended basing the clinical diagnosis of BV on the presence of at least three of the following signs: (i) homogeneous white adherent discharge; (ii) vaginal pH > 4.5; (iii) release of a fishy amine odour from vaginal fluid when mixed with 10% KOH; and (iv) presence ≥ 20% of clue cells. A vaginal smear was also obtained for Gram stain and graded for the presence of BV using Nugent's criteria [22]. This is a standardized 0–10-point scoring system for evaluation of Gram-stained vaginal smears based on three morphotypes: large Gram-positive rods (lactobacilli), small Gram-negative or variable rods (Gardnerella and anaerobic rods) and Mobiluncus. A score ≥ 7 is positive for BV by Nugent's criteria.

After confirmed diagnosis of BV according to Nugent and Amsel's criteria, and after reading and signing an informed consent, 30 nonpregnant women were enrolled in the study. The women were randomly assigned to receive either 5 g ACIDFORM gel (n = 13) or 10% metronidazole gel (n = 17) intravaginally once daily for five consecutive days. ACIDFORM gel (TOPCAD; licensed to Instead) was manufactured by the same pharmaceutical company as that used in previously published trials [20]; metronidazole 10% gel is marketed in Brazil (Flagyl gel®; Aventis, São Paulo, Brazil). The tubes used for ACIDFORM were identical to the marketed tubes of Flagyl gel® and therefore indistinguishable from each other.

Participants were evaluated in two follow-up visits (7–12 days and 28–35 days after completion of treatment). Therapeutic success was defined as the presence of less than three of Amsel's criteria. If three or more criteria were present at the first or second follow-up visit, the woman was excluded from the study and treated orally with metronidazole. Nugent scores were recorded at each visit but these were not used to define cure. Therefore, both Nugent and Amsel criteria were used to confirm the diagnosis and select patients for the study, but only Amsel criteria were used to define therapeutic success. Criteria for yeast infection used in this study were based on Sabouraud vaginal culture positive for yeast.

Statistical comparisons were made using the χ² test of association, Fisher's exact test or the McNemar nonparametric test, as appropriate. The level of significance was established at 5%.

Results

There were no significant differences between the ACIDFORM and the metronidazole groups with respect to sociodemographic characteristics or history (Table 1). The groups were not balanced (13 in the ACIDFORM and 17 in the metronidazole group) because the randomization schedule was designed for 36 participants and only 30 were enrolled. It occurred due to a misunderstanding by the manufacturer who was responsible for the randomization. They asked the statistician to prepare the randomization for 30 participants plus an additional 20% (out of 30) in case of any dropout. It was supposed to be only ‘an extra randomization’ to cover eventual dropouts. Instead, the statistician included it in the total (and made the randomization for 36 participants). The investigators received the tubes with the patients’ numbers from the manufacturer and therefore they did not know about this distribution until the study was completed and the code was broken.

Table 1.

Selected characteristics of the participants in the ACIDFORM and metronidazole treatment groups

Characteristics/antecedents	ACIDFORM(n = 13)	Metronidazole(n = 17)	P-value
Age (years)
Mean	30.7	31.9	0.69***
SD	5.2	8.0
Parity
=2	11 (85%)	12 (71%)	0.42**
>2	2 (15%)	5 (29%)
Marital status
Married/stable union	8 (62%)	11 (65%)	>0.99**
Unmarried not living with	5 (38%)	6 (35%)
partner or no partner
Ethnicity
White	10 (77%)	10 (59%)	0.44**
Not white (black or brown)	3 (23%)	7 (41%)
Education (years in school)
=6 years	7 (54%)	9 (53%)	>0.99*
>6 years	6 (46%)	8 (47%)
Smoking habit
Yes	3 (23%)	6 (35%)	0.69**
No	10 (77%)	11 (65%)
Contraceptive method used
while in study
IUD/abstinencea	10 (77%)	14 (82%)	>0.99**
Hormonal injectable	3 (23%)	3 (18%)
History of BV in the past
Yes	7 (54%)	10 (59%)	>0.99*
No	6 (46%)	7 (41%)

^*χ² test of association with continuity correction.

^**Fisher's test.

^***Mann–Whitney test.

^aOnly one woman in the ACIDFORM group relied on abstinence.

The results with reference to treatment and BV cure rates are summarized in Figure 1. At the first follow-up visit (mean = 9.8 days, SD = 4.8), 15 (88%) of the women in the metronidazole group were cured, whereas only three (23%) of the women in the ACIDFORM group were cured (P < 0.001). The remaining 12 women (10 of the ACIDFORM group and two of the metronidazole group) were considered as failure and were treated orally with metronidazole. At the second follow-up visit (mean = 32.6 days, SD = 9.1), two of the ACIDFORM-treated women and six of the metronidazole-treated women presented recurrent BV (Figure 1).

Figure 1.

Flowchart showing the overall efficacy of ACIDFORM gel and metronidazole gel for the treatment of bacterial vaginosis (BV)

The cure rates were significantly lower in the subgroup of women with history of previous BV treated with ACIDFORM compared with those women with history of previous BV treated with metronidazole: 2/7 (28.6%) women vs. 9/10 (90%) women, respectively.

Figure 2 shows the response to therapy at the first follow-up visit, evaluated using Amsel's criteria. All these criteria showed significant differences between the two treatment groups, except regarding abnormal vaginal discharge (P = 0.119). The mean Gram stain score at the first follow-up was also significantly lower in the group of women treated with metronidazole compared with those women treated with ACIDFORM (2.9 vs. 6.9, P < 0.001).

Figure 2.

Status of each clinical criterion for diagnosis of bacterial vaginosis at first follow-up (1 week after treatment with ACIDFORM gel () or with metronidazole gel (▪), once daily, 5 g intravaginally for 5 days). * Fisher's exact test. ** Chi-square test

Symptomatic relief related by the women, i.e. the absence of vaginal discharge or malodour, did not correspond to the BV outcome. All 12 participants who had persistent BV at the first follow-up (10 in the ACIDFORM group and two in the metronidazole group) incorrectly reported that they were partially or totally cured.

While using the assigned product, five women reported symptoms of genital irritation: four in the ACIDFORM group and one in the metronidazole group (P = 0.138). After completion of treatment, the most frequent symptom reported by women was vaginal itching. The number of women with symptoms of itching and/or burning at any time during follow-up and the association of these symptoms with yeast is shown in Table 2.

Table 2.

Distribution of women with genital symptoms and presence of yeast infection at any time during the follow-up visits

Symptoms/yeast	ACIDFORM(n = 13)	Metronidazole(n = 17)	Total(n = 30)
Yeast positive*	7	7	14
Patient has symptoms (itching and/or burning)**	9	6	15
Symptoms and yeast positive	4	3	7

^*Based on Sabouraud vaginal culture results (none of the women had candidiasis at the start—it was an exclusion criterion).

^**The symptoms were quite mild.

Discussion

Our results showed that once-daily administration of ACIDFORM gel intravaginally for 5 days was significantly less effective than the same regimen of 10% metronidazole gel for the treatment of BV. In spite of the number of patients in our study being quite small, our findings were similar to those of other studies in which vaginal acidification proved to be an ineffective treatment for BV compared with metronidazole [15, 16] and with placebo [17]. The small size sample may affect the power and reproducibility of our data, although the differences observed were large.

The higher concentration of metronidazole (10% compared with 0.75%) commercially available in Brazil produced similar cure rates at the 1-week follow-up visit to those reported for the 0.75% gel (twice daily dosing), 75–87%[9, 23, 24]. However, it is known that very high local concentrations of metronidazole and clindamycin can inhibit the growth of lactobacilli in vitro, which is not desirable [12, 25].

It is possible that the women who were cured after using ACIDFORM gel as well as using metronidazole gel would have been cured even if they had used a placebo. Cure rates with placebo (vaginal cream or gel) have been reported to be approximately 17–27% with the same duration of follow-up [9, 23, 26], and this is similar to the 23% cure rate found in this study for ACIDFORM.

The BV cure rates in this study fell greatly between the second and third visits in both the metronidazole and ACIDFORM groups. At the second follow-up visit, two of the ACIDFORM-treated women and six of the metronidazole-treated women presented recurrent BV. Duration of follow-up is probably the most important variable in determining the effectiveness of any BV treatment regimen. Follow-ups that extend for 1 month after treatment generally show resolution rates that are significantly lower than those calculated only a few days following treatment.

Both treatment groups well tolerated the study product in this trial, with no discontinuation because of adverse events (AE). Genital irritation was the symptom most frequently reported by women and tended to occur more frequently in ACIDFORM users (four vs. one women). However, this difference was not statistically significant and it is hard to distinguish persistent BV from adverse effect of the treatment gel.

It is well known that neither symptoms nor vaginal characteristics alone are reliable bases for the diagnosis of BV [27]. In this study, women were unaware of persistent BV. All women who had persistent BV at the first follow-up incorrectly believed that they were cured. These findings reinforce previous recommendations that the clinician cannot make a reliable diagnosis of BV based only on the patient's impression of the symptoms or vaginal discharge. Low-cost and easily performed clinical tests (such as vaginal pH measure and whiff test) should always be performed.

Vulvovaginal candidiasis is a common sequela of BV treatment. Symptomatic yeast vaginitis develops at a rate of 10–30% in women whose BV is treated with other topical microbicides and conventional antibiotics [18, 24]. In the present study, the occurrence of yeast infection following treatment was similar in both treatment groups. The BV-associated organisms have been shown to produce amines that reduce yeast survival and replication [28]. Thus, the suppression of BV organisms by any effective treatment could release this inhibition, resulting in the frequent occurrence of this vaginal infection following treatment of BV.

In conclusion, ACIDFORM gel was significantly less effective than high-dose metronidazole gel for the treatment of symptomatic BV in this randomized, double-blind, clinical trial. These results increase the doubts already raised with respect to any actual benefit of using intravaginal acidification as a feasible alternative for the treatment of symptomatic BV.