Objective
The purpose of this study was to compare the efficacy of 2 different doses of tinidazole with metronidazole for the treatment of bacterial vaginosis and to compare the side effects of the drugs.
Study Design
Women were assigned randomly to receive metronidazole 500 mg twice daily, tinidazole 500 mg twice daily, or tinidazole 1 g twice, all for 7 days. Follow-up visits were conducted at days 14 and 28.
Results
Five hundred ninety-three women were enrolled. There were no significant differences between the treatment arms. Overall cure rates were 76.8% at 14 days and 64.5% at 1 month. Women who admitted to engaging in sexual intercourse during the study were significantly more likely to have bacterial vaginosis at the follow-up visit. There were no significant differences in adverse events across treatment arms.
Conclusion
There were no differences in cure rates between metronidazole and either of the tinidazole dosing regimens that were studied. In addition, there were no important differences in the side-effect profiles of metronidazole and tinidazole.
Bacterial vaginosis (BV) is the most prevalent cause of symptomatic vaginal discharge in the United States and has been associated with complications that include preterm delivery of infants, pelvic inflammatory disease, urinary tract infections, and the acquisition/transmission of sexually transmitted diseases (STDs) that include human immunodeficiency virus (HIV). Control of BV has been advocated as a means of decreasing the prevalence of these complications. However, available treatment regimens have been disappointing in terms of initial cure and recurrence rates. Tinidazole was licensed recently in the United States for the treatment of trichomoniasis and BV. The indication for BV was based on a tinidazole vs placebo trial. There are no data on the efficacy of multidose tinidazole vs metronidazole for treatment of this infection. Tinidazole has the potential of being a more efficacious drug for BV based on its longer half-life and its seemingly superior side-effect profile, compared with oral metronidazole. We compared the efficacy of 2 different doses of tinidazole with oral metronidazole for the initial treatment of symptomatic BV and short-term recurrence rates. In addition, side-effect profiles of the 2 drugs were compared.
Materials and Methods
Heterosexual women who attended the Jefferson County Department of Health STD Clinic with symptomatic BV (as defined by the criteria of Amsel et al ) and no evidence of STD were invited to participate in this randomized, double-blinded study. To be eligible the subject had to report symptoms of vaginal discharge and/or vaginal odor and have a vaginal pH >4.5, positive “whiff” test, and clue cells on microscopy. Women with known HIV infection were excluded from the study.
The study was approved by the institutional review board at the University of Alabama at Birmingham. Subjects were administered standardized questionnaires about symptoms and a comprehensive sexual history. A pelvic examination was performed, and specimens were collected for vaginal pH, microscopy, Gram stain, gonorrhea and chlamydia testing, and trichomonas culture.
Women were assigned randomly to receive oral metronidazole 500 mg twice a day for 7 days, tinidazole 500 mg twice a day for 7 days, or tinidazole 1 gm twice a day for 7 days. The study was initiated before the Food and Drug Administration (FDA) approval of tinidazole for BV, thus dosing was based on reviews of the literature.
Follow-up visits were conducted at days 14 and 28 of the study and then monthly for an additional 2 visits. At each of the follow-up visits, a standardized questionnaire was administered; a pelvic examination was conducted, and specimens for the diagnosis of BV were collected. Women with symptomatic BV at the follow-up visits were treated with the current standard therapy of 7 days of metronidazole and discontinued from the study at that time. Any woman who was found to have a positive screening test for gonorrhea and/or chlamydia was treated appropriately and discontinued from the study, with the exception of women with positive culture tests for Trichomonas vaginalis because this infection would be treated adequately with either metronidazole or tinidazole. Women with intercurrent vaginal yeast infections were treated and continued in the study.
Microbiologic cure was defined as a Nugent score of <7. Clinical improvement was defined as normalization of 2 of 3 criteria (pH, whiff test, and clue cells), and cure was defined as normalization of all 3 criteria. The persistence of ≥2 abnormalities constituted clinical failure.
Microbiologic methods
Vaginal pH, amine odor (whiff test), and microscopy were performed, as previously described. Vaginal smears were Gram stained and interpreted according to the method of Nugent et al. Neisseria gonorrhoeae and Chlamydia trachomatis were Gen-Probe Aptima Combo 2 assay (Gen-Probe Incorporated, San Diego, CA) according to manufacturer’s instructions. Trichomonas cultures were performed with the In Pouch TV test (BioMed Diagnostics Inc, White City, OR).
Statistical considerations
Sample size and power
Sample-size estimates were computed for the efficacy analysis based on the cure rates that were defined as 70% for the metronidazole group, 85% for the tinidazole 500-mg twice-a-day group, and 90% for the tinidazole 1-gm group. Assuming a 20% loss to follow-up overall, a sample size of 192 per group (576 women) was assigned randomly to the 3 treatment arms, with an alpha level of .025. No formal comparison was planned for the 2 tinidazole arms. Based on 160 randomly assigned subjects per group, we would be able to detect a difference of 15% in side-effects among the treatment arms. This would be an event proportion of approximately 15% in the 2 tinidazole arms with an alpha of .025.
Statistical analyses
The proportion of women who were cured of BV at the 14 day follow-up visit (7 days after completion of therapy) was compared between the metronidazole and 500-mg tinidazole group with the chi-square test of proportions; exact confidence intervals were computed. The proportion of women who were cured was compared between the metronidazole and 1-g tinidazole groups in a similar fashion. Descriptive statistics of the population (such as age, sexual history, STD history) were compared between the treatment arms by the t test for compare continuous variables and the Fisher’s exact test or the χ 2 test was used for nominal variables.
The short-term recurrence rates at 1, 2, and 3 months among subjects who were cured successfully after the initial treatment (at the 14-day visit) were compared across treatment groups with the chi-square test. Confounding variables that were tested in a repeated measures model to predict BV included number of partners (>1 vs 1), new partner (yes vs no), consistent condom usage (yes vs no), current douching (yes vs no), and baseline BV score (≥7 vs 4-7). For all analyses, a probability value of < .05 was considered statistically significant.
Adverse events were categorized as gastrointestinal, vaginal/gynecologic, neurologic, and other. The proportion of women who experienced an event was compared across the 3 treatment groups by the Fisher’s exact test for each type of event and for overall events. In addition, the total number of events (>1 event possible per person) was compared across the 3 treatment groups in a similar fashion.
The study statistician was responsible for preparing computer-generated randomization lists for the clinical trial, and a blocked randomization scheme was used to generate the list. The randomization scheme was provided only to the study pharmacist who was responsible for preparing the study medication in a blinded fashion. The actual randomization assignments were kept in a secured fashion by the Biostatistics Unit.
Results
Five hundred ninety-three patients were enrolled in the study between November 2004 and October 2009. The allocation of subjects for the trial up to and including the 1 month follow-up visit is shown in the Figure . The most common reason for termination was loss to follow-up observation (158 women; 26.7%) followed by preexisting protocol violation (57 women; 9.6%), of whom 52 women (91.2%) had an STD at baseline. The demographic and behavioral characteristics of subjects by randomized treatment group are shown in Table 1 . There were no significant differences with respect to race, age, or behaviors (such as douching and sexual behavior). Of note, nearly 97% of the women reported a history of STD; 83% of the women reported a history of BV.
