Monthly itraconazole versus classic homeopathy for the treatment of recurrent vulvovaginal candidiasis: a randomised trial

Authors


A Witt, MD, Department of Obstetrics and Gynecology, University of Vienna Medical School, Währinger Gürtel 18-20, A-1090 Vienna, Austria. Email armin.witt@meduniwien.ac.at

Abstract

Objective  Antimycotics effectively treat sporadic and recurrent vulvovaginal candidiasis (RVVC). Classic homeopathy (CH) is also used to treat this condition. We compared the efficacy of CH and itraconazole in reducing the frequency of RVVC episodes.

Design  Single-centre, prospective, randomised trial.

Sample  One hundred-and-fifty patients with a history of RVVC and an acute episode of VVC.

Methods  Women were randomised into 3 groups: itraconazole with lactobacilli (group 1), itraconazole without lactobacilli (group 2) and CH (group 3). Itraconazole treatment of acute infection was followed by a 6-month maintenance regimen with monthly single-day itraconazole (200 mg bid). Women in group 1 were given additional vaginal lactobacilli for 6 days per month throughout the maintenance regimen Thereafter, patients were followed without treatment for 6 months. CH treatment was performed for 12 months.

Results  Women in groups 1 and 2 reached a culture-free status significantly earlier than women in group 3 (log-rank test; P < 0.0001). Specifically, before the start of the maintenance regimen, 44 of 49 women (89.8%) in group 1 and 40 of 47 women (85%) in group 2 were free of Candida detectable by culture, 22 of 46 (47%) women in group 3 reached a culture-free status after the first visit, but had a recurrence significantly earlier compared with women in groups 1 and 2 (log-rank test; P = 0.002). After 12 months, 19 of 25 (76%) women in group 1, 18 of 23 (78%) women in group 2 and 9 of 23 (39%) women in group 3 were free of culture-detectable Candida. Assessment of RVVC-associated complaints by VAS score showed that women in group 3 had a significantly higher level of discomfort (36.8, 25.1 and 27.7 respectively; P < 0.001) and were significantly less satisfied (59.2, 68.2 and 71.7 respectively; P < 0.001) than patients in groups 1 and 2.

Conclusions  Monthly cycle-dependent itraconazole is more effective than CH in the treatment of RVVC. Lactobacilli do not confer an added benefit.

Introduction

Between 70 and 75% of women are affected by vulvovaginal candidiasis at least once during their lifetime,1 and 40 to 50% of these patients will experience a recurrence.2 Candidiasis is currently the second most common vaginal infection after bacterial vaginosis.3 The most frequent yeast pathogen responsible for vulvovaginal candidiasis is Candida albicans, isolated from 85 to 90% of clinical specimens.4 Some of the host factors predisposing to vulvovaginal candidiasis are pregnancy, oral contraceptive use, diabetes mellitus, use of systemic antibiotics, corticosteroid therapy, human immunodeficiency virus (HIV) infection and other factors such as tight, poorly ventilated nylon underclothing with increased perineal moisture and temperature.4–7 The most common symptom of vulvovaginal candidiasis is pruritus, followed by vulvar burning, especially at the time of micturition or during sexual intercourse. Also, there is an increase in vaginal discharge, the vagina is oedematous and the vulva is hyperaemic. For some patients, especially those with recurrent episodes of vulvovaginal candidiasis, the symptoms lead to psychological disturbances and sexual dysfunction.8

Recurrent vulvovaginal candidiasis (RVVC) is usually defined by at least four episodes of vulvovaginal candidiasis per year and affects 5 to 8% of women in their reproductive years.9 RVVC has been associated with immunological factors such as circulating heat shock proteins10 or reduction of vaginal mannose-binding lectin.11 The pathophysiological mechanisms of RVVC, however, are not completely understood and clinical management remains problematic. Effective therapy involves identifying and eradicating the underlying and/or predisposing causes of the disease, but in most cases, no such factors are identified.12

Acute episodes of vulvovaginal candidiasis are usually treated with topical antifungal agents of the azole group. In recent years, the use of oral azoles has become more widespread. In women with RVVC, clinical studies demonstrated the success of maintenance regimens with oral ketoconazole 100 mg daily,12 clotrimazole 500 mg suppositories13 once-weekly and single oral fluconazole 100 mg14 in reducing the frequency of symptomatic episodes of vulvovaginal candidiasis. Recently, Sobel et al.9 published a randomised, multicentre trial using a maintenance regimen of oral fluconazole. However, whatever the chosen regimen, cessation of therapy is followed by relapse in about 50% of women within 6 months.9,12–14

Itraconazole is a well-tolerated antifungal agent of the azole group with activity against nearly all yeasts and is effective against vulvovaginitis caused by Candida.7,15 We chose to assess a prophylactic maintenance regimen using itraconazole once a month shortly before ovulation to produce high antimycotic levels when estrogen levels are highest, that is, at the time of ovulation. This strategy was based on the assumption that estrogen enhances the production of glycogen by vaginal epithelial cells. Glycogen in turn breaks down into glucose, which acts as a substrate for bacteria and yeasts. Thus, antimycotic treatment may be most effective during this time of the menstrual cycle.

Lactobacilli are major constituents of the normal vaginal flora and are thought to inhibit pathogen colonisation by the production of bactericins, lactic acid and hydrogen peroxide. Hilton et al.16 tested a yogurt product containing Lactobacillus acidophilus taken orally every day for a period of 6 months and found that the number of infections was significantly lower in patients receiving the yogurt-containing diet. The same study group also reported subjective improvement with the 7-day use of vaginal suppositories impregnated with 109 L. GG organisms, a variant of L. casei, in an open trial of women with RVCC.17 As the administration of lactobacilli may confer an added benefit in the treatment of RVVC, we included a second treatment arm using a L. acidophilus agent containing 1 × 1010 to 1 × 1011 CBU/g administered intravaginally monthly for 6 days, additional to itraconazole.

In this study, itraconazole was compared with classic homeopathy (CH). CH is a form of alternative medicine, claiming that diluting and succussing causes a toxic remedy to lose its toxic effect and win a deeper action of healing. The hypothesis is that remedies capable of producing symptoms in healthy persons will cure similar symptoms in ill people (simile rule). Although controversial, CH is widely used as an alternative treatment for numerous illnesses, which include among them cancer, pain, depression and chronic infections.18

We performed a prospective, randomised, single centre study to compare the clinical and mycologic efficacy of menstrual-cycle dependent itraconazole with and without lactobacilli and CH in reducing the frequency of clinical episodes of RVVC over a study period of 12 months.

Methods

This study was performed with the approval of the Ethics Committee of University of Vienna Medical School (RIB-no. 134/2000) and took place in our outpatient department between 2000 and 2006. Written informed consent was obtained from each patient. All patients were at least 18 years old, had had at least four episodes of Candida vaginitis in the year before study inclusion and complained of symptoms of acute Candida vaginitis at first presentation to our department. Exclusion criteria were pregnancy, mixed infections, infection with Candida glabrata or Candida krusei and a positive HIV or hepatitis status. Education years were noted by self-evaluation of the patients.

If both microscopical examination and culture were Candida albicans positive, patients were assigned to three study groups, that is, itraconazole, itraconazole followed by lactbacilli and CH, using a computerised randomisation list. For the treatment of acute infection, patients in the itraconazole groups received an induction regimen consisting of a single-day treatment of 200 mg bid itraconazole (Sporanox®; Janssen-Cilag, Vienna, Austria). After their next menstruation, patients continued with the same single-day treatment twice weekly, usually taken every Tuesday and Friday until their next menstrual period occurred. Within 1 week after menstrual bleeding, patients were requested to return for another visit for the initiation of itraconazole maintenance therapy. Patients in the itraconazole group continued with the single-day treatment of 200 mg bid itraconazole once a month and patients in the itraconazole and lactobacilli group received the same regimen plus one tablet of a vaginal lactobacilli (containing 2 × 108 to 2 × 109Lactobacillus gasseri-lyophilisates; Döderlein med®; Gebro Pharma, Fieberbrunn, Austria), to be taken every evening for six consecutive days. In both groups, the maintenance therapy continued for 6 months, with follow-up visits scheduled every month. Thereafter, patients were followed without treatment for another 6 months, with bimonthly follow-up visits (Figure 1).

Figure 1.

 Treatment algorithm.

Patients in the CH group did not receive any antimycotic induction or maintenance treatment and were treated with high potencies of a single homeopathic remedy according to each patient’s condition based on the simile rule throughout the entire 12-month study period. They were also followed monthly during the first 6 months and bimonthly for another 6 months. CH treatment was provided by a licensed CH practitioner (MB). Specifically, a personal history was taken and an individualised treatment scheme was prescribed. The most often used homeopathic remedies were carcinosin M, nux vomica, pulsatilla M, ferrum metallicum and sepia M. Potencies of homeopathic remedies ranged from C 30 to C 1000.

At every follow-up visit, a clinical history was taken and a vaginal smear for microscopic evaluation and culture was obtained in all patients. The vaginal swabs were sent to the Department of Clinical Microbiology, processed on Sabouraud’s dextrose agar and identified at the species level. Patients’ RVVC-associated subjective complaints were assessed with seven questions using a 100-mm visual analogue scale (VAS) as presented in Table 1. Patient satisfaction with the treatment received was evaluated with two questions using a 100 mm VAS (Table 1).

Table 1.   Patients’ disease-associated subjective complaints (questions 1–7) and patients’ satisfaction with the treatment received (questions 8 and 9)
1. How strong was your vaginal pain within the last month?
2. How strong was vaginal itching within the last month?
3. How strong was vaginal burning within the last month?
4. How strong was the vaginal discharge within the last month?
5. How did the infection affect sexual intercourse within the last  month?
6. How did the infection disturb sportive activities within the last  month?
7. How did the infection disturb social activities (shopping, leisure  activities, etc.) within the last month?
8. How satisfied are you with the medication?
9. How satisfied are you with the medical attendance?

Statistics

A power calculation demonstrated that, with a sample size of 150, the study has a power of >95% to detect a 30% difference in treatment efficacy at a significance level of 0.05 using the Yates correction factor based on published efficacy rates of antimycotics in women with RVVC.12 Chi-square was used for comparison of frequencies and cross-tabulations and Students t-test was used on means. Descriptive statistics (means, SD and ranges) were calculated for demographic data and for total scores of the VAS. Differences were compared using chi-square and configuration frequency analysis. A Kaplan–Meier curve was constructed for evaluating the time to culture-free and time to relapse. Log-rank test was used to compare the three treatment groups, that is, itraconazole, itraconazole and lactobacilli, and CH. We performed univariate and multivariate linear regression models with VAS scores as the dependent variable and patient age and education (<5 school years versus >5 school years) as the independent variables. We used the statistical software SPSS 11.0 for Windows (SPSS 11.0, SPSS Inc., Chicago, IL, USA) for statistical analysis.

Results

Five hundred-and-fifty women were screened for this study and a total of 150 Caucasian patients were enrolled during January 2000 and March 2006. Enrolled women were randomised into 3 groups: itraconazole without (Group 1) and with (Group 2) local lactobacilli and CH (Group 3) (Consort diagram—Figure 2). Median patient age was 30.4 (range, 17 to 56) years. After randomisation, two women in the itraconazole group and four women in the CH group dropped out of the study because of diarrhoea and pregnancy and use of co-medication with antimycotics respectively. Twenty-five women in the itraconazole group, 25 women in the itraconazole + lactobacilli group and 23 women in the CH group were lost to follow up or withdrew from the study. Seventy-one women completed all 12 months of the study period (Table 2). Therefore, the per-protocol (PP) analysis was based on data of 71 women.

Figure 2.

 Consort diagram.

Table 2.   Number of patients free of culture-detectable Candida
VisitItraconazoleItraconazole +  lactobacilliCH
  1. CH, classic homeopathy.

  2. *Treatment of acute episode of vulvovaginal candidiasis.

  3. **Start of maintenance phase.

  4. ***Monthly visits.

  5. ****Start of observation phase.

  6. *****Bi-monthly visits.

Randomisation0/48 0/50
Visit 2*28/48 (58%)32/50 (64%)
Visit 3**40/47 (85%)44/49 (90%)0/46
Visit 4***34/42 (81%)34/48 (71%)22/46 (48%)
Visit 5***31/40 (78%)32/47 (68%)20/43 (47%)
Visit 6***28/38 (74%)31/46 (67%)15/42 (36%)
Visit 7***21/32 (66%)30/45 (66%)9/40 (23%)
Visit 8****19/31 (61%)30/45 (67%)15/39 (38%)
Visit 9*****20/29 (69%)24/38 (63%)15/33 (45%)
Visit 10*****20/27 (74%)25/34 (74%)14/30 (47%)
Visit 11*****18/23 (78%)19/25 (76%)9/23 (39%)

Figure 3 shows the Kaplan–Meier curves comparing the time from study entry to initial treatment success, that is, free of candida in vaginal culture, in the itraconazole, itraconazole plus lactobacilli and the CH groups. This difference was statistically significant (log-rank test; P < 0.0001) with women in the itraconazole and the itraconazole plus lactobacilli groups reaching a culture-free status earlier than women in the CH group. Specifically, before the start of the maintenance regimen i.e. after a median time of 45 days, 40 of 47 women (85%) in the itraconazole group and 44 of 49 women (89.8%) in the itraconazole plus lactobacilli group were free of Candida detectable by culture. Table 2 shows the proportion of Candida-negative women throughout the study period.

Figure 3.

 Kaplan–Meier curve representing time from initial treatment to a status of candida-free vaginal culture in the three study groups. The full line represents women in the itraconazole group, the dashed line represents women in the itraconazole plus lactobacilli group and the dotted line represents women in the classic homeopathy group.

To start the twice weekly regimen at the same day of the menstrual cycle we chose a ‘pre-treatment’ to reduce complaints at the beginning. Because of the different time interval between the first day of treatment and the start of the more aggressive part of the induction regimen, we found a relatively high proportion of culture positive women at visit 2 (Table 2).

At the end of the maintenance regimen, that is, 6 months before the end of the study, 19 of 31 women (61%) in the itraconazole group, 30 of 45 women (67%) in the itraconazole plus lactobacilli group were free of culture-detectable Candida. 15 of 39 (38%) women in the CH group were free of culture-detectable Candida at the end of the study.

Figure 4 shows women who reached a culture-free status after initial treatment and compares the time from culture-free status to recurrence of culture-detected Candida recurrence. Women in the CH group had a recurrence of vaginal Candida infection sifnificantly earlier compared with women in the itraconazole and itraconazole plus lactobacilli groups (log-rank test; P = 0.002).

Figure 4.

 Kaplan–Meier curve representing time from a status of candida-free vaginal culture to recurrence of culture-detected vaginal Candida infection in the three study groups. The full line represents women in the itraconazole group, the dashed line represents women in the itraconazole plus lactobacilli group and the dotted line represents women in the classic homeopathy group.

At the end of the study period, that is, after 12 months, 18 of 23 (78%) women in the itraconazole group, 19 of 25 (76%) women in the itraconazole + lactobacilli group and 9 of 23 (39%) women in the CH group were free of culture-detectable Candida.

Interpretation of the median VAS results of the seven items regarding RVVC-associated complaints shows that women in the CH group had a significantly higher level of discomfort than women in the itraconazole group and the itraconazole plus lactobacilli group (36.8, 25.1 and 27.7 respectively; P < 0.001). Also, women in the CH group were significantly less satisfied than patients in the itraconazole group and the itraconazole plus lactobacilli group (59.2, 68.2 and 71.7 respectively; P < 0.001) based on the two items regarding patient satisfaction. In a general linear model, analysis of the VAS results of the seven items regarding RVVC-associated complaints showed a significant influence of Candida colonisation, but not age and education on the value of VAS scores. In accordance, a linear regression analysis with dichotomed long-term success of treatment as dependent variable showed no influence of age and education.

During the monthly follow-up visits, vaginal swabs were also analysed for accompanying bacterial colonisation. No significant differences were found between the three treatment groups (data not shown). The same was true when Candida-negative and Candida-positive patients were compared. In both of these patient subgroups, the most common isolates were ureaplasma (35.6% versus 36.9%; P = 0.7), group B streptococci (21.8% versus 21.2; P = 0.9), coagulase-negative staphylococci (22.5% versus 21.2; P = 0.7) and Gardnerella (15.8% versus 21.2; P = 0.1). Interestingly, no significant differences were found with regard to the colonisation with lactobacilli when Candida-positive and Candida-negative women were compared (90.6% versus 92.2%; P = 0.5).

An adverse event thought to be attributable to itraconazole occurred in two patients and led to discontinuation of treatment, that is, a local allergic reaction and diarrhoea. No adverse events were recorded in the CH group.

Discussion

In this prospective, randomised, single-centre trial, we found that monthly cycle-dependent itraconazole is more effective than CH in the treatment of RVVC. Lactobacilli do not confer an added benefit to itraconazole treatment. Women using itraconazole had a vaginal culture free of Candida infection significantly earlier than women treated with CH and the time to recurrence of vaginal Candida infection was significantly longer in women using itraconazole. In our study, monthly itraconazole for a period of 6 months and a subsequent observation period of another 6 months achieved a cure rate of 77%. This reduction compares favourably with what has been described in previous studies.9,12–14

The results of our study suggest that the maintenance regimen investigated was successful in our study population and could be possibly successful when confirmed in a larger series of patients. They also show that the additional use of local Lactobacillus preparations does not confer an added benefit in terms of improved Candida eradication, which is why application of additional local measures is not recommended. We are not able to explain the lack of effectiveness of lactobacilli, but it might be also because of an ineffective colonisation of this special types, because there was no difference in the amount of vaginal Lactobacillus between the 3 groups during the study. CH was significantly less effective than itraconazole in reducing the frequency of RVVC.

The rationale of administering the antimycotic agent in a menstrual cycle-dependent fashion, that is, after menstruation, is to counteract the high vaginal glucose concentrations present at the time of ovulation. In view of both its pharmacokinetics and its long half-life, administration of itraconazole in the second week of the menstrual cycle allows adequate antimycotic tissue levels to be reached at the time when they are most needed.

Whereas the long duration of our study emphases the difficulty of including ‘true’ cases of RVVC, the high drop-out rate indicates that it is equally challenging to guide and follow patients with this disease and accompanying physical and psychological stress through a 12-month study. In our study, we observed a high discontinuation rate. Lack of treatment compliance is a well-known clinical problem in women with RVVC. This shortcoming has to be acknowledged when interpreting the results of our study. Based on our results, treatment discontinuation was not automatically followed by treatment failure. Rather, as our study demonstrates, more than 50% of women were Candida-free at the time they discontinued treatment. Overall, despite the high discontinuation rate, results remained highly significant throughout the study. Moreover, the costs of medication of once-monthly itraconazole are lower than those of maintenance regimens used in earlier studies, particularly those with once-weekly antimycotic treatment.

Our study has limitations. The high drop-out rate may dilute the effect of both itraconazole and CH treatments. Also, we cannot rule out that the effect seen in the CH group with a success rate of 39% is a placebo effect. Our study also has strengths, which include among them the randomised study design, the high number of patients and a long study duration of 12 months.

In conclusion, our results indicate that RVVC can be successfully controlled by a single-day treatment of 200 mg bid itraconazole given in the second week of the menstrual cycle over a period of 6 months, achieving long-term vaginal eradication of Candida organisms in 77% of patients.

Disclosure of interests

The author and all co-authors have no conflict of interest in connection with this paper.

Contribution to authorship

Armin Witt coordinated the study and wrote the paper. All investigators participated in the study design, data collection, analysis, interpretation and writing of the paper.

Details of ethics approval

This study was performed with the approval of the Ethics Committee of University of Vienna Medical School (EK-Nr. 134/2000).

Funding

None.

Acknowledgements

None.

Ancillary