DOI: 10.2337/dc06-1469 © 2007 by the American Diabetes Association
Prevalence of Candida glabrata and Its Response to Boric Acid Vaginal Suppositories in Comparison With Oral Fluconazole in Patients With Diabetes and Vulvovaginal Candidiasis
1 Department of Endocrinology and Metabolism, All India Institute of Medical Sciences, New Delhi, India Address correspondence and reprint requests to Dr. Ravinder Goswami, MD, DM, Associate Professor, Department of Endocrinology and Metabolism, All India Institute of Medical Sciences, New Delhi 110029, India. E-mail: Ravinder Goswami at gosravinder{at}hotmail.com or Uma Banerjee at uma_banerjee{at}hotmail.com
OBJECTIVEA large proportion of vulvovaginal candidiasis (VVC) in diabetes is due to nonalbicans Candida species such as C. glabrata and C. tropicalis. Observational studies indicate that diabetic patients with C. glabrata VVC respond poorly to azole drugs. We evaluated the response to oral fluconazole and boric acid vaginal suppositories in diabetic patients with VVC. RESEARCH DESIGN AND METHODSA total of 112 consecutive diabetic patients with VVC were block randomized to receive either single-dose oral 150-mg fluconazole or boric acid vaginal suppositories (600 mg/day for 14 days). The primary efficacy outcome was the mycological cure in patients with C. glabrata VVC in the two treatment arms. The secondary outcomes were the mycological cure in C. albicans VVC, overall mycological cure irrespective of the type of Candida species, frequencies of yeast on direct microscopy, and clinical symptoms and signs of VVC on the 15th day of treatment. Intention-to-treat (ITT; n = 111) and per-protocol (PP; n = 99) analyses were performed. RESULTSC. glabrata was isolated in 68 (61.3%) and C. albicans in 32 (28.8%) of 111 subjects. Patients with C. glabrata VVC showed higher mycological cure with boric acid compared with fluconazole in the ITT (21 of 33, 63.6% vs. 10 of 35, 28.6%; P = 0.01) and PP analyses (21 of 29, 72.4% vs. 10 of 30, 33.3%; P = 0.01). The secondary efficacy outcomes were not significantly different in the two treatment arms in the ITT and PP analyses. CONCLUSIONSDiabetic women with C. glabrata VVC show higher mycological cure with boric acid vaginal suppositories given for 14 days in comparison with single-dose oral 150-mg fluconazole.
Abbreviations: HVS, high vaginal swab ITT, intention to treat PP, per-protocol VVC, vulvovaginal candidiasis
Patients with diabetes are at increased risk of developing vulvovaginal candidiasis (VVC) (18). Unlike nondiabetic women, these patients have a higher proportion of colonization/infection due to nonalbicans Candida species such as C. glabrata and C. tropicalis (4,68). In contrast, C. albicans VVC is more frequent in nondiabetic women (9). Moreover, diabetic patients with C. glabrata VVC respond poorly to single oral doses of 150 mg fluconazole (8). C. glabrata demonstrates intrinsically reduced susceptibility to azole drugs (1013). Redondo-Lopez et al. (10) and other investigators have reported the successful use of vaginal boric acid suppositories in the management of patients with C. glabrata (10,14,15) as well as C. albicans VVC (1618). In view of the higher frequency of C. glabrata VVC among diabetic women and their poor response to fluconazole, there is a need for a therapeutic regimen that is effective against both C. glabrata and C. albicans. To the best of our knowledge, there is no previous study reporting the effect of boric acid vaginal suppositories in diabetic women with VVC. In the current open-label, randomized trial, we report the response to oral fluconazole and boric acid vaginal suppositories in diabetic patients with VVC.
This was a randomized, open-label comparison of 2 weeks duration between single-dose oral 150-mg fluconazole and 14 days of treatment with boric acid vaginal suppositories (600 mg/day) in 112 consecutive patients with diabetes (type 2 diabetes, n = 77 and type 1 diabetes, n = 35) and VVC conducted during 20042006. Diabetes was diagnosed as per the American Diabetes Association Expert Committee criteria (19). Patients already on insulin or oral hypoglycemic agents also were included. Any patient with onset of diabetes before age 30 years who had received continuous insulin treatment since diagnosis was considered to have type 1 diabetes. Subjects excluded were those who were pregnant, sexually inactive girls, those aged >65 years, those with renal insufficiency (serum creatinine >1.8 mg/dl), and those on steroid therapy. Patients who did not give consent for pelvic examination, those treated for vaginal discharge during the past 3 months, and symptomatic patients in whom Candida growth was not detected on fungal culture were also excluded. A1C was measured in all the subjects to assess their glycemic control (normal range: 5.47.0%) at visit 1. The institutional ethics committee of the All India Institute of Medical Sciences approved the study protocol, and all the subjects were examined after their informed consent.
Diagnosis of VVC Detection and identification of the yeast up to the species level was done in the mycology laboratory of the All India Institute of Medical Sciences as per the standard protocols (21,22). The criteria for laboratory diagnosis were 1) direct demonstration of the yeast by Grams staining and 2) culture on Sabourauds dextrose agar supplemented with 2 mg/ml gentamicin, with and without cyclohexamide (0.5%). Identification of Candida species was performed by examining colony morphology, germ tube test, hyphal morphology, and chlamydospore formation on cornmeal agar, triphenyl tetrazolium reduction test, fermentation and assimilation of different sugars, and cycloheximide susceptibility.
Randomization and concealment A statistician generated the entire randomization sequence list in advance, and allocations were sequentially numbered from the beginning to end. A female clinical investigator (D.R.) carried out genital examinations for VVC, including an HVS for fungal culture for all consecutive diabetic patients attending the endocrine outpatients clinic and gave a serial number to all the patients examined. Following a report of yeast growth, two clinicians (D.R. and R.G.) assigned patients to a treatment arm, as per the serial order of the clinical examination and randomization list sequence number, in an irreversible manner. One of these two clinicians (R.G.) was not involved with clinical examination. Patients were allocated to treatment arm following reports of yeast growth but before the results of Candida species were available. The serial order of the randomization list was not disclosed to the patients until the treatment was assigned. The intake of fluconazole was supervised to ensure compliance. Patients receiving boric acid were instructed to insert one boric acidfilled gelatin capsule (600 mg) intravaginally every night for 14 uninterrupted days either by a plastic applicator or dispensed by hand as per their preference. Two extra capsules were given, and unused capsules were counted on the follow-up visit to assess compliance. Patients in both treatment arms were reassessed on the 15th day of therapy (visit 3). To minimize interobserver variation, the same investigator carried out the repeat clinical assessment at visit 3 when HVSs were taken again for demonstration of yeast by direct microscopy and fungal culture. All the HVSs were of the same color and size and sent to the mycological laboratory in identical containers labeled with a numerical identity. The plan of block randomization, patients treatment assignments, the pre- and postnature of the specimen and clinical findings were not disclosed to the mycologist and the laboratory staff who performed fungal culture and species identification till the completion of the 2-year study.
Efficacy outcomes
Adverse effects
Sample size and power calculations
Statistical analysis
Of 112 subjects randomized to the two treatment arms, 111 were included in the ITT analysis and 99 were analyzed in the PP analysis. One patient was excluded from the ITT analysis because she fell into exclusion criteria (sexually inactive, in whom a HVS was not taken). Twelve patients were dropped for PP analysis because of the following reasons: two stopped boric acid therapy due to vaginal burning sensation appearing on the 7th day of therapy, two did not come for follow-up because of diabetic foot and pulmonary consolidation (fluconazole arm), and eight patients were lost to follow-up for unknown reasons (four in each of the two treatment arms). The most common species isolated in the cohort of 111 patients included in the ITT analysis was C. glabrata (n = 68, 61.3%), followed by C. albicans (n = 32, 28.8%) and C. tropicalis (n = 4, 3.6%). The baseline demographic and clinical characteristics, including the proportion of patients with type 1 and type 2 diabetes, frequencies of C. glabrata and C. albicans species isolated on HVS culture, number of postmenopausal women, and those using commercially available protective sanitary napkins, were comparable in both the treatment arms in the ITT and PP analyses groups (Table 1). Only two women were on oral contraceptive pills (one in each of the two treatment arm). The mean A1C value at baseline was lower in the boric acid treatment arm than in the fluconozaole arm in the ITT as well as in the PP analyses group. The primary and secondary efficacy outcomes were, therefore, adjusted for A1C at baseline (Table 2).
The mycological cure rate in patients with C. glabrata VVC was significantly higher in the boric acid treatment arm compared with the fluconazole arm in both the ITT and PP analyses (P values = 0.01 for both after adjustment for baseline A1C). The odds in favor of mycological cure with boric acid vaginal suppositories in diabetic patients with C. glabrata VVC was 4.0 (95% CI 1.411.6). The goodness of fit of the logistic regression model was assessed by Nagelkerke R2. Although the Nagelkerke R2 value for primary outcome was only 15.3%, the model should not be interpreted as inadequate (23). No significant difference in mycological cure was observed in the two treatment arms in subjects with C. albicans VVC. The overall mycological cure on the 15th day, when analyzed irrespective of the type of Candida species isolated at baseline HVS culture, tended to be higher in those receiving boric acid in the ITT (P = 0.07) and PP (P = 0.06) analyses. The improvement in vaginal pruritus, discharge, and other clinical signs and symptoms was comparable in the two treatment arms in the ITT or PP analyses. The duration of different symptoms and signs related to VVC, yeast positivity on direct microscopy, and mean A1C values were comparable in diabetic subjects who grew C. albicans and C. glabrata in the ITT and PP analyses groups. The mean A1C tended to be higher in subjects who continued to grow Candida on repeat HVS culture compared with those who did not grow Candida in the ITT (9.3 ± 2.0% vs. 8.6 ± 2.1%; P = 0.11) and PP groups (9.5 ± 2.2% vs. 8.6 ± 2.1%; P = 0.06).
Adverse effects
The 59.9% prevalence of C. glabrata infection observed in the current study confirms the findings of our earlier studies that non-albicans VVC is frequent in diabetic women (4,8). de Leon et al. (6) observed the 54% vaginal carriage rate of C. glabrata in type 2 diabetes. The comparable frequency of clinical symptoms and signs between diabetic women with C. glabrata or C. albicans infection is similar to that reported by Geiger et al. (24) in nondiabetic women. Increased prevalence of C. glabrata infection in diabetic women has clinical relevance because poor therapeutic response and innate resistance to azoles has been reported for C. glabrata VVC in nondiabetic women (1013). Similar information is lacking in diabetic subjects, as they are often excluded in antifungal efficacy studies (8,10,2527). Poor mycological cure in diabetic women with C. glabrata VVC to single-dose oral 150-mg fluconazole, observed in the current study, is in accordance with our earlier case-control study (8). Mechanisms implicated for resistance of C. glabrata to fluconazole include increased fungal ergosterol synthesis and up to eightfold higher expression of azole efflux pump protein coded by CgCDR1 transporter genes (26,28). Diabetic patients with VVC also respond poorly to itraconazole and ketoconazole (29,30) In the current study, the higher mycologoical cure (72.4%) to boric acid therapy in diabetics with C. glabrata VVC is similar to that reported in nondiabetic individuals (10,1418). In 1974, Swate and Weed (16) first reported boric acid to be a "safe, effective and inexpensive form of therapy" for VVC. Sobel and Chaim (14) reported clinical improvement in 81% and mycological cure in 77% of nondiabetic patients with symptomatic C. glabrata/Torulopsis vaginitis treated with vaginal boric acid. There are also reports of better therapeutic response to boric acid compared with topical nystatin and miconazole in nondiabetic women with C. albicans VVC (17,18). Otero et al. (31) have reported a greater in vitro susceptibility of C. albicans to boric acid compared with C. glabrata. Boric acid or boracic [B(OH)3] is a weak acid, and its mode of antifungal action is not clear. Shiohara and Tasker (15) proposed that its acidic properties lead to disruption of the fungal cell wall. The pH of boric acidSabouraudss broth (5.05.09) required for in vitro inhibition of C. albicans growth has been reported to be similar to the vaginal pH in untreated Candida vaginitis, indicating that the effect of boric acid is not specific to its acidic properties (17). Alkaline compound (i.e., Na2B4O3 [borax]) also has antiseptic properties (32). The doses of the boric acid used to treat VVC have ranged from 600 mg daily for 7 days (16) to twice daily for 14 days (18). We used 600 mg boric acid daily for 14 days because >90% mycological cure has been reported with this schedule in nondiabetic patients with symptomatic C. albicans (17) or C. glabrata VVC (14). Boric acid therapy (600 mg/day for 14 days) has been found to be safe (14,17), except for local burning sensation and vestibular erythema in occasional patients (14,17,18). Two of our patients also reported this side effect. Improvement in clinical features of VVC was comparable in two treatment arms. However, we cannot exclude an element of bias in the analysis of this secondary outcome, as the same investigator carried out the clinical examination following treatment. In our previous studies, poor glycemic control was linked with Candida growth (4) but was unrelated to the type of species isolated or response to the fluconazole therapy (8). In the current study, there was a trend of higher mean A1C levels in those showing persistence of Candida growth following either of the two therapies. The cause of increased C. glabrata isolation in diabetic women is not clear but may involve frequent use of antifungal drugs leading to its reduced susceptibility to azoles (33) and consequent polarization/homing in diabetic women. Feng et al. (34) reported lesser susceptibility of C. glabrata in comparison with C. albicans to ß-defensins, natural cationic antimicrobial/antifungal peptides expressed in human epithelia. In diabetic milieu, ß-defensins expression is reduced (35). Reduced expression of defensins in association with resistance of C. glabrata to fungicidal activity of drugs like fluconazole may also explain the high prevalence of C. glabrata VVC in diabetic women. One of the relevant findings in this study is the importance of species identification of Candida isolates for proper management of VVC in diabetic women. It is emphasized that conventional inexpensive testing procedures, though time consuming, are good enough to identify Candida up to species level and would help in the management of diabetic women with VVC. As an alternative, boric acid therapy could be considered as the frontline therapy for treating VVC in diabetic women because it is effective against both C. albicans and C. glabrata compared with fluconazole, which is effective against C. albicans only.
The authors are thankful to the financial support by ICMR, New Delhi, for this study.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C Section 1734 solely to indicate this fact. Received for publication July 18, 2006. Accepted for publication October 22, 2006.
This article has been cited by other articles:
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||