Abstract
For practitioners, selecting successful therapy for vulvovaginal candidiasis is anything but trivial. The dominant problem; however, lies not with selecting the correct antimycotic agent, but with making the correct diagnosis and not treating non-yeast infections and noninfections as yeast-induced. Moreover, not all cases of vulvovaginal candidiasis are equal and practitioners owe patients the obligation of selecting appropriate therapy based upon the specific type and severity of vulvovaginal candidiasis. Uncomplicated candidiasis is readily treated with short-term oral or topical therapy, whereas complicated candidiasis needs additional strategies using the plethora of therapies available. Problematic refractory cases still abound and the pipeline for new, more potent antifungal agents is largely empty. Management strategies for complicated Candida vaginitis are discussed.
Vulvovaginal candidiasis (VVC) also called Candida or yeast vaginitis, is widespread, found worldwide and treated predominantly by primary care practitioners and specialist gynecologists, but also by virtually all categories of physician. The track record for appropriate therapy is anything but impressive. This occurs primarily because VVC is trivialized and seen as a nonserious, transient maladie that should respond to any form of antimycotic therapy. The initiative for providing therapy has now passed to women themselves, with ready access to a wide variety of topical antimycotic agents without simultaneous provision of essential diagnostic aids. Accordingly, the frequently blind empiric approach of practitioners is now perpetuated by women suffering from vulvovaginal symptoms. In the majority of episodes of VVC, short-term antifungal therapy without much additional consideration will suffice to rapidly resolve symptoms; however, complicated VVC provides additional challenges. This review provides insights into therapeutic options in selecting antifungal therapy.
Epidemiologic considerations
Definitive statistical information on the incidence of VVC is incomplete, particularly since vaginitis is not a reportable entity. Furthermore, the collection of data is hampered by inaccuracies of diagnosis. Unfortunately, VVC is routinely diagnosed in clinical practice without the benefit of microscopy or culture, and in as many as half the cases so diagnosed, the women may be uninfected or have other conditions [1]. Published epidemiologic data on incidence, where diagnostic data are based on definite clinical and mycologic findings, are exceptional [2]. Moreover, most studies have been carried out in sexually transmitted disease clinics, family planning or student health clinics, largely ignoring women seen by private practitioners, especially older women. Most studies suggest a VVC prevalence of 5–15% depending on the population studied. The availability of over-the-counter antifungal medications and the potential for their widespread abuse precludes any current and future epidemiologic studies [3].
VVC affects an estimated 70–75% of females at least once during their lives, and 40–50% will experience a recurrence of vaginitis [2–6]. A small subpopulation of undetermined size, probably approximately 5% of adult women, have recurrent, often intractable episodes of VVC. Diagnosis and therapy of VVC together with lost productivity result in an estimated cost of US $1 billion annually in the USA. Statistical data in the UK derived from patients whose conditions were diagnosed at genitourinary medicine centers show a sharp increase in the incidence of VVC from 118/100,000 to 200/100,000 women during the last decade [7]. In the USA, Candida is currently the second most common cause of vaginal infections, with bacterial vaginosis as the most common diagnostic entity [8,9]. Based on the number of prescriptions written to treat fungal infections between 1980 and 1990, the incidence of VVC almost doubled; these numbered approximately 13 million in 1990.
Point-prevalence studies indicate that Candida spp. may be isolated from the lower genital tract of approximately 10–20% (occasional studies have found the upper to be 55%) of asymptomatic healthy women without vulvovaginal symptoms [10,11]. These asymptomatic women are colonized with Candida organisms usually with low numbers of yeast organisms. Longitudinal studies indicate that most women will be found to be colonized with Candida, often transiently, with up to 90% of women being colonized at one time or another [12,13]. Most studies indicate that VVC is a frequent diagnosis among young women, affecting as many as 15–30% of symptomatic women visiting a clinician.
Microbiology
Between 85–95% of yeasts isolated from the vagina are C. albicans strains [14]. The remainder is usually the non-albicans Candida spp., the most common of which is C. glabrata. Non-albicans Candida spp. can induce vaginitis that is clinically indistinguishable from C. albicans and should never be ignored; moreover, they are often more resistant to therapy [14]. Rare anecdotal case reports of symptomatic vaginitis are reportedly caused by trichosporon spp., zygomycetes spp. and Saccharomyces cerevisiae.
It has been claimed, but not proven, that the percentage of non-albicans Candida vaginal infections has risen dramatically [15,16]. The use of single-dose, oral and topical regimens, together with the popularity of low-dosage azole maintenance regimens and the availability of over-the-counter antimycotics, have been blamed for the appearance of non-albicans spp., given their lesser susceptibility to azole agents. Three recent large multicenter studies failed to demonstrate any increase in non-albicans Candida spp. isolated [17–19]. The apparent increase in non-albicans Candida spp. may reflect increased awareness, performance of more cultures or a reporting bias from referral centers.
C. glabrata is by far the most frequent non-albicans cause of acute and chronic vulvovaginitis. Diabetes mellitus, especially when poorly controlled, is a major risk factor for C. glabrata vaginitis. In HIV seropositive women, isolation of C. glabrata and other non-albicans Candida spp. has been linked to fluconazole exposure [20]. Other infrequent causes of VVC include C. parapsilosis and C. tropicalis, although virtually every species has been associated with vaginitis. Recurrent VVC is usually the result of relapse caused by identical persisting strains rather than new infection caused by different strains of C. albicans.
Treatment of vulvovaginal candidiasis
Asymptomatic colonization
Asymptomatic vaginal colonization with Candida spp. is generally not treated, since colonization is common, microbiologic eradication is infrequent and there are no major sequelae associated with colonization in pregnant women.
Symptomatic vulvovaginal candiasis
Uncomplicated vulvovaginal candidiasis
The majority of women with VVC suffer from uncomplicated vaginitis (Table 1) characterized by sporadic, infrequent attacks of mild-to-moderate severity due to C. albicans. These attacks occur in healthy adult women without unique host predisposing factors. In contrast, approximately 10–20% of women suffer from complicated VVC in which attacks are either more severe, occur on a recurrent basis, or are due to non-albicans Candida spp. Patients with complicated Candida vaginitis more frequently have underlying host factors.
Uncomplicated and complicated vulvovaginal candidiasis.
Several highly effective topical azole agents are now available in a variety of formulations (Table 2) [21]. No strong evidence exists that any one formulation results in superior cure rates, neither is there convincing evidence of the superiority of any specific azole agent in the treatment of VVC [21]. Overall cure rates for topical or systemic azole agents, defined as the eradication of symptoms and mycologic negative cultures, are of the order of 80–95% [18,19]. Oral systemic azole agents achieve comparable, equivalent or marginally higher therapeutic cure rates and patients enjoy the convenience of oral administration, which eliminates local side effects and messiness [22]. On the other hand, oral azoles suffer the drawback of potential systemic toxicity, which has limited the use of ketoconazole, although this is a lesser consideration in prescribing itraconazole and fluconazole. Side effects of single-dose fluconazole (150 mg) tend to be mild and infrequent, including gastrointestinal intolerance, headache and rash [18,19].
Azole therapy of vulvovaginal candidiasis.
Oral systemic therapy.
Less frequently used today are nonazoles, including topical gentian violet, boric acid and nystatin suppositories. These older agents offer no advantage for C. albicans infections, except being less expensive, but they require more prolonged therapy and achieve lower cure rates (60–80%).
Several surveys have consistently shown that most women prefer the convenience of oral therapy [23]. Of the three oral azoles, ketoconazole is unlikely to gain widespread usage in VVC due to a rare but serious hepatotoxicity [24]. Itraconazole 200 mg twice daily (single day) and fluconazole 150 mg (single dose) are both highly effective triazoles, although neither achieves substantially higher cure rates than topical agents [21].
Oral agents cannot be expected to accomplish immediate local relief, hence severe local symptoms may necessitate adjunct topical treatment for the first 48 h. Oral azoles are contraindicated in pregnancy.
Over the years, conventional use of vaginal antimycotics has been remarkably safe and free of any untoward side effects. Certainly, no major morbidity or serious complications have been reported, although high-dose regimens of terconazole were associated with fever and flu-like symptoms that resulted in the withdrawal of this topical agent in these concentrations [25]. Mild-to-moderate vulvovaginal burning is an underestimated and not infrequent side effect of topical azoles. Accordingly, physicians should respect the preference of the patient when prescribing, both in relation to choice of route and the individual agent. Both itraconazole and fluconazole are extremely well tolerated with good safety profiles and are not comparable to ketoconazole with regard to risk of complications. Nevertheless, untoward drug interactions have been reported between astemizole and terfenadine (both commonly used antihistamines) and itraconazole. Accordingly, the ideal antimycotic agent continues to evade us. Among the major shortcomings of the current agents is their reduced efficacy in vaginitis caused by non-albicans Candida spp., as well as the potential for C. albicans strains to become resistant to fluconazole.
There has been a growing tendency to use shorter and shorter courses of both topical and oral agents [21,26]. Although justified on the basis of convenience and improved compliance, these new regimens have also been introduced to provide a marketing edge for one therapeutically equivalent agent over another. Nevertheless, single-dose therapy by any route is effective in uncomplicated vaginitis. Many of the single-dose drugs (e.g., vaginal clotrimazole 500 mg suppository and fluconazole 150 mg oral tablet) possess pharmacokinetic properties such that following a single administration of these agents, concentrations of the antimycotic persist in the vagina for up to 4 days [27,28]. Hence, single-dose therapy may be more than ‘single-day’ therapy.
It is of interest that a single dose of fluconazole 150 mg will suffice for acute VVC, whereas successful therapy for oropharyngeal candidiasis (OPC) due to an identical strain requires daily administration for at least 7 days. The explanation for this major difference underlies the basis for the remarkable VVC success rate observed with fluconazole. Firstly, fluconazole has a long half-life, persisting in vaginal but not oral secretions for 72–96 h. Of greater significance, Moosa and colleagues recently demonstrated potent synergy between fluconazole and acetic acid produced by vaginal bacteria. The documented synergy converted the anti-Candida fungistatic action of fluconazole to frank cidal activity, thereby enhancing in vivo potency [29].
Complicated vulvovaginal candidiasis
Severe vulvovaginal candidiasis
Although not specifically addressed in the majority of efficacy studies, a subpopulation analysis in a large prospective randomized fluconazole/clotrimazole study concluded that severity of vaginitis (signs and symptoms) independently influenced response to antifungal therapy [18]. In the same study, women with severe vaginitis found short-course therapy (single-dose fluconazole) less effective than more prolonged therapy with 7 days of clotrimazole. In a second prospective study examining short versus longer duration oral fluconazole, the latter regimen achieved significantly higher cure rates [19]. In a prospective study, 556 women with severe or recurrent VVC were randomly assigned to therapy with a single dose of fluconazole (150 mg) or two sequential doses given 3 days apart [19]. Severity of disease was based upon a scoring system involving degree of pruritus and physical signs (erythema, edema, excoriation and fissures). The two-dose regimen resulted in significantly higher clinical cure at evaluation on day 14 (94 vs 85%) and day 35 (80 vs 67%) in women with severe but not recurrent disease. Accordingly, regardless of route of therapy, severe vaginitis should not be treated with short-course azole therapy. Although not structurally defined, longer duration may be defined as the presence of therapeutic concentrations of the antifungal drug in the vagina for at least 7 days.
Recurrent vulvovaginal candidiasis
Prior to initiation of treatment, diagnosis of recurrent VVC (RVVC) must be confirmed by culture. Thousands of women carry the label of recurrent yeast infections when, in fact, recurrent symptoms are due to noninfectious etiologies such as allergic and hypersensitivity vulvitis. Following confirmation of VVC, every effort should be made to eliminate factors predisposing to VVC; however, in the majority of women, no reversible or correctable causal factors are present [30]. Secondary causes of RVVC and specific management are listed in Table 3.
Causes and management of recurrent vulvovaginal candidiasis.
Recommendation controversial and unsubstantiated
The majority of women with RVVC are infected with C. albicans that is highly sensitive to all azoles [31–33]. Whereas 90–95% of women with sporadic yeast infections are infected with C. albicans, this percentage does drop to 80–85% in some, but not all, studies of women with RVVC. On this basis, the initial step in managing recurrent infections is to identify the Candida spp. The approach detailed below is based upon the proven observation that recurrent vaginitis is due to a sensitive strain of C. albicans [33]. Species identification is sufficient to predict antifungal susceptibility, and resistance testing is unnecessary except when patients fail to clear the yeast, which is rare.
Initial antimycotic therapy requires an induction course of either oral or vaginal antimycotic therapy, which must be continued daily until the patient is completely asymptomatic and a culture-negative status has been achieved. In RVVC, failure to initiate a maintenance regimen will result in mycologic and clinical relapse of vaginitis in 50% of patients within 3 months [31].
Maintenance suppressive regimens include daily oral ketoconazole or itraconazole (100 mg) or once-weekly regimens of either 500 mg clotrimazole vaginal suppositories or fluconazole 150 mg orally. All maintenance regimens are likely to be effective in preventing breakthrough vaginitis provided that the suppressive agent is given frequently enough [32,33]. The superior safety profiles of fluconazole and clotrimazole have resulted in these two agents largely replacing ketoconazole and itraconazole suppressive prophylaxis. Similarly, weekly or daily topical vaginal azole agents are capable of reducing recurrence rate [34]. Whatever the maintenance regimen, cessation of therapy is accompanied by symptomatic relapse in half the women within a short time of discontinuing therapy [30–32]. Re-establishment of the azole maintenance regimen is then required and some women may require a long-term antifungal suppressive regimen.
The role of treatment of male sexual partners was reviewed and no benefit was demonstrated in several large studies [14]. Similarly, a study by Fong using systemic ketoconazole therapy to treat male partners failed to reduce the recurrence rate in women with RVVC [35]. Nevertheless, Spinillo and colleagues, in an uncontrolled study, did report decreased RVVC in women when attempts were made to eradicate all Candida sources in both partners [36]. Dennerstein reported a reduced rate of recurrence in RVVC in 15 patients during a 3-month period of depot medroxyprogesterone acetate therapy [37]. In a small study using patients as their own controls, Hilton and colleagues reported fewer episodes of VVC in women placed on oral yogurt [38]. Given the small numbers and lack of controls in this unblinded study, the role of yogurt in preventing Candida vaginitis remains unproven. Personally, the author has not found it necessary to recommend yogurt, since effective alternative methods are available in most Western countries. The majority of commercially available yogurt products do not contain viable lactobacilli. No evidence of human vaginal lactobacillus deficiency has been published.
An alternative approach to long-term maintenance antifungal therapy is the use of hyposensitization with Candida antigen preparation. Two small studies achieved encouraging results [39,40]. Although the lay press is replete with suggestions of alternative and home-made remedies none, including the probiotic therapy, have been validated as effective.
Vulvovaginal candidiasis due to non-Candida albicans species
Vaginitis caused by azole-resistant strains of C. albicans has been extremely rare, whereas fluconazole-resistant strains of C. albicans have emerged as a major problem with refractory OPC in advanced immunodeficient patients with AIDS [41]. It should be emphasized that azole-resistant strains of C. albicans are rarely the cause of RVVC, although in a recent study 4% of isolates of C. albicans from women with complicated VVC were infected with fluconazole-resistant C. albicans [19]. Thus, in vitro susceptibility tests are rarely indicated unless compliant patients have no response to adequate therapy. In contrast, RVVC is not infrequently due (10–20%) to non-albicans Candida spp., the majority of which show reduced susceptibility to all azoles [42]. Particularly common is C. glabrata, and approximately half the strains show marked reduced sensitivity to available azoles [43]. Boric acid (600 mg administered vaginally once daily in a gelatin capsule) has been shown to be highly effective in this clinically resistant infection [44]. Therapy should be continued until cultures are negative (usually 10–14 days), and when a history suggests RVVC, a maintenance regimen of alternate-day and then twice-weekly boric acid is prescribed. Unfortunately, there is still little published data on the efficacy of the boric acid maintenance regimen, nor has the long-term safety of intravaginal boric acid been confirmed. The only alternative to vaginal boric acid in resistant infections is 17% flucytosine cream; however, the latter agent is not available commercially and must be prepared by a compounding pharmacist [45].
Flucytosine for vaginal use should be limited due to the potential for the acquisition of resistance, hence long-term regimens are not recommended. An alternative for long-term use in women with recurrent infections not responding to boric acid or azole suppressive prophylaxis is to use nystatin 100,000 unit vaginal suppositories daily or on alternate days. Recently, Sobel and colleagues reported their experience with flucytosine in women failing boric acid, achieving a clinical response greater than 90% [44]. Similar good results have been seen with a vaginal compounded mixture of amphotericin B and flucytosine [46].
Vulvovaginal candidiasis in HIV-infected women
Despite the original descriptions of intractable Candida vaginitis in HIV-positive women, evidence of increased incidence of RVVC and vaginitis not responding to conventional antifungal therapy has not been forthcoming [47–53]. Similarly, although Spinillo reported a higher frequency of non-albicans Candida spp. in HIV seropositive women with RVVC [54], other studies have not shown an increased prevalence of non-albicans Candida species in HIV-infected women [51–53]. Thus, the clinical and microbiologic spectrum of VVC appears similar in HIV-positive and -negative women, and treatment principles should be identical to those for HIV-negative women. Schuman and colleagues demonstrated that suppressive prophylactic therapy with fluconazole (200 mg once weekly) effectively decreased recurrences of symptomatic VVC in HIV-positive women [51]. Nevertheless, despite the efficacy and safety of this regimen, primary prophylaxis is expensive, unnecessary and not recommended. One of the observations to emerge from this study was a tendency for C. glabrata to replace C. albicans as a colonizing organism [55]. To date, azole resistance among C. albicans vaginal isolates has been rare in contrast to OPC.
Whether Candida vaginitis with overt inflammation influences transmission of HIV and specifically increases susceptibility to HIV infection in unknown. In a recent prospective study of women in Tanzania, Kapiga and colleagues found that the presence of VVC at the beginning of the period of observation conferred a relative risk for HIV seroconversion of 1.98 (95% confidence interval [Cl]: 1.17–3.33) and remained an independent risk factor in multivariate analysis [56]. In a large pregnancy study, Burns and colleagues concluded that VVC was not associated with an increased risk of mother-to-infant transmission of HIV [53].
Conclusion
Candida vulvovaginitis is most commonly caused by C. albicans (>85%) with little evidence of an increase in vaginitis due to non-albicans Candida spp. Epidemiologic studies are no longer possible in developed countries in the era of self-diagnosis and self-treatment by women empowered by the availability of over-the-counter antimycotics. Classifying VVC into uncomplicated and complicated vaginitis has simplified choice and duration of antifungal therapy. Vaginitis due to C. albicans responds well to currently available therapy; in contrast, vaginitis due to C. glabrata is associated with a high failure rate. Candida vaginitis infection rates in HIV-seropositive women remain undetermined and reports of refractory VVC have not been substantiated. Despite the wide array of antifungal agents currently available, considerable limitations in available therapy exist in the effective management of complicated vaginitis.
Future perspective
Need for a diagnostic test for vulvovaginal candidiasis
There exists an enormous need for a rapid, inexpensive, point-of-care diagnostic test aimed at both practitioners and women themselves. Self-diagnosis is as inaccurate as empiric physician diagnosis, hence all are in need of a diagnostic device. Several prototypes are under study. Existing nonmicroscopy based tests, including DNA probes and polymerase chain reaction measurement are too complex, too expensive and too time-consuming. Diagnostic confusion results from the lack of specificity of signs and symptoms of VVC, together with poor microscopy skills on the part of practitioners. Point-of-care measurement of vaginal pH is also useful in identifying bacterial vaginosis, trichomoniasis and atrophic vaginitis, but does not distinguish Candida from the myriad of noninfectious causes with identical signs and symptoms.
Do currently available antifungals suffice or do we need additional agents?
For the most part, once a correct diagnosis is obtained and the patient placed in the correct category of complicated or uncomplicated VVC, more than 90% of episodes of VVC can be effectively handled using existing agents for the recommended duration. Of course, we can always aspire to have even more potent agents achieving more rapid relief of symptoms and vaginal yeast eradication. Unfortunately, the pipeline of new, yet to be introduced, antimycotics is almost entirely empty. There exists the perception in the pharmaceutical industry that no need exists for additional vaginal antimycotics; especially since fluconazole, the most successful anti-VVC drug ever, became generic in many countries. Accordingly, the only ongoing study of new antifungals is one of short-course itraconazole in a new enhanced oral formulation for acute VVC.
Executive summary
A worldwide problem, vulvovaginal candidiasis (VVC) continues to affect all strata of society. New figures of epidemiology are elusive given the tendency for practitioners to treat vulvovaginal symptoms empirically, together with empowerment of women to self-treat with over-the-counter antimycotics.
More than 90% of VVC episodes are caused by Candida albicans with little evidence of increased frequency of non-albicans spp., even in women with HIV and others with recurrent VVC.
Candida glabrata strains are frequently resistant to fluconazole and, overall, poor results are obtained with this species with all azole treatment regimens; nevertheless, overall azole resistance is uncommon and resistance in C. albicans strains is rare.
The clinical separation of the more common uncomplicated variety of VVC (90%) from women with complicated VVC (10%) has proven to be a useful and valuable therapeutic aid.
Uncomplicated VVC responds to short-course therapy with either topical or oral systemic azole drugs. Women should select and determine the route of therapy. The majority prefer the convenience of oral therapy, although relief may occur sooner with topical agents.
Complicated infections dictate longer duration of therapy, specifically for severe infections and in compromised hosts (pregnancy, diabetes etc.).
Complicated VVC due to non-albicans Candida spp. require special consideration. C. tropicalis and C. andidan arapsilosis infection, while uncommon, will respond to conventional azoles as long as the treatment is not abbreviated. However, C. krusei and especially C. glabrata often require boric acid or flucytosine therapy.
Complicated recurrent VVC requires attention to the underlying cause, but is controlled by long-term maintenance suppressive therapy with oral once-weekly fluconazole. Unfortunately, although the recurrence rate is extremely low (approximately 7%) while on the well-tolerated fluconazole, a high rate of recurrence can be anticipated with cessation of prophylaxis.
There is a need to treat C. glabrata infections as well as fungicidal agents to achieve cure rather than control in women with RVVC.
What of the newest antifungals used for systemic candidiasis?
The echinocandin class of antifungals, available only intravenously, would make ideal vaginal antimycotics, being broad spectrum and Candida-cidal. All these agents; caspofungin, micafungin and anidulofungin, are highly active against C. glabrata, C. krusei and fluconazole-resistant C. albicans and, therefore, ideal agents. Unfortunately, these drugs are not soluble and hence not available orally, and stability issues have precluded their topical use.
Of the newest generation of triazoles (voriconazole and posaconazole), little likelihood exists for their use in VVC. Both agents are available orally and have considerably broader spectrums of activity than fluconazole against all Candida spp. Even against fluconazole-sensitive C. albicans, these agents are approximately 30-100-times more active in vitro than fluconazole. However, despite their enhanced in vitro activity, this in no way guarantees superior in vivo clinical cure rates compared with fluconazole, given the latter agent's superior tissue penetration, prolonged tissue half-life and extraordinary safety profile. Hence the likelihood of new agents appears dismal. Fortunately fluconazole-resistant infections are rare in women with C. albicans, as are fluconazole-intolerant patients in whom fluconazole is contraindicated. New drugs are needed against C. glabrata and this appears to be the single biggest challenge.