Mucocutaneous candidiasis occurs in 3 forms in persons with HIV infection: oropharyngeal, esophageal, and vulvovaginal disease. Oropharyngeal candidiasis (OPC) was among the initial manifestations of HIV-induced immunodeficiency to be recognized (1,2) and typically affects the majority of persons with advanced untreated HIV infection. Presenting months or years before more severe opportunistic illnesses, OPC may be a sentinel event indicating the presence or progression of HIV disease.(3-5) Although usually not associated with severe morbidity, OPC can be clinically significant. Severe OPC can interfere with the administration of medications and adequate nutritional intake, and may spread to the esophagus.(6) Esophageal candidiasis remains one of the most common opportunistic infections in countries where combination antiretroviral therapy (ART) is a routine part of the standard of care.(7) Vulvovaginal candidiasis is an important concern for women with HIV infection, although the relationship of vulvovaginal candidiasis to HIV infection remains unclear.(8) In resource-poor nations, mucocutaneous candidiasis is a formidable problem.(9,10) Despite the frequency of mucosal disease, disseminated or invasive infections with Candida and related yeasts are surprisingly uncommon.
|Microbiology and Epidemiology|
Yeasts are fungi that grow as single cells and reproduce by budding. They are distinguished from one another on the basis of the presence or absence of capsules, their size and shape, the mechanism of daughter formation, the formation of true hyphae or pseudohyphae, and the presence or absence of sexual spores, along with physiologic data from biochemical testing. Candida albicans is the predominant causative agent of all forms of mucocutaneous candidiasis. Less frequently, C glabrata, C parapsilosis, C tropicalis, C krusei, and several other species may cause disease. C dubliniensis, a species that is phenotypically similar to C albicans, may cause approximately 15% of infections previously ascribed to C albicans.(11-14)
Candida are normal inhabitants of the human gastrointestinal (GI) tract and may be recovered from up to one third of the mouths of normal individuals and two thirds of those with advanced HIV disease.(15,16) Oral colonization with inherently drug-resistant organisms is more common in advanced HIV infection (CD4 lymphocyte counts <50 cells/µL).(16) Although vaginal colonization is more prevalent among HIV-infected women compared with HIV-negative women, no association of colonization rate with CD4 counts is apparent.(17) The majority of disease is caused by organisms that are part of the normal flora of an individual, although rare cases of person-to-person transmission have been documented.(18)
The individual Candida strains affecting persons with HIV infection are not different from those in other immunosuppressed hosts.(19) C dubliniensis is more commonly identified in HIV-infected persons, though it is currently indistinguishable from C albicans in its clinical presentation.(11-13) There are no detectable differences in the virulence of strains isolated from HIV-infected or HIV-uninfected persons. Recurrent disease can result from the same or from different species or strains of Candida.(20-23) The emergence of different strains or species is more likely in persons with lower CD4 lymphocyte counts and exposure to antifungal therapy.(19)
Oropharyngeal and vulvovaginal disease are the most common forms of mucocutaneous candidiasis. Up to 90% of persons with advanced untreated HIV infection develop OPC, with 60% having at least 1 episode per year with frequent recurrences (50-60%).(16,21,24-32) Esophageal candidiasis occurs less frequently (10-20%) but is the leading cause of esophageal disease.(33-35) Vaginal candidiasis has been noted in 27-60% of women, similar to the rates of oropharyngeal disease.(36-38) However, the incidence appears to be similar in HIV-infected and HIV-uninfected women.(8) Of note, 75% of all women of childbearing age develop vaginal candidiasis and 40% will have a second occurrence. Few women (<5%) experience frequent recurrences (defined as >=4 infections in a 12-month period).
The use of combination ART results in a significant decline in the incidence of a number of opportunistic illnesses (eg, Pneumocystis jiroveci pneumonia and cytomegalovirus).(39-41) Thus, in resource-rich nations, the incidence of mucocutaneous candidiasis has declined precipitously. For example, a significant difference has been found in the incidence of recurrent OPC in patients treated with protease inhibitors compared with those not treated with protease inhibitors (7% vs 36%).(42) Similarly, a decline in the incidence of OPC from 30% to 4% over a 1-year period was reported in persons on ART.(43) More frequent use of antifungal medications for secondary prevention also may have reduced the prevalence of mucocutaneous disease.
A number of factors are important in the development of mucocutaneous candidiasis.(44) The level of immunosuppression is paramount.(45) Other host factors important in the defense of Candida infections include blood group secretor status (such as presence or absence of specific Lewis antigens), salivary flow rates, condition of the epithelial barrier, antimicrobial constituents of saliva, presence of normal bacterial flora, and local immunity.(27,46) Several studies suggest an impairment in a number of anti-Candida host defense mechanisms in persons with HIV infection.(27,29,30) Higher levels of HIV-1 RNA in the plasma also have been associated with increased rates of mucocutaneous candidiasis and colonization with Candida.(42,47) Of note, the correlation between the level of immunosuppression and vaginal candidiasis may not be as strong. In a cross-sectional study of 833 HIV-infected and 427 HIV-uninfected women, the annual incidence of vaginal candidiasis in the 2 groups was similar (9%).(38)
Symptoms of OPC may include burning pain, altered taste sensation, and difficulty swallowing liquids and solids. Many patients are asymptomatic. Most persons with OPC present with pseudomembranous candidiasis or thrush (white plaques on the buccal mucosa, gums, or tongue) and less commonly with acute atrophic candidiasis (erythematous mucosa) or chronic hyperplastic candidiasis (leukoplakia, distinct from "hairy leukoplakia"; see chapter on Oral Manifestations) involving the tongue, or angular cheilitis (inflammation and cracking at the corners of the mouth).
Esophageal candidiasis usually is accompanied by the presence of OPC. Typically, dysphagia and odynophagia are described. In as many as 40% of patients with OPC, esophageal involvement may be asymptomatic.(6) Occasionally, esophageal disease may occur in the absence of clinically detectable oropharyngeal disease.
Vulvovaginal candidiasis generally presents with marked itching, watery to curdlike discharge, vaginal erythema with adherent white discharge, dyspareunia, external dysuria, erythema, and swelling of labia and vulva with discrete pustulopapular peripheral lesions. The cervix usually appears normal. Symptoms typically exacerbate the week preceding menses with some relief once menstrual flow begins. Vaginal candidiasis frequently is associated with pregnancy, high-estrogen oral contraceptives, uncontrolled diabetes mellitus, tight-fitting clothes, antibiotic therapy, dietary factors, intestinal colonization, and sexually transmitted disease. Specific additional risk factors for recurrent vulvovaginal candidiasis have not been identified.(48) Female-to-male transmission remains questionable, although male sexual partners may experience a transient rash, erythema, pruritus, or burning sensation of the penis minutes to hours after unprotected sexual intercourse. Occasionally, Candida balanitis may occur.
The diagnosis of OPC usually is made by its characteristic clinical appearance; recovery of an organism is not required. Oropharyngeal cultures often demonstrate Candida species, but alone are not diagnostic because colonization is common.(24) The diagnosis of OPC can be confirmed by examining a 10% potassium hydroxide (KOH) slide preparation of a scraping from an active lesion. Pseudohyphae and budding yeast are characteristic findings. The appearance of the lesion and presence of yeast forms on microscopic examination of the oropharynx are sufficient to confirm the diagnosis. A KOH preparation is not mandatory for diagnosing OPC. A presumptive diagnosis of OPC can be made by visual detection of characteristic lesions with resolution of those lesions in response to antifungal therapy. Culture usually is not necessary unless the lesions fail to clear with appropriate antifungal therapy. In patients with poorly responsive OPC, a culture should be obtained to look for inherently drug-resistant yeast or those that respond poorly to certain azoles (eg, C krusei or C glabrata). Clinicians should note that many microbiology laboratories report yeast cultures as either C albicans or non-albicans species based upon the germ tube test, and further characterization requires making a specific request. Biopsy of oral lesions rarely is helpful or indicated for the diagnosis of oral candidiasis.
A presumptive diagnosis of Candida esophagitis can be made in a patient with dysphagia and/or odynophagia who has OPC. Barium swallow or upper GI endoscopy can confirm a suspicion of esophageal involvement. These studies are not uniformly required, however, unless a patient fails to improve with appropriate systemic antifungal therapy.(35) If esophageal symptoms in a patient with OPC do not resolve despite resolution of the oral lesions, endoscopy is indicated to exclude other causes of esophagitis (eg, cytomegalovirus, herpes simplex virus, aphthous ulcers) in persons with HIV infection. The diagnosis of Candida esophagitis is confirmed by the presence of yeast forms on histologic examination of esophageal lesions. Cultures to look for drug-resistant yeast are warranted for patients who require endoscopy. Barium swallow rarely is indicated in HIV-infected patients with esophageal disease because it usually is not possible to determine the cause of an abnormality by its radiologic appearance alone.
The diagnosis of Candida vulvovaginitis is made by the presence of a characteristic clinical appearance and observation of yeast forms on microscopic examination. A KOH preparation of the vaginal discharge should be made to confirm the diagnosis of candidiasis and to differentiate from a number of other conditions that can be similar in appearance (eg, trichomoniasis). Because yeast are normal inhabitants of the vaginal mucosa, routine fungal cultures rarely are helpful when the KOH preparation is negative. A fungal culture should be obtained if a patient fails to respond to standard antifungal therapy.
Antifungal susceptibility testing has improved over the past few years but remains problematic. In 2002, the National Committee on Clinical Laboratory Standards (NCCLS) published standardized methods and definitions for in vitro susceptibilities to selected agents (Table 1).(49,50) The most common methods for in vitro testing are the macrotube and microtiter broth dilution assays. Alternative methods such as agar-based assays and flow cytometry are under evaluation.(51) In vitro susceptibilities should not be used routinely to guide the choice of antifungal agents because the correlation between clinical response and susceptibility testing is not clear. Despite the technical limitations, a number of studies have documented that in vitro resistance to antifungal medications is common.(16,52-64) The incidence of resistance varies widely in these studies. Overall, the rates of fluconazole resistance vary from 5% to 56%.(16,52-57) The rates of ketoconazole and itraconazole resistance have been reported less frequently but vary from 0% to 25%.(58-60) Amphotericin B resistance is extremely uncommon but has been reported.(64) Resistance to newer agents including echinocandins and triazoles (eg, voriconazole and posaconazole) is less common but has been documented.(65,66) Much of the variance in the rates of in vitro resistance can be explained by several factors: differences in the level of host immunosuppression, prior exposure to antifungal agents, the design of the study (longitudinal vs cross-sectional), the prevalence of non-albicans species of yeast, and differences in the in vitro methods used.
Several mechanisms may contribute to in vitro resistance to antifungals. Some yeasts have single-drug resistance, whereas others are multidrug resistant. Azole resistance has been demonstrated in yeasts that contain alterations in the enzymes that were the target of azole action or were involved in ergosterol biosynthesis. The cytochrome P450-dependent 14alpha-sterol demethylase (P450DM) and the delta5,6 sterol desaturase are enzymes that, when altered, result in azole resistance.(67,68) Reduced cell permeability is another mechanism of azole resistance.(69) Finally, active efflux of drug also has been observed.(70) The prevalence of these mechanisms, however, is unknown. Further, it is not clear whether certain mechanisms of resistance may be overcome by increasing the dosage of the drug.
A wide variety of agents are effective for the treatment of candidiasis (Table 2). Important factors that determine clinical response, besides the choice of antifungal agent, include the extent and severity of disease, patient adherence, and the pharmacokinetic properties of the drug. Treatment of OPC and vaginal candidiasis is relatively simple, with most types responding to therapy. Overall, randomized studies show little difference between topical and systemic therapy. Mild OPC or vulvovaginal disease often can be treated with topical therapy. Moderate and severe episodes typically require systemic therapy. Esophagitis always requires systemic therapy.
Classes of antifungal agents include polyenes (nystatin and amphotericin B), which bind to ergosterol in the fungal cell membrane and induce osmotic instability and loss of membrane integrity; azoles, including the imidazoles (clotrimazole) and triazoles (ketoconazole, itraconazole, fluconazole, voriconazole, ravuconazole, and posaconazole), which inhibit fungal cytochrome P450-dependent enzymes, resulting in the impairment of ergosterol biosynthesis and depletion of ergosterol from the fungal cell membrane; pyrimidine synthesis inhibitors, including 5-fluorocytosine (flucytosine), which inhibits DNA and RNA synthesis in fungal organisms; and the echinocandins (caspofungin, micafungin and anidulafungin), cyclic lipopeptides that inhibit beta-1:3 glucan synthase, an enzyme involved in fungal wall cell biosynthesis.
Nystatin is used in a topical preparation. The oral form is not absorbed and has minimal side effects other than dysgeusia. Flucytosine is available as a tablet and is associated with such side effects as nausea, vomiting, diarrhea, GI bleeding, renal insufficiency, hepatitis, thrombocytopenia, anemia, and leukopenia. Although the manufacturer recommends maintaining flucytosine levels between 25 and 100 µg/mL, most clinicians monitor laboratory parameters (complete blood count, liver function tests, and renal function tests) and the patient for adverse effects. Clotrimazole is available as a spray, solution, and troche for oral use. Clotrimazole has few side effects, and is absorbed from the GI tract poorly. Ketoconazole is available as a tablet or cream. Oral absorption is enhanced when the gastric pH is <4.0. Achlorhydria has been documented in HIV-infected patients and, when present, may interfere with ketoconazole absorption.(71) Itraconazole is available in a cyclodextrin oral solution, capsule, and parenteral form. The suspension and intravenous formulations have enhanced bioavailability compared with the capsule formulation. Absorption is improved when itraconazole is taken after a meal. Fluconazole, the first triazole compound released in the United States, is absorbed more completely than itraconazole or ketoconazole because absorption is not dependent on gastric acidity or food intake. Fluconazole is available in suspension, tablet, and parenteral form. In general, the side effects of ketoconazole, itraconazole, fluconazole, posaconazole, and voriconazole are similar, the more common being headache, dyspepsia, diarrhea, nausea, vomiting, hepatitis, and skin rash.(72) Voriconazole can cause reversible mild abnormal vision.(73) Prolonged administration of azoles may require surveillance of liver enzymes to monitor for hepatotoxicity. Significant drug interactions with each of these medications are provided in Table 3. The echinocandins are available only in parenteral forms. Caspofungin and micafungin are approved by the U.S. Food and Drug Administration (FDA) for the treatment of esophageal candidiasis. Adverse events including fever, nausea, infused-vein complications, and vomiting typically are mild.(74)
Most antifungal treatment studies for mucocutaneous candidiasis are difficult to interpret, given the small numbers of patients, heterogeneous populations, short follow-up, and nonblinded design. Specifically, no study has stratified patients by CD4+ lymphocyte count. This point is important because persons with low CD4+ lymphocyte counts may respond more slowly to treatment, have lower rates of fungal eradication, and have higher relapse rates than persons with less advanced disease. No treatment trials for vulvovaginal candidiasis in women with HIV infection have been published. Recommendations for the treatment of vulvovaginal disease are made based on data from the non-HIV-infected population.
Most of the published controlled trials for the treatment of oral and esophageal candidiasis are listed in Table 4. Response rates range from 34% to 100% in studies of treatment for oral and esophageal disease.(31,75-82) In clinical experience, the response rates to standard antifungal treatments are on the order of 75-95%. There are few significant differences in response rates between topical and systemic therapies or among the different systemic therapies for OPC. Thus, it is reasonable to conclude that clotrimazole, ketoconazole, fluconazole, and itraconazole probably are equivalent in the acute treatment of most cases of OPC.
The treatment of esophageal candidiasis has not been studied so well as the treatment of OPC. Most experts recommend systemic therapy because of the significant morbidity of esophageal candidiasis and the absence of evidence supporting the use of topical therapy. Response rates to systemic therapies generally are quite good. Fluconazole has proved to be more effective than ketoconazole in one trial.(80) There have been no comparative studies of itraconazole in tablet or solution vs fluconazole. Itraconazole solution probably is equivalent to fluconazole for treating esophageal candidiasis.
The cure rates for vulvovaginal candidiasis range from 72% to 98% in most trials of persons without HIV infection.(83-87) In the past, the standard treatment for vulvovaginal candidiasis typically consisted of topical clotrimazole or miconazole for 7 days. However, shorter courses have proved effective. Topical therapy for 3 days generally is equivalent to treatment with 7 days of topical medication. In a study comparing a 1-time dose of fluconazole (150 mg orally) with 7 days of topical clotrimazole therapy (100-mg vaginal suppositories), the clinical cure rate (75%) in the 2 groups was equivalent by day 35.(87) Mycologic eradication rates at day 35 were 63% for the fluconazole group and 57% for the clotrimazole group. Either topical or systemic therapy generally is effective in women with HIV infection, but relapse rates may be quite high.(38)
There are no prospective trials using real-time, in vitro susceptibility testing to guide the choice of antifungal therapy. A likely explanation is that most Candida infections respond to empiric therapy, and in vitro testing for antifungal resistance is not yet as reliable as antibiotic susceptibility testing of bacterial isolates. Some clinical fungal isolates found to be "resistant" by in vitro testing nevertheless respond to therapy. Less commonly, some patients fail to respond to therapy despite having a relatively "sensitive" organism isolated. Thus, despite the determination of standard definitions for what constitutes in vitro resistance, more work must be done in this area before susceptibility testing can be used as a guide to antifungal therapy.
There are a number of newer antifungals in varying phases of clinical development, including triazoles, echinocandins, sordarins, chitin synthase inhibitors, and topoisomerase inhibitors. Several new agents in the former 2 categories are now approved in the United States. In vitro activity of 3 new triazoles (posaconazole, ravuconazole, and voriconazole) appears to be quite good for Candida species, the latter agent having been licensed by the FDA in 2002.(88-94) A blinded, randomized study of voriconazole (200 mg twice daily) vs fluconazole (200 mg daily) for the treatment of esophageal candidiasis found no difference in the number of persons with endoscopically proven cure after 2-6 weeks of therapy: 94.8% of the voriconazole group (n = 115) vs 90.1% of the fluconazole group (n = 141).(73) Posaconazole (100 mg suspension/day) compared favorably to fluconazole (100 mg suspension/day) in the treatment of OPC in persons with HIV infection.(93) The clinical cure rate was 92% (n = 169) for posaconazole vs 93% (n = 160) for fluconazole. Similarly, posaconazole compared favorably to fluconazole in a dose-ranging study for the treatment of oral candidiasis in HIV infection.(94) Caspofungin, micafungin, and anidulafungin are members of the echinocandins, a novel class of antifungals. These agents also show promise in the treatment of Candida infections but are limited to parenteral administration at present.(95-97) In a study of 21 persons with esophageal or oral candidiasis, most of whom had HIV infection, who were treated with caspofungin, a favorable response was noted in 82% with esophageal disease and 100% with oral disease.(65) In a randomized trial comparing caspofungin with fluconazole for esophageal candidiasis, response rates and relapses at 4 weeks were similar. In the caspofungin group, 81% (n = 81) responded vs 85% (n = 94) in the fluconazole group.(98) Similarly, micafungin at a dose of 150 mg/day (n = 59) was similar in efficacy to fluconazole 200 mg/day (n = 60) for esophageal candidiasis (89.8% vs 86.7% response rate, respectively) in a randomized, double-blinded, multicenter study.(99)
Reports of refractory OPC and esophageal disease began emerging in 1990.(16,58,59,100-114) Refractory vaginal candidiasis has remained relatively uncommon.(115) Refractory disease is defined as the failure to respond to antifungal treatment with appropriate doses for a standard duration of time (eg, 14 days).(16,115) Fluconazole-refractory disease has received particular attention because of significant morbidity, treatment often requiring the use of parenteral agents, and the frequency with which fluconazole has been prescribed. The annual incidence of fluconazole-refractory OPC was reported to be 4-5% in advanced HIV infection in developed countries prior to the introduction of combination ART.(16,37) Like most other opportunistic infections, fluconazole-refractory OPC is less common with the widespread use of effective ART. Candidiasis refractory to amphotericin B is exceedingly uncommon.(116-120) Of note, clinical failures also may result from inadequate drug absorption or drug interactions that decrease the levels of some antifungal medications.(71,121,122)
Refractory candidiasis tends to occur in persons with advanced HIV disease (CD4 lymphocyte counts <50 cells/µL) who have been exposed to antifungal therapy on a continuous, chronic basis.(16) A longer median duration of exposure to antifungal therapy (419 vs 118 days, p < .001) and to systemic azole therapy (272 vs 14 days, p < .001) has been reported in comparing persons who had fluconazole-refractory OPC with matched controls.(104) In a randomized trial of episodic vs continuous use of fluconazole to determine which strategy for the prevention and treatment of mucocutaneous candidiasis would more likely lead to refractory candidiasis, no difference was observed in the rates of refractory disease (4.3% vs 4.1%, respectively).(123) Other factors that may predict the development of refractory candidiasis include use of prophylactic trimethoprim-sulfamethoxazole and a history of prior opportunistic illnesses such as Mycobacterium avium complex disease.(16) Similarly, chronic exposure to itraconazole results in higher rates of in vitro resistance, but these isolates typically remain susceptible to fluconazole.(124)
Refractory candidiasis often is difficult to treat and may become increasingly unresponsive to therapy over time. The most important step is to determine what medications and dosages have been tried and whether adherence to therapy has been adequate. Removing any interacting medications or increasing the dose of the antifungal agent may be curative in some persons. In general, persons with OPC that is unresponsive to clotrimazole, nystatin, ketoconazole, or itraconazole tablets will respond to fluconazole. Persons with OPC unresponsive to fluconazole 200 mg daily given for 2 weeks are less likely to respond to higher doses but sometimes do respond. Additionally, flucytosine may be added for synergy.
Options for managing fluconazole-refractory disease are listed in Table 5. There have been few controlled, comparative studies of these approaches. Parenteral amphotericin B (or liposomal preparations of amphotericin B) remains the drug of choice for persons with severe disease or esophageal involvement. For mild to moderate fluconazole-refractory OPC, amphotericin B oral suspension, itraconazole solution, or the addition of flucytosine are reasonable therapeutic strategies.(125-131) Overall, the response rate is 50-60% for itraconazole solution and slightly lower for oral amphotericin B solution. Other options for treating fluconazole-resistant isolates include voriconazole, caspofungin, micafungin, and anidulafungin.(88,89) ART should be optimized in persons with refractory candidiasis. Treatment with protease inhibitors has been noted to result in clinical improvement in difficult-to-treat cases.(132) Protease inhibitors have been shown to inhibit Candida secretory aspartic proteases, demonstrating direct antifungal activity against Candida.(133-135) The duration of antifungal treatment for refractory disease is based upon the response, but typically a course of 14 days for OPC or vaginal disease, and 21-28 days for esophageal disease, is necessary. Relapse rates are high in persons with refractory disease, and maintenance suppressive therapy is universally required.(131) In challenging cases, therapy 2-3 times weekly, or daily suppressive therapy on occasion, may be needed to avoid relapse.
Despite the frequency of mucosal candidiasis, invasive disease is uncommon in persons with HIV infection. There are few studies describing the incidence and prevalence of nonesophageal invasive candidiasis in HIV-infected persons. Most studies are restricted to case series or anecdotal reports. A retrospective review found the incidence of candidemia to be 0.09 episodes per 100 person-years from 1992 to 1996 compared with 1.1 episodes per 100 person-years from 1997 to 2001, suggesting a decrease following the advent of effective ART.(136) Overall, the incidence of candidemia in AIDS is probably <1%.(137,138) Individuals who develop candidemia usually have risk factors known to be associated with invasive candidiasis (eg, indwelling intravenous catheter, neutropenia, chemotherapy, parenteral alimentation).(136) The explanation for the lack of invasive disease in persons with HIV infection may be that the pathogenesis of invasive candidiasis usually involves disruption of a mucosal or skin surface barrier with subsequent hematogenous dissemination secondary to neutrophil and macrophage dysfunction--conditions that are not characteristic of HIV disease.(139,140)
In a review of 14 cases of Candida meningitis in HIV-infected patients, a history of at least 1 predisposing risk factor was noted in 10 of the patients, 9 of whom were intravenous drug users.(140) Of interest, 4 of the patients had no risk factors identified. Candida meningitis remains a rare entity, even among other immunosuppressed patients, and optimal therapy is unknown. It seems prudent for HIV-infected patients with Candida meningitis to continue chronic suppressive therapy with fluconazole on a regimen similar to that used in the management of cryptococcal meningitis.
The most important method of preventing mucocutaneous candidiasis is reversal of the immunodeficiency associated with HIV infection. Combination ART is the single best intervention to reduce the incidence of mucocutaneous candidiasis. Several studies demonstrate a decline in the rate of colonization and clinical disease with the use of potent ART.(42,43,47) This decline has been correlated with reduction in HIV-1 RNA levels in the plasma. Other possible interventions include smoking cessation, good oral hygiene, avoidance of unnecessary antibiotics and steroids, and specific antifungal medications.
Although recurrent mucocutaneous candidiasis is frequent in persons with untreated advanced HIV infection, the indications for prophylactic antifungal therapy remain uncertain. A randomized study comparing clotrimazole and fluconazole demonstrated that fluconazole can prevent invasive fungal infections such as cryptococcosis and esophageal candidiasis.(141) However, that study found no survival advantage. Weekly fluconazole prophylaxis also has been studied for the prevention of OPC and vulvovaginal disease.(38,142,143) Another study reported decreases in the incidence of both OPC (relative risk [RR] = 0.50; 95% confidence interval [CI], 0.33-0.71) and vulvovaginal disease (RR = 0.56; 95% CI, 0.41-0.77) in a group of 323 women with moderately advanced HIV infection who took weekly doses of fluconazole 200 mg (median follow-up, 29 months).(38) Thus, although prophylaxis can reduce the risk of mucocutaneous candidiasis, there is no associated survival advantage. Furthermore, several studies have demonstrated that continuous, long-term exposure to antifungals such as fluconazole can lead to the emergence of resistance and refractory infections.(16) Consequently, most experts do not recommend universal primary antifungal prophylaxis.
The use of secondary prophylaxis should be individualized. In a randomized study of continuous fluconazole prophylaxis vs episodic treatment with fluconazole in persons with prior mucosal candidiasis, continuous fluconazole resulted in fewer cases of OPC or esophageal disease (0.29 vs 1.08 per patient-year; p < .001) and invasive fungal infections (15 vs 28 episodes; p = .04) though there was no difference in survival.(123) There was no increase detected in resistance among the continuous usage arm. The proportion of patients in whom the final candidal isolate was resistant to fluconazole was 50 of 110 (45%) in the continuous fluconazole arm and 79 of 218 (36%) in the episodic fluconazole arm (p = .11). Some experts recommend prophylaxis in persons with a prior episode of esophageal candidiasis.(144,145) In general, persons with occasional disease or infrequent recurrences of OPC (fewer than 3 episodes per year) can be treated for each episode. An alternative approach is to provide the patient with a supply of antifungal medications that can be initiated at the earliest sign of recurrence. This alternative may be useful for adherent, well-educated patients with frequent or disabling episodes, particularly at low CD4+ lymphocyte counts. Some experts recommend prophylaxis in persons with advanced HIV disease when prescribing antibiotics or corticosteroids, such as in a patient with Pneumocystis jiroveci pneumonia. If the decision is made to use prophylaxis, daily, thrice-weekly, or weekly fluconazole are the options most frequently cited in published studies. Ketoconazole and itraconazole probably are useful as well but have not been evaluated in controlled trials. Topical therapy may be useful in some patients. In summary, reserving continuous use of antifungal agents to those persons with frequent or severe recurrences of mucosal candidiasis is recommended in order to avoid the emergence of drug resistance, avoid drug interactions, simplify already complex drug regimens, avoid drug toxicity, and lower the cost of treatment.