Despite these reports, a complete understanding of the epidemiology and management of this increasingly common complication of therapy has not yet been developed. There is little published data, for example, on the true incidence, prevalence, and risk factors for the development of fluconazole resistance. A recent case-control study from Johns Hopkins indicated that fluconazole resistance is most likely to develop in patients with advanced HIV disease (i.e. CD4 cell counts below 50 cells/mm3) who have had significant prior exposure to the drug and many prior episodes of candidiasis (1). (In this study the median duration of prior fluconazole therapy was > 200 days.) These findings are hardly a surprise, but they do give us a better sense of which patients are at greatest risk of developing resistance to azole therapy.
The current estimates are that 5% to 7% of patients with advanced HIV infection will develop candidiasis that is refractory to standard fluconazole therapy. Preliminary analysis of an ACTG study of this patient population suggests that clinicians should anticipate an annual incidence of azole-resistant candidiasis of about 4% in patients with CD4 counts below 50 cells/mm3. These data also indicate that prior episodes of candidiasis (especially esophageal disease) and other OIs (especially MAC) are risk factors for the development of resistance.
Most of the initial cases of candidiasis seen in HIV-positive patients are caused by Candida albicans, and in about 75% of the cases of azole-resistant candidiasis C. albicans can be cultured from the mouth. Patients who are taking fluconazole to suppress mucosal candidiasis are more likely to develop other species of candida such as C. (Torulopsis) glabrata, C. parapsilosis, and C. krusei. These organisms tend to be less susceptible to fluconazole or any of the other azoles.
Recently, a consensus methodology was developed that will henceforth be used to test in vitro susceptibility of various strains of candida to azole antifungals. This methodology has been successfully correlated with outcomes for patients with AIDS and oropharyngeal candidiasis. Isolates with a MIC < 8 µg/mL to fluconazole are regarded as susceptible; isolates whose MIC is > 64 µg/mL are regarded as resistant. Isolates with intermediate MICs are held to have reduced susceptibility, since infection caused by these isolates may respond to higher-than-standard doses of fluconazole or itraconazole. However, useful as these guidelines are in a research context, they are of limited use in the clinic, where treatment decisions are generally made on clinical grounds.
Although the clinical experience with patients who develop fluconazole-resistant candidiasis is largely anecdotal, it does tell us that these patients are difficult to manage and often suffer significant morbidity. In clinical terms the disease seen in these patients is identical to that seen in patients with sensitive infection, but it tends to be both progressive and more symptomatic, due to the lack of effective antifungal therapy. There is no information to suggest that fluconazole-resistant infection is more virulent, but esophagitis is much more common in patients with resistant disease.
Clinicians confronted with fluconazole-resistant thrush have a number of therapeutic options. Unfortunately, we do not yet have definitive data that identify the most appropriate and efficacious therapeutic choice for patients with resistant disease. Therefore, the clinician should begin by determining the obvious: Has the patient's candidiasis failed to respond to standard doses of fluconazole because the drug is not being taken as prescribed? Is the patient compliant? Are serum levels adequate? If not, the first step is to remind the patient of the paramount importance of compliance; the second step is to recheck the patient, after an interval of several weeks, to determine if enhanced compliance has led to enhanced efficacy.
If the patient's candidiasis has still not responded to therapy, higher doses of fluconazole -- up to 800 mg per day -- may be tried. Higher doses can be effective in patients whose infection is caused by organisms with intermediate susceptibility to azole antifungals. Elevated doses are not likely to be effective in patients with truly resistant strains of candida. The exceptions to this rule are the relatively rare cases of resistance that occur as the result of drug-drug interactions that lower serum levels of fluconazole. These interactions are most often seen in patients taking rifampin or rifabutin.
Other azoles may also be efficacious -- because some fluconazole-resistant isolates retain sensitivity to itraconazole and ketoconazole. In some cases fluconazole resistance may be mediated by an increase in active transport of fluconazole out of the cell wall of the fungus; in those cases either itraconazole or ketoconazole, which are more lipophilic than fluconazole, may still be able to remain within the cell (3).
Therapy with itraconazole or ketoconazole is usually ineffective in patients whose isolates are completely resistant in vitro. Unhappily, the opposite is not equally true, for in vitro susceptibility does not always imply therapeutic efficacy. In any case itraconazole capsules are rarely effective against resistant candidiasis. Recent data do suggest a role for the cyclodextrin oral-suspension formulation of itraconazole, however. In a retrospective evaluation of 19 patients with candidiasis that was resistant to azole therapy, 64% showed definite clinical improvement on 100 or 200 mg per day of this investigational agent (4). A prospective Canadian trial, which used a regimen of 100 mg twice daily for 14 days, reported a similar response rate (65% of 34 patients), and this study was able to show an association between response and microbiologic susceptibility (5).
Similar results have been seen in a larger study, reported last month at the 4th Conference on Retroviruses and Opportunistic Infections in Washington, D.C. (see "Oral suspension of itraconazole effective against refractory candidiasis," in this issue). Although still investigational, the oral suspension of itraconazole is available through a compassionate-release program. Interested clinicians should contact Janssen Phar-maceutica at 1-800-378-4779.
Improved antiretroviral therapy itself has been reported to lead to the clearance of refractory thrush (6), with the credit in this instance given to the addition of a protease inhibitor to the patient's prior regimen (see "Resolution of candidiasis in a patient receiving ddI plus saquinavir," Vol. 2, No. 5, page 112). This anecdotal report, and others like it, suggest that aggressive antiretroviral therapy which combines a protease inhibitor with nucleoside analogs may have a beneficial impact on resistant candidiasis.
Intravenous amphotericin B should be used initially in all patients with severe candidiasis, especially those with esophagitis, and it should be used in patients who fail azole therapy. A short course of therapy at low doses (0.3 mg/kg for 7 to 14 days) is usually effective, although relapses are common in the absence of some form of maintenance therapy. Some cases of amphotericin-B-resistant infection have been reported.
Lozenges containing amphotericin B, which are available in Europe but not in the United States, may be effective in controlling milder disease but are unlikely to have much impact on resistant infection. An oral suspension of amphotericin B is currently being evaluated in patients with fluconazole-resistant candidiasis. The preliminary data suggest that high doses (5 mL q.i.d.) will probably be needed to control infection.
Another alternative agent is flucytosine, which is active against Candida albicans but tends to be less effective against other species of candida. C. albicans rarely develops resistance to flucytosine, even with chronic therapy. One possible approach to the management of fluconazole-resistant C. albicans infection is treatment with amphotericin B followed by maintenance therapy with flucytosine (at a dose of 25 mg/kg q.i.d.). Clinicians who elect this course of therapy should be wary of the potential for considerable toxicity with flucytosine, which can produce myelotoxicity and gastrointestinal disturbances such as abdominal pain, nausea, diarrhea, and GI bleeding.
Finally, there have been anecdotal reports that GM-CSF is effective in cases of refractory candidiasis (7). Further trials of this agent are warranted, as are trials of investigational agents such as terbinafine and the echinocandidin derivatives.
William G. Powderly, M.D., is with the Division of Infectious Diseases, AIDS Clinical Trials Unit, Washington University School of Medicine, St. Louis, MO.