Current issues of ACP Journal Club are published in Annals of Internal Medicine


Fluconazole prevented fungal infections but not mortality in advanced HIV infection

ACP J Club. 1995 Sept-Oct;123:43. doi:10.7326/ACPJC-1995-123-2-043

Source Citation

Powderly WG, Finkelstein DM, Feinberg J, et al. A randomized trial comparing fluconazole with clotrimazole troches for the prevention of fungal infections in patients with advanced human immunodeficiency virus infection. N Engl J Med. 1995 Mar 16;332:700-5.



To evaluate the effectiveness of fluconazole as primary prophylaxis against fungal infections in patients with advanced human immunodeficiency virus (HIV) infection.


Randomized controlled trial with median 35-month follow-up.


29 centers in the United States.


428 patients (median age, 36 y; 95% men) with HIV infection and a CD4+cell count < 200 × 106/L who had tolerated ≥ 1 month of zidovudine at a dose of ≥ 500 mg/d. Exclusion criteria were history of systemic fungal infection; allergy or intolerance to imidazoles or azoles; high serum aminotransferase levels; positive serum cryptococcal antigen during screening; maintenance therapy with systemic antifungal agents; therapy for mycobacterial infection; or active mucosal fungal infection. Follow-up was 82%.


217 patients were assigned to receive oral fluconazole, 200 mg/d, and 211 patients were assigned to receive clotrimazole troches, 10 mg 5 times daily.

Main Outcome Measures

Fungal infections (invasive, superficial, and esophageal candidiasis) and survival.

Main Results

Invasive fungal infections developed in 9 patients (4.1%) assigned to fluconazole compared with 23 patients (10.9%) assigned to clotrimazole {95% CI for the 6.8% absolute risk reduction, 1.9% to 12.1%; P = 0.008; relative risk reduction, 61.9%; number needed to treat, 15, CI, 8 to 54} (numbers calculated from data in article). Of the 32 invasive fungal infections, 17 were cryptococcosis (2 in the fluconazole group and 15 in the clotrimazole group). The benefit of fluconazole was greater for patients with a CD4+ cell count ≤ 50 × 106/L. Esophageal candidiasis developed in 3 patients (1.4%) receiving fluconazole compared with 17 (8.1%) receiving clotrimazole (P < 0.001). 33 patients (15.2%) in the fluconazole group developed superficial fungal infections compared with 100 (47.4%) in the clotrimazole group (P < 0.001). 98 patients (45%) in the fluconazole group died compared with 89 (42%) in the clotrimazole group (P = 0.26). More toxicities were reported in the fluconazole group than in the clotrimazole group (need for transfusion, nausea, and abdominal pain).


Fluconazole taken as primary prophylaxis reduced the frequency of fungal infections in patients with advanced HIV infection. The benefit of fluconazole was greater among patients with a CD4+ cell count ≤ 50 × 106/L. Survival was similar in the 2 groups.

Sources of funding: National Institute of Allergy and Infectious Disease; the National Center for Research Resources; Pfizer Central Research.

For article reprint: Dr. S.A. Bozzette, San Diego Veterans Affairs Medical Center, Mail Code 111N-1, 3350 La Jolla Village Drive, San Diego, CA 92161. FAX 610-552-4321.


This well-designed study by Powderly and colleagues has clearly defined end points and good long-term follow-up and addresses an important controversial subject: efficacy and indications for primary prophylaxis of cryptococcal meningitis in patients with HIV infection. This study confirms previous reports that fluconazole is an effective primary prophylactic agent and helps to resolve the remaining issue—for which patients and at what cost?

Esophageal candidiasis and selected invasive fungal infections associated with HIV infection, such as cryptococcal meningitis, are relatively easy to diagnose and treat. Esophageal candidiasis is rarely life-threatening, and cryptococcal meningitis has a relatively low attributable mortality. Thus, no study (including this one) has shown improved survival with primary prophylaxis. Cryptococcal meningitis rarely occurs with CD4+ cell counts > 100 × 106/L (78% of patients in this study had CD4+ cell counts < 50 × 106/L). The high financial costs, potential toxicities (increased transfusion requirement and abdominal pain), a 14% incidence of at least 1 episode of severe nausea that did not require discontinuation of drug therapy, and the all-important resistance concerns (1) (not addressed by this study) strongly argue against routine use of fluconazole solely for prevention of invasive fungal disease in patients with HIV infection and CD4+ cell counts > 100 × 106/L.

For patients with CD4+ cell counts < 100 × 106/L, the waters are much murkier. Patients with recurrent superficial fungal infection certainly should be considered for long-term fluconazole therapy because it is clearly effective and provides prophylaxis for both superficial and invasive disease, and for esophageal candidiasis. Many patients with advanced HIV infection fall into this group because of the ubiquity of oropharyngeal candidiasis. The study by Powderly and colleagues supports primary prophylaxis with fluconazole, even without recurrent candidiasis, for patients with CD4+ cell counts < 50 × 106/L. For CD4+ cell counts > 50 × 106/L, insufficient data are available to draw firm conclusions.

Aaron E. Glatt, MD
Catholic Medical Center of Brooklyn and Queens, Inc. Jamaica, New York