Flucytosine

Study objective: To assess the role and the therapeutic efficacy of fluconazole and itraconazole-flucytosine association compared with placebo, in the treatment of endoscopically diagnosed esophageal candidiasis in a selected population of AIDS patients. Design: Double-blind, placebo-controlled study. Setting: University Hospitals and AIDS Centers. Patients: Eighty-five HIV-positive patients (53 men and 32 women; mean age, 28 years) at first episode of esophageal candidiasis diagnosed by endoscopy (grades I to 11 of Kodsi's endoseopic classification and grades I to IIa of Barbaro's clinical classification). All the patients selected for the study provided informed consent. Interventions: The patients have been double blindly randomized in 3 groups of patients in relation to pharmacologic therapy: (1) the patients of the first group (n = 3 ) received fluconazole (3 mg/kg daily orally) and placebo (100 mg/kg/daily orally); (2) the patients of the second group (n = 30) received itraconazole (3 mg/kg daily orally) and flucytosine (100 mg/kg daily orally); and (3) the patients of the third group (n = 25) received placebo (3 mg/kg daily orally) and placebo (100 mg/kg daily orally). After 2 weeks of treatment, the patients previously randomized to receive placebo only were double blindly randoniized to receive fluconazole+placebo or itraconazole+flucytosine. To evaluate the efficacy of pharmacologic therapy, clinical and endoscopic examinations were performed at weeks 2 and 4 and at the end of forow-up (3 months). Results: At week 2, endoscopic cure (grade 0) was observed in 68.9% of the fluconazole+placebo group and in 72.4% of the itraconazole+flucytosine group (relative risk, 0.95; 95% confidence interval [CI], 0.68 to 1.33; p = 0.772); partial endoscopic response (grade 1) was observed in 22.7% of the placebo group. Clinical cure (grade 0) was observed in 75.8% of fluconazole+placebo group and in 72.4% of itraconazole+flucytosine group (relative irisk, 1.05; 95% CI, 0.77 to 1.42; p = 0.764), with a difference statistically significant for both treatments in comparison to placebo group (p < 0.001). Partial efiffical response (grade 1) was observed in 27.3% of the placebo group. At the end of follow-up, endoscopic cure was observed in 89.8% of the fluconazole+placebo group and in 94.8% of the itraconazole+flucytosine group (relative risk, 0.97; 95% CI, 0.83 to 1.08; p = 0.695). Clinical cure was observed in 94.8% of the fluconazole+placebo group and in 97.3% of the itraconazole+flucytosine group (relative irisk, 0.97; 95% CI, 0.89 to 1.07; p = 0.981). Conclusions: The results of this study have demonstrated that both fluconazole and itraconazole+ flucytosine association are efficacious in short-term treatment of esophageal candidiasis in AIDS patients with a statistically significant difference in comparison to placebo. Both therapeutic regimens demonstrated a good therapeutic efficacy, without statistically significant difference, between them, in the rate of endoscopic and clinical cure. Itraconazole+flucytosine association may represent an alternative therapeutic regimen for patients with fluconazole-resistant Candida esophagitis.

Key words: AIDS: candidiasis; fluconazole; flucytosine; itraconazole

Abbreviations: CEMIS = Candida Esophagitis Multicenter Italian Study; CI = confidedee interval; Hb = hemoglobin; NGC = neutrophil granulocyte count

Esophageal candidiasis is one of the most common mycotic opportunistic infections in HIV-positive patients, with a mean incidence of 15 to 20%. In 3 to 10% of cases, Candida esophagitis represents the first opportunistic infection and may precede the outbreak of more severe infections in more than 50% of cases.[1] The most clinically important species is Candida albicans (80% of cases); other species reported are as follows: Candida tropicalis, Candida pseudotropicalis, Candida krusei, Candida parapsilosis, and Candida glabrata (15 to 20% of cases).[1]

Contrasting opinions exist about the pharmacologic treatment of esophageal candidiasis in HIV-positive patients. A number of agents have been used to treat Candida esophagitis, including topical medications such as nystatin, clotrimazole, and miconazole, oral compounds such as ketoconazole and flucytosine, and IV agents such as amphotericin B.[2-6] Nevertheless, little information is actually available regarding the response of Candida esophagitis to antifungal therapy.

In previous trials, we have demonstrated that clinical cure can be observed in 77 to 78% of fluconazole-treated patients compared with 73% of itraconazole-treated patients and 63% of flucytosine-treated patients; endoscopic cure can be observed in 75% of fluconazole-treated patients compared with 38% of itraconazole-treated patients and 33% of flucytosine-treated patients.[6-8] The aim of this study has been to assess, by a double-blind multicenter randomization, the role, the therapeutic efficacy, and the safety of use of fluconazole and itraconazole-flucytosine association, compared with placebo, in the treatment of endoscopically diagnosed esophageal candidiasis in a selected population of AIDS patients, providing quantitative data on clinical and endoscopic responses to treatment.

Materials and Methods

Criteria of Selection of the Patients and Mycologic Assessment

HIV-positive patients without history of candida esophagitis and with mild-moderate grade symptoms of odynophagia, dysphagia, or retrosternal pain mere eligible for the study if upper GI endoseopy revealed white plaques or exudate consistent with diagnosis of Candida infection and if pseudohyphae were identified on brushing and/or biopsy specimens of esophageal mucosa, in the absence of other detectable opportunistic infections of the esophagus.

The study has not considered the following: patients younger than 18 years; those who had evidence of noncandidal systemic fungal infection; those who had received previous therapy with antifungal drugs; those with a history of allergy to imidazoles; those who had therapy with anticoagulants, rifampin, ansamycin, cyclosporine, medications that raise gastric pH ([H.sub.2]-receptor blocking agents, antacids, omeprazole), myelotoxic, cytotoxic, nephrotoxic, or experimental drugs; those who had used illicit drugs in the last 6 months; those in pregnancy or during breast feeding; those with total WBC count (less than 2,500/[mm.sup.3] and with neutrophil granulocyte count (NGC) less than 1,000/[mm.sup.3]; those with hemoglobin (Hb) value less than 10 g/dL; those with transaminase and bilirubin levels more than 3 times the normal values and/or with prothrombin time more than 5 s over control; those with levels of creatinine more than 2 times the normal values: those with serious acute infection; those with serious diseases not specifically related to HIV infection (cardiomyopathies, diabetes. hypertension, epilepsy, neoplastic diseases); those with a life expectancy of less than 3 months; and lastly those not able to provide the informed consent.

Mycologic assessment was made by culture of endoscopic brushing and/or bioptic samples of esophageal mucosa on Sabouraud 4% glucose agar. After 48 to 72 h at 37[degrees]C. yeast colonies were isolated and identified. For identification purposes of the various Candida species, germ tube formation in bovine serum, and the presence of chlamydospores and pseudomycelia on polysorbate (Tween)-rice agar were tested. Finally the yeasts were characterized by means of a system (API20C or ATB32C system: Merieux; Lyon, France).

Endoscopic examination was made using an instrument (Olympus GIF Q-IT20). Endoscopic lesions were graded on a scale of 0 to III by a modification of the endoscopic classification by Kodsi et al[9] of Candida esophagitis: grade 0, normal esophageal mucosa; grade I, individual raised plaques, each less than 2 mm in size; grade II, individual raised plaques more than 2 mm in size; grade III, confluent plaques combined or not with ulceration and/or increased friability of mucosa. Clinical symptoms were graded on a scale of 0 to III by a modification of a chimecal classification proposed by Barbaro et al[6-8] in prexious studies: grade 0, absence of symptoms; grade I, symptoms of mild entity not compromising the capability to swallows both solid and liquid substances; grade II, symptoms of moderate entity not compromising (grade IIa) or compromising (grade IIb) the capability to swallon, solid substances, but not compromising the capability to swallow liquid substances; grade III, symptoms of severe entity, compromising either the capability to swallow solid substances onily (grade IIIa) or compromising the capability to swallows both solid and liquid substances (grade IIIb).

HIV infection was confirmed by two subsequent enzyme-linked immunosorbent assay (ELISA) and Western blot analysis. The value of T cells subset CD4+ was cheeked by indirect immunofluorescence, using monoclonal antibodies (Ortho Diagnostic System; Raritan, NJ).

Randomization

The randomization scheme of the study is shown in Figure 1. The patients selected for the study were double blindly randomized and assigned to receive identical capsules of fluconazole (3 mg/kg daily orally) and placebo (100 mg/kg daily orally) or itraconazole (3 mg/kg daily orally) and flucytosine (100 mg/kg daily orally) or placebo (3 mg/kg daily orally) and placebo (100 mg/kg daily orally), according to a computer-generated randomization schedule, kept by an independent investigator, in a 1:1:1 ratio, for 2 weeks. After 2 weeks of treatment, the patients previously randomized to receive placebo were double blindly randomized to receive fluconazole+placebo or itraconazole+flucytosine, according to a computer-generated randomization schedule in a 1:1 ratio for 2 weeks. Patients randomized to receive placebo could receive active phannacologic treatment if no clinical improvement and/or worsening of clinical symptoms, as compared with baseline, were observed in the first week of treatment. The doses could be doubled if no improvement of clinical symptoms was observed after 2 weeks of treatment. Patients not presenting any clinical improvement could continue the pharmacologic treatment for a maximum of 5 weeks. The capsules given to the patients of two groups of randomization were provided by Department of Clinical Pharmacology of the University of Pavia, Italy.

Compliance

Compliance with prescribed therapy was verified during the clinical follow-up of the patients selected for the study by counting the remaining capsules at each visit. Patients of the different groups were considered compliant if they too at least 75% of the prescribed dosage of the drug during the entire study.

Clinical Evaluation and Follow-up

To evaluate the efficacy of pharmacologic therapy, clinical and endoscopic examinations were performed at baseline, at weeks 2 and 4, and at the end of follow-up (3 months). Endoscopic examination could be performed also if relapses of esophageal symptoms were observed during the follow-up period, in order to assess if Candida infection was or was not responsible for symptomatic relapses. Routine hematochemical tests (CBC count, BUN, creatinine, liver tests, electrolytes, glucose, and urinalysis) were performed at baseline, at weeks 2 and 4, and at the end of follow-up. The patients' charts were purged of information that might identify the study drugs by two independent investigators and then scored in a blinded fashion by two other investigators using a computerized database. Endoscopic cure was defined as the complete disappearance of visible candidiasis at endoscopy as compared with baseline. The endoscopic lesions observed in the different times of follow-up were graded according to the above-mentioned modified scale of Kodsi et al.[9] Clinical cure was defined as the complete resolution of symptoms as compared with baseline. Clinical symptoms reported by patients in the different times of follow-up were graded according to the above-mentioned modified clinical scale of Barbaro et al.[6-8] Relapse of Candida esophagitis was defined as the onset of esophageal symptoms with endoscopic findings consistent with diagnosis of Candida infection, in the absence of other detectable opportunistic infections of the esophagus, after a previous positive response to pharmacologic treatment.

Reduction of Dosage of Drugs and Interruption of

Pharmacologic Treatment

The dosage of the drugs in the groups of active treatment could be reduced in the presence of the following: anemia (Hb <8 g/dL); granulocytopenia (NGC <750/[mm.sup.3]), and leukopenia (WBC <1,500/[mm.sup.3]); serious alterations of liver and kidney functions; and manifestations of intolerance to pharmacologic therapy not requiring the interruption of the treatment. The pharmacologic treatment could be interrupted in the presence of progressive myelo-inhibition (Hb <7g/dL; WBC <1,000/[mm.sup.3], NGC <500/[mm.sup.3]); serious relapsing infections; serious toxic effects; needing to begin treatments interfering with antifungal therapy; manifestation of severe intolerance to pharmacologic treatment; and serious and progressive worsening of clinical conditions.

Statistical Analysis

Sample size has been estimated on the basis of a mean expected rate of positive response of 75 to 80% in the groups of active treatment[5-8] and of 50 to 60% in the placebo group;[7] with regard to a difference of 20 to 25% in expected response between the active treatment groups[5-8] and difference of 25 to 30% between the active treatment groups and the placebo group,[7] with a test power of 80% ([alpha]=0.05; [beta]=0.20, one-tailed test) and with a 90% compliance of selected patients,[6-8] a sample of 30 patients for each group was considered sufficient. Proportional data regarding endoscopic and clinical responses and the incidence of side effects in the patients of the study groups have been evaluated using the [X.sup.2] test for 2x2 tables with Yates' correction.[10] Nonproportional data, with 95% confidence interval (CI) for the differences, have been evaluated using the t test for independent samples.[10] The relative lisk for the rates of endoscopic and chrtical cure (grade 0 of Kodsi et al[9] and grade 0 of Barbaro et al[6-8]) and for the rates of relapses of esophageal symptoms observed in the patients of the active treatment groups (rate of cures and relapses observed in fluconazole+placebo group/rate of cures and relapses observed in itraconazole+flucytosine group), with 95% CI has been also calculated.[10] The probability of endoscopic and clinical cure over time was done using the Gray[11] test for comparing treatment groups for cumulative incidence of competing risks; 95% CIs for the probabilities of cure were computed based on normal approximation of the estimates of probabilities and their variances.[10,11]

Informed Consent

The research has been carried out in accordance with the Helsinki Declaration. The study protocol was explained to all the patients selected for it. All the patients selected for the study gave their informed consent.

RESULTS

Enrollment and Baseline Characteristics of the

Patients

From September 1994 to March 1995, a total of 353 HIV-positive outpatients with esophageal symptoms were observed at the Candida Esophagitis Multicenter Italian Study (CEMIS) Centers. Eighty-five patients (24%) fulfilled the selection criteria and entered the trial. In 75 patients (88.3%), C albicans was responsible for esophagitis; in the other 10 patients, C glabrata (6 patients [7%]) and C tropicalis (4 patients [4.7%]) eere responsible for esophageal infection. Thirty patients were assigned to receive fluconazole+placebo, 30 patients were assigned to receive itraconazole+flucytosine, and 25 patients were assigned to receive placebo only. Patient characteristics at baseline in the study groups are shown in Table 1.

[TABULAR DATA OMITTED]

Withdrawal From the Trial

One patient with C albicans esophagitis in the fluconazole+placebo group and 1 patient with C tropicalis esophagitis in the itraconazole+flucytosine group withdrew from the study after 1 week of treatment because of side effects and/or individual intolerance to continue pharmacologic treatment; 3 patients with C albicans esophagitis in the placebo group withdrew from the study because of individual intolerance to continue therapy after 1 week of treatment. One patient math C albicans esophagitis in the fluconazole+placebo group and 2 patients with C albicans esophagitis in the itraconazole+flucytosine group were unavailable for follow-up after clinical and endoscopic assessment at week 4. These patients were not considered for endoscopic and clinical evaluation of therapeutic efficacy of the assigned pharmacologic treatment. The dose of study medication was doubled in 2 patients (1 math C tropicalis and 1 mith C glabrata esophagitis) treated with fluconazole+placebo at week 2 and in 1 patient with C albicans esophagitis treated mith itraconazole+flucytosine at week 2 (p=0.553). No patient assigned to receive placebo only required active treatment during tbe first 2 weeks of treatment.

Endoscopic and Clinical Response at Week 2

At week 2, endoscopic cure (Kodsi et al[9] grade 0) was observed in 20 of 29 patients (68.9%) in the fluconazole+placebo group (18 mith C albicans, 1 with C tropicalis, and 1 with C glabrata esophagitis) and in 21 of 29 patients (72.4%) in the itraconazole+flucytosine group (19 with C albicans, 1 with C tropicalis, and 1 with C glabrata esophagitis) (relative risk, 0.95; 95% CI, 0.68 to 1.33; p=0.772); partial endoscopic response (grade 1) was observed in 5 patients with C albicans esophagitis of 22 of the lacebo group (22.7%). Clinical cure (Barbaro et al[6-8] grade 0) was observed in 22 of 29 patients (75.8%) of the fluconazole+placebo group (20 with C albicans, 1 with C tropicalis, and 1 with C glabrata esophagitis) and in 21 of 29 patients (72.4%) of the itraconazole+flucytosine group (19 with C albicans, 1 with C tropicalis, and 1 with C glabrata esophagitis) (relative risk, 1.05; 95% CI, 0.77 to 1.42; p=0.764), with a difference statistically significant for both treatments in comparison to placebo group (p<0.001). Partial clinical response (grade I) was observed in 6 of 22 patients (27.3%) of the placebo group (5 with C albicans and 1 with C tropicalis esophagitis).

Endoscopic and Clinical Response at Week 4

At week 2, the remaining 22 patients previously assigned to receive placebo were double blindly randomized to receive active treatment: 11 patients (10 with C albicans and 1 with C tropicalis esophagitis) received fluconazole+placebo, while tbe other 11 patients (9 witb C albicans, 1 with C tropicalis, and 1 with C glabrata esopbagitis) received itraconazole+flucytosine.

At week 4, endoscopic cure was observed in 37 of 40 patients (92.5%) of the fluconazole+placebo group (34 with C albicans, 2 with C tropicalis, and 1 with C glabrata esophagitis) and in 38 of 40 patients (95%) of the itraconazole+flucytosine group (35 with C albicans, 2 with C tropicalis, and 1 with C glabrata esophagitis) (relative risk, 0.97; 95% CI, 0.87 to 1.09; p=0.644). Clinical cure was observed in 38 of 40 patients (95%) of the fluconazole+placebo group (35 with C albicans, 2 with C tropicalis, and 1 with C glabrata esophagitis) and in 37 of 40 patients (92.5%) of the itraconazole+flucytosine group (35 with C albicans, 1 with C tropicalis, and 1 with C glabrata esophagitis) relative risk, 1.03; 95% CI, 0.92 to 1.15; p=0.644).

Endoscopic and Clinical Response at the End of

Follow-up

At the end of follow-up, endoscopic cure was observed in 35 of 39 patients (89.8%) of the fluconazole+placebo group (32 with C albicans, 2 with C tropicalis, and 1 with C glabrata esophagitis) and in 36 of 38 patients (94.8%) of the itraconazole+flucytosine group (33 with C albicans, 2 with C tropicalis, and 1 with C glabrata esophagitis) (relative risk, 0.97; 95% CI, 0.83 to 1.08; p=0.695). Clinical cure was observed in 37 of 39 patients (94.8%) of the fluconazole+placebo group (33 with C albicans, 3 with C tropicalis, and 1 with C glabrata esophagitis) and in 37 of 38 patients (97.3%) of the itraconazole+flucytosine group (33 with C albicans, 2 with C tropicalis, and 2 with C glabrata esophagitis (relafive risk, 0.97; 95% CI, 0.89 to 1.07; p=0.981).

Course of Therapy and Relapses

The mean duration of pharmacologic treatment was 3[+ or -]0.88 weeks in the fluconazole+placebo group and 3.5[+ or -]0.6 weeks in the itraconazole+flucytosine group (95% CI for the difference: -0.89 to -0.11; p=0.013). Relapses of esophageal symptoms were observed in 3 of 29 patients in the fluconazole+placebo group (10.3%) (1 at week 7 and 2 at week 8) and in 2 of 29 patients in the itraconazole+flucytosine group (6.8%) (1 at week 8 and 1 at week 10) (relative risk, 1.50; 95% CI, 0.27 to 8.32; p=0.641). In these patients, an endoscopic control with mycologic examination of bioptic specimens of esophageal mucosa demonstrated that Candida infection was responsible for symptomatic replapses in absence of other detectable opportunistic infections of the esophagus. C albicans was responsible for esophagitis in all the cases with symptomatic relapses. Concomitant treatment with zidovudine did not influence the course of therapy and the incidence of symptomatic relapses in both groups. On the contrary, the values of T cells CD4+ influenced significantly both the course of therapy and the incidence of symptomatic relapses; in fact, the mean value of T cells CD4+ was 62[+ or -]12/[mm.sup.3] in the patients of both groups of active treatment who presented endoscopic and clinical cure without symptomatic relapses compared with a mean value of 55[+ or -]7/[mm.sup.3] checked in nonresponder patients and in the patients who presented symptomatic relapses after a previous positive response (95% CI for the difference: 2.63 to 11.37; p=0.002). The patients math symptomatic relapses received again a pharmacologic treatment with fluconazole+placebo or with itraconazole+flucytosine, according to the number of randomization previously assigned in the stage of enrollment, with complete resolution of esophageal symptoms. The mean duration of pharmacologic treatment after relapses was 2.3[+ or -]0.5 weeks in the fluconazole+placebo group and 2[+ or -]0.3 weeks in the itraconazole+flucytosine group (95% CI for the difference: -0.79 to 1.79; p=0.305).

Probability of Cure

The probabihty of cure evaluated by the Gray test regarding both endoscopic and clinical response in the patients of the 2 groups of active treatment is shown in Figure 2.

At week 2, the probability of endoscopic cure was 66.6% in the fluconazole+placebo group and 70% in the itraconazole+flucytosine group ([delta]=3.4%; 95% CI, -0.269 to 0.201; p=0.996); the probability of clinical cure was 73.5% in the fluconazole+placebo group and 70% in the itraconazole+flucytosine group ([delta]=3.5%; 95% CI, -0.193 to 0.263; p=0.989). At the end of follow-up, the probabibty of endoscopic cure was 84% in the fluconazole+placebo group and 80% in the itraconazole+flucytosine group ([delta]=4%; 95% CI, -0.132 to 0.212; p=0.873); the probability of clinical cure was 90.5% in the fluconazole+placebo group and 90% in the itraconazole+flucytosine group ([delta]=0.5%; 95% CI, -0.128 to 0.138; p=0.756).

Side Effects and Compliance

Side effects were mirtimal and comparable in the 2 groups receiving active treatment, without statistically significant difference in comparison to the placebo group (Table 2).

In previous trials, we demonstrated that 29.6% of flucytosine-treated patients and 35% of itraconazole-treated patients had resolution of symptoms without an endoscopic cure, whereas this occurred, respectively, in 3.3 to 7.4% of fluconazole-treated patients (95% CI for the difference vs flucytosine: 0.015 to 0.429; p=0.080; 95% CI for the difference vs itraconazole: 0.176 to 0.458; p<0.001).[6-8] Thus, the association therapy increased the therapeutic action of flucytosine and itraconazole by 30% (27% for flucytosine and 32% for itraconazole) compared with the rates of clinical and endoscopic response that we observed using the same drugs alone.[6-8] Nevertheless, no therapeutic response was observed in 5.1% of the fluconazole+placebo group and in 5.3% of the itraconazole+flucytosine group (95% CI for the difference: -0.101 to 0.097; p=0.636).

Concomitant treatment vath zidovudine did not influence the course of therapy and the incidence of symptomatic relapses compared with the state of immunodeficiency of the patients; in fact, in both groups of active treatment, the mean value of T cells CD4+ in nonresponder patients and in the patients who presented symptomatic relapses after a previous positive response was 12% less than that checked in the patients who presented endoscopic and clinical cure without symptomatic relapses. Therefore, the state of immunodeficiency of the patients represented the principal factor responsible for treatment failure.

It is to be underlined that symptoms of Candida esophagitis may resolve despite continued endoscopically demonstrable gross involvement of the esophageal mucosa.[3,5-8] According to our experience, a spontaneous remission of clinical symptoms can be observed up to 30% of patients without resolution of endoscopic lesions.[6-8] However, some investigators have noted that approximately 10 to 20% of HIV-positive patients who have esophageal symptoms have no evidence of esophageal abnormalities;[3,5,14] according to our experience, no evidence of endoscopic lesions has been observed in 15% of HIV-positive patients with esophageal symptoms.[6-8]

As regards the rates of cure related to Candida species isolated by culture, in our study, endoscopic and clinical cures were observed in 96.2% of the patients with C albicans esophagitis (95.5% in fluconazole+placebo group and in 97% in itraconazole+flucytosine group; 95% CI for the difference: -0.100 to 0.070; p=0.800), in 91.6% of the patients with C tropicalis esophagitis (83.3% in fluconazole+placebo group and 100% in itraconazole+flucytosine group; 95% CI for the difference: -0.291 to -0.043; p=0.026), and in 50% of the patients with C glabrata esophagitis (50% in fluconazole+placebo group and 50% in itraconazole+flucytosine group; 95% CI for the difference: -0.223 to 0.223; p=0.820).

For the chnical purposes of our study, in vitro sensitivity of Candida species isolated by culture to fluconazole and itraconazole was not tested because the discrepancies existing between in vitro and in vivo responses of different Candida species to antifungal drugs are well known. In fact, in vitro susceptibihty testing of antifungal drugs against fungal organisms is of questionable value because of the limited correlation of the results with clinical response.[15] Moreover, comparison of minimal inhibitory concentrations from laboratory to laboratory and from study to study should be interpreted with caution because the results are influenced by a variety of factors, including the growth medium used, pH, temperature, and size of the inoculum.[15] Furthermore, the discrepancies between in vitro and in vivo responses are greater in HIV-positive patients; in fact, a different in vivo pharmacologic spectrum of action of antifungal drugs (including azole derivatives) has been described in HIV-positive patients compared with healthy subjects, in relation both to a reduced oral absorption and to the state of immunodeficiency of the patients.[15-19] These aspects may justify a different therapeutic action of antifungal drugs in HIV-positive patients that cannot be foreseen by in vitro tests.

The potential explanation for the difference in therapeutic response to fluconazole compared with flucytosine may be related to a better oral absorption of fluconazole together with a greater half-life (22 h compared with 3 to 5 h of flucytosine).[12,13] In addition, fluconazole has a greater in vivo and in vitro activity against Candida species than does flucytosine.[12] The difference in therapeutic response to fluconazole compared with itraconazole may be related to a greater bioavailability of fluconazole for a significantly reduced protein binding, to a higher peak plasma concentration (10.2 [mu]g/ml of fluconazole vs 0.2 to 0.4 [mu]/ml of itraconazole) and to a more consistent oral absorption of fluconazole (gastric pH-independent).[5-8,12,15-19] The effect of gastric acid secretion on therapeutic activity of itraconazole, in fact, must be considered, since most patients with AIDS have gastric acid hyposecretion possibly as a residt of a decrease in parietal cell function in the presence of antiparietal cell antibodies.[5,20] The absorption of itraconazole, like that of ketoconazole, may be impaired in AIDS patients as a result of reduced gastric acid secretion;[12,15,18-21] pharmacokinetic studies have demonstrated that the absorption of oral itraconazole is reduced by approximately 50% in AIDS patients compared with healthy volunteers.[22] While it has been demonstrated that fluconazole has greater activitv in vivo against Canclida species than does ketoconazole,[5,18,19] a possible greater activity of fluconazole in vivo compared with itraconazole can be assumed, but it is still to be defined.[6,8,12,18,19]

The resistance phenomena to flucytosine by Candida species are more frequent than those observed to fluconazole and itraconazole and are to be considered on a clinical point of view.[7,23] The mechanism of drug resistance may include loss of deaminase and decreased permeability to the drug.[7,12] According to our experience, the resistance phenomena to flucytosine are more frequent in long-term treatment and in patients who received a previous antifungal therapy.[7] Nevertheless, resistance phenomena to fluconazole and itraconazole are increasing among HIV-positive patients with Candida infection as confirmed by laboratory and epidemiologic data.[19,24] Furthermore, the state of immunodeficiency of the patients plays a significant role in the onset of resistance phenomena.[25]

The resistance phenomena to flucytosine and to azole drugs may be resolved with the association of antifungal drugs provided with a different mechanism of action. A synergistic interaction between flucytosine and fluconazole against C albicans has been described[26] and in our study, a synergistic action between itraconazole and flucytosine association has been observed, the therapeutic action of both drugs being increased by 30%. Under this point of view, itraconazole+flucytosine association may represent an alternative therapeutic regimen for patients with fluconazole-resistant Candida esophagitis.

In summary, both fluconazole and itraconazole-flucytosine association are efficacious and well tolerated, with a good therapeutic action and with a significant compliance in short-term treatment of Candida esophagitis in AIDS patients. The association of itraconazole with flucytosine association is provided with a synergistic therapeutic action and may be suggested in the long-term treatment of Candida esophagitis, reducing the incidence of resistance phenomena to pharmacologic treatment. Further controlled clinical trials are needed to confirm this hypothesis.

References

[1] Rondanelli EG. AIDS: clinical and laboratory atlas. Pavia: Edizioni Medico-Scientifiche, 1989; 86-91 [2] Deschamps MM, Pape JW, Verdier RI, et al. Treatment of Candida esophagitis in AIDS patients. Am J Gastroenterol 1988; 83:20-1 [3] Connolly GM, Hawkins D, Harcourt-Webster JN, et al. Oesophageal symptoms, their causes, treatment and prognosis in patients with the acquired immunodeficiency, syndrome. Gut 1989; 30:1033-39 [4] Smith DE, Midgley J, Allan M, et al. Itraconazole vs ketoconazole in the treatment of oral and oesophageal candidosis in patients infected with HIV. AIDS 1991; 5:1367-71 [5] Laine L, Dretler RH, Conteas CN, et al. Fluconazole compared with ketoconazole for the treatment of Candida esophagitis in AIDS: a randomized trial. Ann Intern Med 1992; 117:655-60 [6] Barbaro G, Di Lorenzo G. Comparison of therapeutic activity of fluconazole and itraconazole in the treatment of esophageal candidiasis in AIDS patients: a double-blind, randomized, controlled clinical study. Ital J Gastroenterol 1995; 27:175-80 [7] Barbaro G, Barbarini G, Di Lorenzo G. Fluconazole vs flucytosine in the treatment of esophageal candidiasis in AIDS patients: a double-blind, placebo-controlled study. Endoscopy 1995; 27:377-83 [8] Barbaro G, Barbarini G, Di Lorenzo G. Fluconazole compared with itraconazole in the treatment of esophageal candidiasis in AIDS patients: a double-blind, randomized, controlled clinical study. Gut 1995; 37(S2):A6-7 [9] Kodsi BE, Wicremesinghe PC, Kozinn PJ, et al. Candida esophagitis: a prospective study of 27 cases. Gastroenterology 1976; 71:715-19 [10] Armitage P, Berry G. Statistical methods in medical research. 3rd ed. Oxford, England: Blackwell Scientific, 1994 [11] Gray RJ. A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat 1988; 16:1141-54 [12] Bennett JE. Antifungal agents. In: Mandell GL, Douglas RG Jr, Bennett JE, eds. Principles and practice of infectious diseases. 3rd ed. New York: Churchill Livingstone, 1990; 361-70 [13] Diasio RB, Bennett JE, Myers CE. Mode of action of 5-fluorocytosine. Biochem Pharmacol 1978; 27:703-07 [14] Bonacini M, Young T, Laine L. The causes of esophageal symptoms in human immunodeficiency virus infection: a prospective study of 110 patients. Arch Intern Med 1991; 151:1567-72 [15] Grant SM, Clissold SP. Itraconazole: a review of its pharmacodynamic and pharmacokinetics properties, and therapeutic potential in superficial and systemic myocoses. Drugs 1990; 39:877-916 [16] Phillips P, Fetchick R, Weisman I, et al. Tolerance and efficacy of itraconazole in treatment of systemic myocoses: preliminary results. Rev Infect Dis 1987; 9(suppl 1):S87-93 [17] Just-Nubling G, Gentschew G, Dohle M, et al. Fluconazole in the treatment of oropharyngeal candidosis in HIV-positive patients. Mycoses 1990; 33:435-40 [18] Saag MS, Dismukes WE. Azole antifungal agents: emphasis on new triazoles. Antimicrob Agents Chemother 1988; 32:1-8 [19] Como JA, Dismukes WE. Oral azole drugs Systemic antifungal therapy. N Engl J Med 1994; 330:263-72 [20] Lake-Bakaar G, Quadros E, Beidas S, et al. Gastric secretory failure in patients with the acquired immunodeficiency syndrome (AIDS). Ann Intern Med 1988; 109:502-04 [21] Lake-Bakaar G, Tom W', Lake Bakaar D, et al. Gastropathy and ketoconazole malabsorption in the acquired immunodeficiency syndrome (AIDS). Ann Intern Med 1988; 109:471-73 [22] Smith D, Woestenborghs R, Heykants J, et al. The pharmacokinetics of oral itraconazole in AIDS patients. Proceedings of the Seventh International Conference on AIDS, Florence, June 16-21, 1991, 225 [abstract] (WB 2174) [23] Francis P, Walsh TJ. Evolving role of flucytosine in immuno-compromised patients: new insight into safety, pharmacokinetics, and antifungal therapy. Clin Infect Dis 1992; 15:1003-18 [24] Cameron ML, Schell WA, Bruch S, et al. Correlation of in vitro fluconazole resistance of Candida isolates in relation to therapy, and symptoms of individuals seropositixe for human immunodeficiency virus type I. Antimicrob Agents Chemother 1993; 37:2449-53 [25] Ruhnke M, Eigler A, Engelmann E, et al. Correlation between antifungal susceptibility testing of Candida isolates from patients with HIV infection and cliniical results after treatment with fluconazole. Infection 1994; 22:132-36 [26] Scalarone GM, Mikami Y, Kurita N, et al. In vitro comparative evaluation of the postantifungal effect: synergistic interaction between flucytosine and fluconazole against Candida albicans. Mycoses 1991; 34:405-10

COPYRIGHT 1996 American College of Chest Physicians
COPYRIGHT 2004 Gale Group