Antifungal azoles [for systemic use]

Mechanisms

Drug interactions with the antifungal azoles are common for several reasons:

• they are substrates of CYP3A4, but also interact with the heme moiety of CYP3A, resulting in non-competitive inhibition of oxidative metabolism of many CYP3A substrates; to a lesser extent they also inhibit other CYP450 isoforms;

• although fluconazole undergoes minimal CYP-mediated metabolism, it nevertheless inhibits CYP3A4 in vitro, albeit much more weakly than other azoles [ , ]; however, fluconazole also inhibits several other CYP isoforms in vitro and interacts with enzymes involved in glucuronidation [ ];

• interaction of antifungal azoles and other CYP3A substrates can also result from inhibition of P-glycoprotein-mediated efflux; P-glycoprotein is extensively co-localized and exhibits overlapping substrate specificity with CYP3A [ ]; in a cell line in which human P-glycoprotein was overexpressed, itraconazole and ketoconazole inhibited P-glycoprotein function, with 50% inhibitory concentrations of about 2 and 6 μmol/l respectively; however, fluconazole had no effect [ ].

• the systemic availability of the antifungal azoles depends in part on an acidic gastric environment and the activity of intestinal CYP3A4 and P-glycoprotein.

For details of interactions with individual antifungal azoles, see individual monographs (fluconazole, itraconazole, ketoconazole, miconazole, and voriconazole).

Inhibition of metabolism by CYP3A4, and inhibition of transport by multidrug transporters. Both were important in a boy with toxicity from a chemotherapeutic regimen containing drugs that are handled by these systems [ ].

• A 14-year-old boy with Hodgkin’s lymphoma was given vinblastine, doxorubicin, methotrexate, and prednisone chemotherapy and low-dose radiotherapy. When he was given itraconazole for a presumed fungal infection during an episode of neutropenia, unexpectedly severe bone marrow toxicity and neuropathy suggested toxicity from the chemotherapy due to enhancement by itraconazole. The itraconazole was withdrawn and the neutropenia and neuropathic pain improved.

The authors suggested that itraconazole had interfered with the metabolism of vinblastine, resulting in neurotoxicity, and with the metabolism of doxorubicin and methotrexate and the transport of doxorubicin, resulting in bone marrow suppression.

Posaconazole is an exception, since it is eliminated unchanged in the feces [ ].

A novel mechanism whereby azoles may take part in drug interactions has been described [ ]. Drug metabolism is controlled by a class of orphan nuclear receptors that regulate the expression of genes such as CYP3A4 and MDR-1 (multi-drug resistance-1). Xenobiotic-mediated induction of CYP3A4 and MDR-1 gene transcription was inhibited by ketoconazole, which acted by inhibiting the activation of human pregnenolone X receptor and constitutive androstene receptor, which are involved in the regulation of CYP3A4 and MDR-1. The effect was specific to this group of nuclear receptors.

Frequency

To assess the frequency of potential drug interactions with azole derivatives and the consequences of interactions between fluconazole and other drugs in routine in-patient care, a retrospective cohort study of patients with systemic fungal infections treated with an oral or intravenous azole derivative was conducted in a tertiary-care hospital [ ]. Of the 4185 admissions in which azoles (fluconazole, itraconazole, or ketoconazole) were given, 2941 (70%) admissions involved potential drug interactions, and in 2716 (92%) there were potential interactions with fluconazole. The most frequent interactions that were potentially moderate or severe were co-administration of fluconazole with prednisone (25%), midazolam (18%), warfarin (15%), methylprednisolone (14%), ciclosporin (11%), and nifedipine (10%). Charts were reviewed for 199 admissions in which patients were exposed to potential fluconazole drug interactions. While four adverse events were attributed to fluconazole, none was thought to have been due to a drug–drug interaction, although in one instance fluconazole may have contributed. The authors concluded that although fluconazole drug interactions were very frequent they had few apparent clinical consequences.

Alfentanil

In a randomized crossover study in 12 healthy volunteers, oral voriconazole (400 mg twice on the first day and 200 mg twice on the second day) increased the AUC of intravenous alfentanil 20 micrograms/kg six-fold, reduced its mean plasma clearance by 85%, from 4.4 to 0.67 ml/minute/kg, and prolonged its half-life from 1.5 to 6.6 hours [ ]. Alfentanil caused nausea in five subjects and vomiting in two, all when they were taking voriconazole. The authors attributed this interaction to inhibition of CYP3A by voriconazole.

All-trans-retinoic acid

See Tretinoin.

Amphotericin

In evaluating possible antagonism between amphotericin and antifungal azoles, details of the experimental set-up are crucial. When filamentous fungi were exposed to subfungicidal concentrations of azoles, before exposure to an amphotericin + azole combination, antagonism could always be shown both in vitro and in vivo [ ].

In vitro studies and experiments in animals have given conflicting results relating to potential antagonism between the effects of fluconazole and amphotericin on Candida species [ ]. However, large, randomized, double-blind comparisons of fluconazole with and without amphotericin for 5 days in non-neutropenic patients with candidemia showed no evidence of antagonism, but faster clearance of the organism from the blood and a trend toward an improved outcome in those who received the combination [ ].

The combination of amphotericin with ketoconazole appears to lead to antagonism [ ]. A study of the effects of combinations of amphotericin with fluconazole, itraconazole, or ketoconazole against strains of Aspergillus fumigatus in vitro showed antagonistic effects in some strains, but different effects in other strains [ ]. In one group of mice infected with Candida , combinations of amphotericin with fluconazole were more effective than fluconazole alone; in another group the combination showed no interaction, but was not better than either drug given alone [ ].

Although there are no clinical data, it can be expected that similar antagonism occurs between amphotericin and squalene oxidase inhibitors, which also eliminate the primary target ergosterol from the fungal cell membrane.

Anidulafungin

In a placebo-controlled study in 17 subjects anidulafungin (200 mg on day 1 then 100 mg/day on days 2–4) had no effect on the pharmacokinetics of voriconazole (400 mg every 12 hours on day 1 then 200 mg every 12 hours on days 2–4) [ ]. There were no dose-limiting or serious adverse events, and all adverse events were mild and consistent with the known safety profiles of the two drugs.

Antacids

The potential for a pH-dependent pharmacokinetic interaction between posaconazole 200 mg and the antacid Mylanta (co-magaldrox) 20 ml has been investigated under fasting and non-fasting conditions [ ]. In a randomized, four-period, crossover, single-dose study in 12 healthy men completed this. Food increased the relative systemic availability of posaconazole by 400%, but antacid co-administration had no statistically significant effect.

The effects of an antacid suspension (aluminium hydroxide 220 mg + magnesium hydroxide 120 mg in 240 ml) on the oral absorption of itraconazole 200 mg from capsules has been investigated in a randomized, open, two-period, crossover study in 12 healthy Thai men [ ]. The t _max of itraconazole was prolonged and its C _max and AUC were markedly reduced by the antacid, implying that the antacid markedly reduced the speed and extent of itraconazole absorption.

Antihistamines

The effects of co-administration of ketoconazole 400–450 mg/day on the pharmacokinetics of ebastine 20 mg/day and loratadine 10 mg/day and on the QT _c interval have been evaluated in two placebo-controlled studies in healthy men (n = 55 and 62) [ ]. Neither ebastine nor loratadine alone altered the QT _c interval. Ketoconazole and placebo increased the mean QT _c by 6.96 ms in the ebastine study and by 7.52 ms in the loratadine study. Mean QT _c was statistically significantly increased during administration of both ebastine + ketoconazole administration (12.21 ms) and loratadine + ketoconazole (10.68 ms) but these changes were not statistically significantly different from the increases seen with placebo + ketoconazole (6.96 ms). Ketoconazole increased the mean AUC for ebastine 43-fold, and that of its metabolite carebastine 1.4-fold. It increased the mean AUC of loratadine 4.5-fold and that of its metabolite desloratadine 1.9-fold. No subjects withdrew because of electrocardiographic changes or drug-related adverse events. Thus, the larger effect of ketoconazole on the pharmacokinetics of ebastine was not accompanied by a correspondingly larger pharmacodynamic effect on cardiac repolarization.

Antiretroviral drugs

Indinavir is metabolized mainly by CYP3A4. There have been two randomized placebo-controlled studies in healthy men of the pharmacokinetic interactions, safety, and tolerance of voriconazole and indinavir [ ]. The first was an open parallel-group study of the effect of indinavir on the steady-state pharmacokinetics of voriconazole in 18 volunteers. The subjects took voriconazole 200 mg bd (days 1–7), then voriconazole 200 mg bd plus either indinavir 800 mg or placebo tds (days 8–17). The second was a double-blind, randomized, crossover study of the effect of voriconazole on the steady-state pharmacokinetics of indinavir in 14 volunteers, who took indinavir 800 mg tds + voriconazole 200 mg or placebo bd for two 7-day treatment periods separated by a washout period of at least 7 days. There were no important changes in the pharmacokinetics of either compound. Voriconazole co-administered with indinavir was well tolerated without serious adverse events.

However, voriconazole has reportedly interacted with other antiretroviral drugs.

• A 10-year-old girl (weight 21 kg: height 130 cm) with vertically acquired AIDS received antiretroviral combination therapy and died of liver failure after starting to take voriconazole [ ]. While taking amprenavir (22.5 mg/kg bd), didanosine (120 mg/m ² bd), nevirapine (4 mg/kg bd), lopinavir (10 mg/kg bd), and ritonavir (2.5 mg/kg bd), she was given voriconazole 200 mg bd for refractory esophageal candidiasis. The next day her liver function tests rose slightly and rapidly deteriorated within 7 days, when voriconazole was withdrawn. Infectious causes were excluded. After 2 days the plasma concentrations of the antiretroviral drugs were increased (lopinavir, 10 μg/ml; nevirapine, 7.7 μg/ml; amprenavir, 10.9 μg/ml) compared with concentrations during the 6 months before admission (lopinavir, 3.9–6.0 μg/ml; nevirapine, 3.5–8.4 μg/ml; amprenavir, 3.5–7.7 μg/ml). There was no fever. She was alert and afebrile and neither had any neurological symptoms nor complained of pain. In the presence of progressive liver dysfunction, voriconazole and HAART were withdrawn. However, irreversible liver failure ensued, followed by hepatic coma. She dies 28 days after the start of voriconazole therapy. A postmortem was not performed.

The authors concluded that an interaction with HAART was the most likely explanation for the ultimately fatal liver failure.

Aripiprazole

Aripiprazole is mainly metabolized in vitro by CYP3A4 and CYP2D6. The effect of itraconazole 100 mg/day for 7 days on the pharmacokinetics of a single oral dose of aripiprazole 3 mg has been studied in 24 healthy adult men [ ]. Itraconazole increased the C _max , AUC, and terminal half-life of aripiprazole by 19%, 48%, and 19% respectively and of its main metabolite OPC-14857 by 19%, 39%, and 53%. Itraconazole reduced the oral clearance of aripiprazole in extensive metabolizers by 27%, with an even greater reduction (47%) in intermediate metabolizers. For C _max , there was no significant difference between extensive metabolizers and intermediate metabolizers, and the percent change by co-administration of itraconazole was less than 20% in both groups. For OPC-14857, the t _max in intermediate metabolizers was longer than that in extensive metabolizers, and the difference was amplified by itraconazole. The AUC was similarly affected by itraconazole in all genotypes. The urinary 6-beta-hydroxycortisol/cortisol concentration ratio was halved by itraconazole, consistent with inhibition of CYP3A4. However, the effect of CYP3A4 inhibition on the pharmacokinetics of aripiprazole was not thought to be clinically significant. On the other hand, there were definite differences in pharmacokinetics between CYP2D6 genotypes.

Atenolol

The effect of itraconazole 200 mg bd for 2 days on the pharmacokinetics of atenolol 50 mg has been investigated in 10 healthy volunteers in a randomized crossover study [ ]. Itraconazole increased the AUC of atenolol and the amount excreted in the urine by about 12%, suggesting a slight increase in systemic availability. However, it had no statistically significant effect on the pharmacodynamics of atenolol.

Benzodiazepines

Carbamazepine

Fluconazole-induced carbamazepine toxicity has been reported [ ].

• A 29-year-old woman taking carbamazepine 1600 mg/day, lamotrigine 400 mg/day, and barbexaclone 100 mg/day developed severe diplopia, oscillopsia, nausea, vomiting, and gait instability within several days after starting to take fluconazole 150 mg/day for tinea corporis. The carbamazepine concentration was 1.5 times above the usual target range. Within 24 hours after withdrawal of fluconazole, the neurological deficits had disappeared and the carbamazepine concentrations had returned to the target range.

This interaction was probably due to inhibition by fluconazole of CYP3A4 and/or CYP2C9, isoenzymes that are involved in the metabolism of carbamazepine.

Cardiac glycosides

The effect of multiple-dose voriconazole on the steady-state pharmacokinetics of digoxin in healthy men has been studied in a double-blind, randomized, placebo-controlled study [ ]. All the subjects took oral digoxin for 22 days (0.5 mg bd on day 1, 0.25 mg bd on day 2 and 0.25 mg/day on days 3–22). On days 11–22 they were randomized to either voriconazole 200 mg bd or placebo. Voriconazole did not significantly alter the C _max , C _min , AUC, t _max , or clearance of digoxin at steady state. There were no significant differences in adverse events, all of which were classified as mild and transient.

Celiprolol

The effects of itraconazole on the pharmacokinetics of celiprolol has been investigated in a randomized crossover study in 12 healthy volunteers who took itraconazole 200 mg orally or placebo bd or grapefruit juice 200 ml tds for 2 days [ ]. On the morning of day 3, 1 hour after drug ingestion, each subject took celiprolol 100 mg with 200 ml of water (placebo and itraconazole phases) or grapefruit juice. During the itraconazole phase, the mean AUC from 0 to 33 hours of celiprolol was 80% greater than in the placebo phase. Cumulative urinary excretion of celiprolol was increased by itraconazole by 59%. Hemodynamic variables did not differ between the phases. Itraconazole almost doubles plasma celiprolol concentrations. This interaction probably results from increased availability of celiprolol, possibly as a result of inhibition of P glycoprotein in the intestine.

Ciclosporin

The extent of the pharmacokinetic interaction between ciclosporin and itraconazole oral solution in eight renal transplant recipients and the effect on daily drug costs has been determined in a single-center, open, non-randomized study [ ]. After transplantation, renal transplant recipients received itraconazole solution 200 mg bd and ciclosporin to achieve target blood concentrations. At steady state blood samples were collected over 12 hours for pharmacokinetic evaluation of ciclosporin, itraconazole, and hydroxyitraconazole. Itraconazole was withdrawn after about 3 months. Ciclosporin doses were again titrated to achieve target blood concentrations and ciclosporin concentrations were once again determined at steady state. Mean peak and trough itraconazole concentrations were 1.64 and 1.23 μg/ml respectively. Mean peak and trough hydroxyitraconazole concentrations were 2.37 and 2.20 μg/ml respectively. Itraconazole caused a 48% reduction in the mean total daily dose of ciclosporin necessary to maintain target concentrations, 171 versus 329 mg). This reduction in ciclosporin dose resulted in a discounted itraconazole daily drug cost of about 30% while providing antifungal coverage with adequate itraconazole trough concentrations.

In a retrospective study of 102 children with steroid-dependent nephrotic syndrome, 78 received daily ketoconazole 50 mg dose and a reduced dose of ciclosporin and 24 received ciclosporin alone [ ]. The mean duration of treatment was 23 months. Co-administration of ketoconazole significantly reduced the mean doses of ciclosporin by 48%, with a net cost saving of 38%. It also resulted in significant improvement in the response to ciclosporin, increased success in the withdrawal of steroids, and a reduced frequency of renal impairment.

A 14-year-old girl with an allogeneic bone marrow transplant stopped taking voriconazole because of worsening liver function tests; the ciclosporin trough blood concentrations fell [ ]. This observation emphasizes the need for careful monitoring and dosage adjustments of ciclosporin in patients who take antifungal azoles.

The outcomes in renal transplant patients have been monitored using simultaneous ciclosporin C0 and C2 concentration measurements and in patients in whom only ciclosporin C2 concentrations were measured [ ]. The latter had higher ciclosporin C2 concentrations, AUCs, and drug doses during the immediate postsurgical period, and at 2 weeks and 4 and 6 months after transplantation. Six of the latter and none of the former had severe liver toxicity, characterized by jaundice and raised liver enzymes, with negative serological tests for CMV, HVC, and HVB. There was a correlation between aspartate transaminase activity and ciclosporin C2 concentrations and both normalized at 15–55 days after ciclosporin dosage reduction. High ciclosporin C2 concentrations, which have been recommended when the drug is used alone in renal transplantation, cannot be used in patients taking ketoconazole, because C2 does not reflect drug exposure and high C2 concentrations can cause liver toxicity.

Cimetidine

The effect of itraconazole on the renal tubular secretion of cimetidine has been investigated in healthy volunteers who received intravenous cimetidine alone and after 3 days of oral itraconazole 400 mg/day [ ]. The cimetidine AUC increased by 25% after itraconazole. Glomerular filtration rate of cimetidine was unchanged, but secretory clearance was significantly reduced presumably due to inhibition of P glycoprotein.

Cyclophosphamide

Cyclophosphamide is a prodrug that is metabolized by CYP450 enzymes to cytotoxic alkylating species, and the extent of metabolism correlates with both efficacy and toxicity. In a randomized study of the safety and efficacy of itraconazole or fluconazole in preventing fungal infections in patients undergoing allogeneic stem cell transplantation, itraconazole (200 mg/day intravenously or 2.5 mg/kg orally tds) or fluconazole (400 mg/day intravenously or orally) were given with from start of conditioning therapy until at least 120 days after transplantation [ ]. After enrolment of the first 197 patients, a data and safety monitoring board reviewed the potentially drug-related adverse effects. Patients who had taken itraconazole had higher serum bilirubin and creatinine concentrations in the first 20 days after transplantation; the highest values were in patients who had taken itraconazole concurrently with cyclophosphamide conditioning. Analysis of cyclophosphamide metabolism in a subset of patients showed greater exposure to toxic metabolites (in particular 4-hydroxycyclophosphamide and 4-ketocyclophosphamide) among recipients of itraconazole compared with fluconazole. In contrast, those who took fluconazole had greater exposure to the unmetabolized drug. Adverse effects occurred preferentially in patients who had greater exposure to cyclophosphamide metabolites. These data suggest that azole antifungals, through differential inhibition of hepatic cytochrome P450 isozymes, affect cyclophosphamide metabolism and conditioning-related adverse effects after allogeneic stem cell transplantation.

In a randomized comparison of itraconazole (200 mg/day intravenously or 2.5 mg/kg tds orally) and fluconazole (400 mg/day intravenously or orally) in preventing fungal infections in patients undergoing allogeneic stem cell transplantation, those who received itraconazole developed higher serum bilirubin and creatinine concentrations in the first 20 days after transplantation, with the highest values in those who received concurrent cyclophosphamide [ ]. There was higher exposure to toxic metabolites (in particular 4-hydroxycyclophosphamide and 4-ketocyclophosphamide) among recipients of itraconazole compared with fluconazole. In contrast, recipients of fluconazole had higher exposure to unmetabolized drug. Adverse effects occurred preferentially in those who had higher exposure to cyclophosphamide metabolites. These data suggest that azole antifungals, by differential inhibition of hepatic CYP isoenzymes, affect cyclophosphamide metabolism.

Agents that are frequently co-administered with cyclophosphamide in high-dose chemotherapy regimens were tested for inhibition of the activation of cyclophosphamide in human liver microsomes. The K _m and V _max of the conversion of cyclophosphamide to 4-hydroxycyclophosphamide were 93 μmol/l and 4.3 mg.nmol/hour respectively; itraconazole was inhibitory at an IC50 of 5 μmol/l, which is higher than the usual plasma itraconazole concentration and was thus considered of no clinical relevance [ ].

Cytarabine

In vitro assays with itraconazole have shown that cytarabine is a substrate of CYP3A4 [ ]. Cytarabine and itraconazole inhibit CYP3A4. Cytarabine metabolism was significantly reduced when it was combined with itraconazole. Inhibition of cytarabine metabolism may have important clinical implications and warrants investigation in vivo.

Dapsone

The formation of dapsone hydroxylamine is thought to be the cause of the high rates of adverse reactions to dapsone in HIV-infected individuals. The effect of fluconazole on hydroxylamine formation in individuals with HIV infection has been investigated in 23 HIV-infected subjects [ ]. Fluconazole reduced the AUC, percent of dose excreted in urine in 24 hours, and formation clearance of the hydroxylamine by 49%, 53%, and 55% respectively. This inhibition of in vivo hydroxylamine formation was quantitatively consistent with that predicted from human liver microsomal experiments. Rifabutin had no effect on the plasma AUC of hydroxylamine or the percent excreted in the urine in 24 hours but increased formation clearance by 92%. Dapsone clearance was increased by rifabutin and rifabutin plus fluconazole (67% and 38% respectively) but was unaffected by fluconazole or clarithromycin. Hydroxylamine production was unaffected by clarithromycin. On the basis of these data, and assuming that exposure to dapsone hydroxylamine determines dapsone toxicity, we predict that co-administration of fluconazole should reduce the rate of adverse reactions to dapsone in people with HIV infection and that rifabutin and clarithromycin will have no effect. When dapsone is given in combination with rifabutin, dapsone dosage adjustment may be necessary.

Dexloxiglumide

Dexloxiglumide is a cholecystokinin CCK1 receptor antagonist under investigation for functional gastrointestinal disorders; it is metabolized by CYP3A4 and CYP2C9. The effect of steady-state ketoconazole on the pharmacokinetics of dexloxiglumide and its primary metabolite O-demethyldexloxiglumide has been studied in healthy subjects in a randomized, two-period, crossover study [ ]. Ketoconazole increased dexloxiglumide C _max by 32% without affecting the C _max of changed O-demethyldexloxiglumide and increased the AUC of dexloxiglumide and O-demethyldexloxiglumide by 36%. There were no changes in the half-lives of dexloxiglumide or O-demethyldexloxiglumide.