Antifungal Drug Interactions and Therapeutic Drug Monitoring

Tacrolimus and cyclosporine are metabolized primarily by CYP3A4 and transported by P-gp; both are inhibited by all azoles to varying degrees. Voriconazole and posaconazole produce the largest increases in calcineurin inhibitor concentrations — typically requiring tacrolimus dose reductions of 50 to 75 percent and cyclosporine reductions of 25 to 50 percent when an azole is initiated. Itraconazole elevates calcineurin inhibitor concentrations substantially through combined CYP3A4 and P-gp inhibition. Fluconazole has a more modest effect, typically requiring tacrolimus reductions of 25 to 50 percent. Isavuconazole elevates calcineurin inhibitor concentrations to a lesser degree than voriconazole or posaconazole, but dose reduction and monitoring are still required. These reductions must be made proactively before the azole reaches steady state — reactive adjustment after the calcineurin inhibitor trough rises risks nephrotoxicity and neurotoxicity from supratherapeutic exposure.

Sirolimus is an mammalian target of rapamycin (mTOR) inhibitor used in renal transplantation and as anti-graft-versus-host disease (GVHD) therapy. It is metabolized almost exclusively by CYP3A4 and has a narrow therapeutic index. Voriconazole and posaconazole increase sirolimus area under the concentration-time curve (AUC) by approximately 500 to 1000 percent — from therapeutic to profoundly toxic — making co-administration categorically contraindicated with both agents. Isavuconazole produces a substantial but smaller increase and requires careful TDM with dose reduction if the combination must be used. Everolimus has a similar but somewhat less extreme interaction profile and requires TDM whenever used with any azole.

Warfarin is metabolized by CYP2C9 (the more pharmacologically active S-warfarin enantiomer) and CYP3A4 (R-warfarin). Fluconazole and voriconazole are the most potent inhibitors of CYP2C9 among the azoles and significantly elevate the international normalized ratio (INR) by reducing S-warfarin clearance — increases of two to three-fold are reported within days of starting the azole. INR monitoring within one to two weeks of starting or stopping an azole is mandatory, and warfarin dose reductions of 25 to 50 percent are commonly required. Posaconazole affects warfarin primarily through CYP3A4 inhibition; the interaction is clinically significant but typically less severe than with fluconazole or voriconazole. Direct oral anticoagulants (DOACs) — including rivaroxaban and apixaban — are cytochrome P450 3A4 and P-glycoprotein substrates and are also affected, with exposure potentially increasing to hemorrhage-risk concentrations when combined with potent azoles.

Simvastatin and lovastatin are highly dependent on CYP3A4 for first-pass and systemic metabolism; itraconazole and other potent CYP3A4 inhibitor azoles increase plasma concentrations dramatically, elevating the risk of myopathy and rhabdomyolysis. These combinations are contraindicated; pravastatin or rosuvastatin should be substituted. Midazolam and other CYP3A4-metabolized benzodiazepines may reach dangerous concentrations with azole co-administration, particularly in perioperative and intensive care unit (ICU) settings. Fentanyl and oxycodone are also CYP3A4 substrates and can accumulate to respiratory-depression concentrations when potent CYP3A4 inhibitor azoles are co-administered.

Four conditions must be present for TDM to deliver meaningful clinical benefit: demonstrable exposure-response relationships for both efficacy and toxicity; a narrow therapeutic index; high interpatient pharmacokinetic variability; and non-linear or unpredictable pharmacokinetics. Voriconazole satisfies all four with exceptional force. The published coefficient of variation for voriconazole trough concentrations at standard doses exceeds 80 percent — driven by cytochrome P450 2C19 (CYP2C19) genotype, comedications, hepatic function, and severity of illness. Two patients receiving identical doses may have trough concentrations differing by a factor of ten or more. Without TDM, approximately 30 to 50 percent of patients will have subtherapeutic concentrations and another 20 to 30 percent will have supratherapeutic concentrations.

Multiple clinical studies have established concentration-outcome relationships for voriconazole. Trough concentrations below 1.0 to 1.5 mg/L are consistently associated with increased treatment failure in invasive aspergillosis. Conversely, trough concentrations above 5.0 to 5.5 mg/L are associated with hepatotoxicity, neurotoxicity (hallucinations, encephalopathy, delirium), and visual adverse effects. The therapeutic window of approximately 1.0 to 5.5 mg/L forms the basis of clinical TDM practice. Achieving this window cannot be reliably accomplished by dose adjustment alone without measured concentrations — the interpatient variability is simply too large.

The CYP2C19 genotype is the most important driver of voriconazole pharmacokinetic variability. Poor metabolizers (PM), who carry two loss-of-function CYP2C19 alleles, achieve plasma concentrations four to five times higher than normal metabolizers at the same dose. Ultrarapid metabolizers (UM) achieve concentrations up to 50 percent lower. Hepatic function significantly affects voriconazole clearance: Child-Pugh class B or C liver disease elevates concentrations substantially. Proton pump inhibitors such as omeprazole inhibit CYP2C19 and further elevate voriconazole concentrations. Inflammation and critical illness reduce CYP enzyme activity and can acutely increase exposure even without dose changes. The combination of these variables makes TDM at steady state — day five to seven of therapy — essential rather than optional in any patient receiving voriconazole for serious infection.

Before prescribing any azole, review the patient's complete medication list against a validated drug interaction database — not memory. The three interaction pairs that most frequently produce serious harm and are most commonly missed are: azole plus calcineurin inhibitor (missed because the immunosuppressant dose is managed by a separate transplant team); azole plus sirolimus (missed because sirolimus is used intermittently and may not appear on the active medication list); and azole plus warfarin (missed because warfarin is prescribed by a primary care provider not involved in the antifungal decision). For any patient in a transplant program, the antifungal prescriber must notify the transplant pharmacist before administering the first azole dose so that calcineurin inhibitor dose adjustment is coordinated.

When more than one antifungal agent is clinically appropriate for the indication, choose the agent with the lowest interaction burden for that patient's specific drug regimen. For invasive candidiasis in a transplant patient on cyclosporine: an echinocandin (micafungin or anidulafungin) avoids the CYP interaction entirely and is first-line anyway. For Aspergillus prophylaxis in a patient on warfarin: posaconazole (CYP3A4 inhibitor, less effect on CYP2C9) has a smaller warfarin interaction than voriconazole or fluconazole, though close INR monitoring is still required. For a patient on rifampin who requires an antifungal for mucormycosis: rifampin renders all azoles subtherapeutic — liposomal amphotericin B is the only appropriate choice and has no CYP interactions.

For any patient receiving voriconazole: obtain trough concentration at day five to seven, after any dose change, after any interacting drug is added or removed, and when hepatic function changes. For posaconazole suspension: obtain trough at day five to seven; repeat if GI function or absorption-affecting medications change. For itraconazole: obtain trough after 14 days of stable therapy; interpret combined itraconazole plus hydroxy-itraconazole concentration if the assay measures both. For echinocandins: TDM is not routinely indicated; focus monitoring on liver function tests, potassium, and magnesium. For liposomal amphotericin B: TDM is not performed; monitor creatinine, potassium, and magnesium with every infusion.