Navigating the complexities of azole antifungal therapy through pharmacokinetic concepts: a case of prolonged isavuconazole toxicity

Le résumé fourni par la source

Pharmacokinetics plays a significant role in azole antifungals effectiveness and tolerability, influencing treatment outcomes for patients. Voriconazole is primarily metabolized by the hepatic cytochrome P450 (CYP) 2C19, while isavuconazole is primarily processed by the CYP3A4/5. In contrast, the metabolism of posaconazole occurs mainly through glucuronidation.1 Understanding these pharmacokinetic profiles is essential for optimizing therapeutic regimens and minimizing the risk of adverse effects. This case study illustrates the complexities of managing a patient on isavuconazole, highlighting the significance of individualized pharmacokinetic considerations in clinical practice. Mr X, a 68-year-old male, was a heart recipient, had Stage 4 chronic kidney disease and a history of several infections. His medications included tacrolimus and esomeprazole. He experienced a series of adverse reactions to different azole antifungals used to treat an aspergillosis. He initially received voriconazole, but hepatic and renal toxicities were observed on Day 11. Voriconazole was not overdosed (trough concentration at 3.6 mg/L, usual target range: 1–4 mg/L).1 Subsequently, isavuconazole was administered from Day 24. Due to a major overdose (serum trough concentration of 9.96 mg/L) on Day 72, treatment was halted, resumed at 100 mg/day from Day 80, but stopped again on Day 93. Mr X showed fatigue, unusual dyspnoea, altered taste, loss of appetite and significant asthenia with isavuconazole. Posaconazole was initiated (Day 94), with acceptable trough concentration throughout the treatment (2.0 mg/L on Day 130, for a target range between 1 and 3 mg/L) (see Figure 1).1 His symptoms persisted for 1 month after he started posaconazole and discontinued isavuconazole, suggesting possible ‘cross’ intolerance to all azole antifungals. However, pharmacokinetics may offer a different explanation. Thirty days after discontinuation, the isavuconazole concentration remained at 2.05 mg/L. The pulmonary aspergillosis showed favourable progress with regression of the lesion, but the patient died on Day 133 due to multiorgan failure following a SARS-CoV-2 infection. Chronological representation of medication periods and isavuconazole and posaconazole concentrations. The figure legend denotes periods of antifungal administration with rectangles (for VORIconazole, ISAvuconazole and POSAconazole), modelled Isavuconazole concentrations using the Shirae model,2 with median concentrations indicated by solid line, observed Isavuconazole concentrations indicated by stars and toxic threshold drawn using a dashed line. Posaconazole trough concentration is also indicated using a bold star. Therapeutic drug monitoring (TDM) may be useful in patients who do not respond to isavuconazole therapy, have unexpected toxicity or drug interactions, or in the treatment of pathogens with high MICs or in sanctuary sites like the CNS.3 In the absence of well-defined therapeutic targets for isavuconazole, a plasma trough level in the range of 2–3 mg/L after Day 5 of treatment (including loading doses) is suggested based on data from Phase II/III clinical studies. This range is considered to indicate adequate drug exposure.3 Toxicity of isavuconazole is considered to occur at trough concentrations >5 mg/L, based on findings that adverse events were observed in patients with median concentrations of 5.025 mg/L.4 The patient’s concentration was nearly twice this threshold, which likely explains the experienced adverse effects. Theoretically, based on usual half-life between 100 and 130 h,1 isavuconazole should have been totally eliminated 30 days after its discontinuation, but its concentration was still in the therapeutic range. Isavuconazole metabolism primarily involves CYP3A4/5 enzymes.1 Genetic polymorphisms, such as CYP3A4*22 and CYP3A5*3, can potentially reduce its clearance. The CYP3A5*3 allele is predominantly expressed in approximately 92% of the European population, which limits its impact on overall enzyme reduced activity. Nevertheless, patients with homozygous CYP3A5*3 variant may experience a prolonged isavuconazole half-life.5 In Mr X’case, significant concentrations of tacrolimus, another CYP3A4/5 substrate, were observed despite minimal dosing (i.e. 6 ng/mL with a 0.5 mg prolonged-release dose), which aligns with the observation that poor CYP3A expressors tend to have a higher tacrolimus concentration-to-dose ratio.6 Additional factors, such as concomitant medication like esomeprazole, may exacerbate isavuconazole exposure, as esomeprazole has been described as a weak CYP3A4 inhibitor.7 The switch to posaconazole further inhibited CYP3A4/5, potentially exacerbating isavuconazole's slow metabolism. Concomitant administration of ketoconazole (a strong CYP3A4 inhibitor) increased the AUC0-∞ of isavuconazole by 422%.8 Posaconazole is a potent but less intense CYP3A4 inhibitor (reducing metabolic clearance of midazolam, a CYP3A4 substrate, by 60% versus by 90% with ketoconazole),9 resulting in an increase in isavuconazole’s AUC by around 250%–300%. Moreover, C-reactive protein was very high from Day 104 to Day 133 (between 53 and 286 mg/L). Inflammation is a major regulator of drug metabolizing enzymes and transporters,10 such as CYP3A4, resulting in increased concentration-to-dose ratios and decreased clearance. In particular, inflammatory cytokines like IL-6 and IL-1β have been shown to significantly downregulate CYP3A4 activity,11 which may alter drug metabolism and response, even if data on isavuconazole are scarse. Prolonged isavuconazole clearance should be expected when switching to another azole antifungal because of CYP inhibition. TDM may be useful in this context to guide a potential washout period and avoid disabling toxicities. We conducted simulations using a population pharmacokinetic model that did not account for genetic polymorphism.2 Simulations were conducted on R software version 4.1.2 (R Foundation for Statistical Computing, Vienna, Austria). The patient’s concentrations were plotted against these simulations to assess their alignment (see Figure 1), but the observed concentrations were significantly higher than expected, confirming that its isavuconazole pharmacokinetics should be explained by other covariables (such as a major CYP inhibition). Estimating elimination clearance via the Bayesian method, given the limited number of available concentrations, would be biased towards the population mean due to the shrinkage phenomenon. The two concentrations used for linear regression were measured on Day 123 and Day 129 (i.e. 2.25 and 2.05 mg/L at 2375 and 2532 h post-isavuconazole introduction, respectively), representing the terminal phase of the drug’s elimination curve. These data points allowed for the estimation of the terminal half-life by calculating the slope of the regression line. The terminal half-life was 1169 h, a 10-fold increased value compared to its usual half-life. Our findings highlight the significance of monitoring plasma isavuconazole trough levels due to the reported linear increase of 0.032 mg/L per day during prolonged therapy4 and the persistence of drug concentrations in the blood following treatment cessation, which may have important implications for patient management and toxicity assessment. Long-lasting toxicities were assumed to be due to high concentrations combined with decreased clearance, resulting from a potential genetic deficit, drug–drug interactions, and hyperinflammation. Moreover, the high volume of distribution of approximately 450 L for isavuconazole indicates extensive tissue distribution,12 which also prolongs the washout process after treatment discontinuation. To optimize patient management and understand residual effects, TDM of isavuconazole should be conducted. Additionally, estimating its elimination half-life seems essential before switching to another antifungal therapy. None. None to declare.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.