Current approaches and future strategies for the implementation of pharmacogenomics in the clinical use of azole antifungal drugs

1. Introduction

Azoles are a heterogeneous group of antifungal drugs. They can be classified into two groups, those with two nitrogens in the azole ring, called imidazoles (include clotrimazole, econazole, ketoconazole, miconazole, and tioconazole) and those with three nitrogens in the azole ring, called triazoles (fluconazole, itraconazole, posaconazole, voriconazole and isavuconazole).

Agents within the azole class work primarily by inhibiting the fungal cytochrome P450 enzyme lanosterol 14α-demethylase. This leads to disruption of ergosterol biosynthesis, a component of the fungal cell membrane, which alters the functionality and structure of the fungal cell membrane, resulting in cell lysis and death [1].

Imidazoles are now usually confined to topical use due to its unfavorable safety profile, while triazoles, with superior pharmacokinetics and improved safety profiles, are widely used for prophylaxis and treatment of invasive fungal disease, such as invasive aspergillosis, for which voriconazole has become the standard of care [1,2].

Pharmacokinetic properties vary between drugs, but all classes of azoles have the potential to be both substrates and inhibitors of different CYP (cytochrome P450) enzymes, conjugative enzymes and transport proteins [2]. Polymorphic expression of these enzymes and transporters has been widely recognized as an important source of variability in the disposition and response to these drugs, which makes them potential biomarkers to predict optimal doses and drug response. However, the implementation of these biomarkers in the clinical setting has been uneven [3].

2. Body

The difficulties to validate and implement the use of these biomarkers are well described in the literature, including low institutional promotion, insufficiency of available evidence and cost-effectiveness, lack of clinical guidelines and protocols, and economic and institutional issues as well as ethical, legal, and social implications [4,5]. Implementation of pharmacogenetics into clinical practice needs of a sequential process. Once a gene-drug interaction is identified, evidence of statistically associations between the pharmacogene and clinical outcomes needs to be provided. The prevalence of the clinically actionable pharmacogenes in the target population is another key aspect to consider. In addition, to assist physicians to apply this information, pharmacogenomic guidelines with consistent recommendations have to be developed. Finally, demonstrating cost-effectiveness and impact on cost is essential to justify and facilitate the adoption of Pharmacogenetics as a tool able to save significant costs to healthcare systems [5,6].

Polymorphic expression of genes coding azole-metabolizing enzymes and transporters has been described, and some of these genetic variations have been associated to interindividual heterogeneity in disposition, response, or toxicity of these drugs. Still, for the majority of azoles there are no clinical guidelines or recommendations that take into account the pharmacogenetic information to achieve an accurate and optimized therapeutic plan.

In a review done in the Pharmacogenomics Knowledgebase website (PharmGKB; www.pharmgkb.org) no actionable pharmacogenetic information was found for imidazole drugs, no clinical or drug label annotations and no prescribing information were available. Only a few variant annotations (genetic variant–drug associations as reported in a single publication) were described for ketoconazole. Information found for most triazoles was in line with the previous group. There is no information about isavuconazole, some variant annotations are available for fluconazole, one for itraconazole, and there is one annotation of EMA Label for posaconazole and CYP3A4 but it is related to metabolism drug–drug interactions with CYP3A4 substrates. On the contrary, there is significant information for voriconazole: clinical annotations, drug label annotations with actionable pharmacogenetics information and prescribing information including clinical guidelines providing drug dosing information based on pharmacogenetic data [7].

The lack of pharmacogenetics information available for azoles except voriconazole can be explained by diverse reasons. All triazoles are substrates of CYP3A4 to a different extent, and although interindividual variability in CYP3A4 activity exists, only a few common genetic variants have been described and so far there is scarce evidence supporting dose adjustment for the majority of drugs metabolized by CYP3A4 [3,8]. Only isavuconazole undergoes metabolism by CYP3A5, but to a lesser extent than by CYP3A4 [9,10]. Thus far, the documented influence of CYP3A4 and CYP3A5 genotype on azoles is mostly related to interactions with CYP3A4/5 substrates, as they may significantly increase its concentrations in individuals homozygous for CYP3A5*3 variant, which is associated with a lack of CYP3A5 activity [3,11].

The triazoles are also substrates and/or inhibitors of conjugative enzymes and transporter proteins such as uridine diphosphate glucuronosyltransferase (UGT) or P-Glycoprotein (P-gp), but further insights are needed to clarify its role in triazole disposition, efficacy, and drug interactions [2,3]. Additionally, polymorphisms in genes coding regulatory proteins that interact with CYP enzymes, such as POR or NR1I2, seem to have an additive effect on the pharmacokinetics of some CYP-metabolized drugs, but not much actionable information is available in relation to this [12].

Two enzymes of the CYP2C subfamily are involved in fluconazole and voriconazole, metabolism, CYP2C19 and CYP2C9 [3,13]. The CYP2C9 gene is highly polymorphic, but the most prevalent allele is CYP2C9*1, associated with normal enzyme activity, being the variants with decreased function notably infrequent [3,14]. Furthermore, fluconazole is mostly excreted as unchanged drug, and its metabolic pathway is not well described.

Conversely, lot of information is available related to Voriconazole metabolism. It is estimated that around 70–75% of its total metabolism is mediated through CYP P450 enzymes, mostly CYP2C19, CYP3A4, and CYP2C9, and the remaining 25–30% is mediated by the FMO (flavin-containing monooxy- genase) family [12]. Despite several SNPs located in the genes associated with voriconazole metabolism have been related to variability in its clearance, to date, dosing guidelines for voriconazole are monogenic and rely exclusively on CYP2C19 [15]. Further, even when guidelines are available, ambiguity in how to apply pharmacogenomic biomarker information exists. The Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for voriconazole and CYP2C19, recommends to select other antifungal agents in those patients with ultrarapid and poor metabolizer genotypes, and to prescribe the usual dosing in other genotypes. In contrast, the Dutch Pharmacogenetics Working Group guideline propose a dose adjustments for ultrarapid and poor metabolizer genotypes. Also, Hicks et al. propose a dose adjustment based on CYP2C19 genotype in children with leukemia, which would be a very good option for this first line drug in invasive aspergillosis [15–18].

Table 1 summarizes for each azole the gene–drug interactions described in the literature and the level of evidence.

Table 1. Summary of gene–drug interaction described in the literature for each azole and level of evidence

		Gene	Description of Gene-Drug Interaction	Level of evidence^α	References
Imidazoles	Clotrimazole¥	Ø
	Econazole¥	Ø
	Ketoconazole ¥	CYP3A5	CYP3A5 *1/*1 and CYP3A5*1/*3 are associated with decreased severity of inhibition of CYP3A-catalyzed metabolism of midazolam or testosterone when treated with ketoconazole as compared to CYP3A5 *3/*3.	Variant annotation -Cohort study. N = 9	PMID: 23,723,360
		CYP2C19	Some CYP2C19 alleles are associated with decreased sensitivity or not sensitivity to ketoconazole, as compared to CYP2C19 *1.	Variant annotation -One study in vitro	PMID: 21,692,664
	Miconazole¥	Ø
	Tioconazole¥	Ø
Triazoles	Fluconazole¥	CYP2C19	Different alleles are associated with different sensitivity to fluconazole as compared to CYP2C19 *1.	Variant annotation -One study in vitro	PMID: 21,692,664
		CYP2C9	Some alleles are associated with decreased inhibition of CYP2C9 when exposed to fluconazole or voriconazole as compared to CYP2C9 *1 and CYP2C9 *3.	Variant annotation -One study	PMID: 26,886,310
	Itraconazole¥	CYP3A5	CYP3A5 *1/*1 and CYP3A5*1/*3 are associated with decreased severity of inhibition of CYP3A-catalyzed metabolism of midazolam or testosterone when treated with itraconazole as compared to CYP3A5 *3/*3.	Variant annotation -Cohort study. N = 9	PMID: 23,723,360
	Posaconazole¥	CYP3A4	Information related metabolism drug-drug interactions with CYP3A4 substrates.	Drug Label Annotations -Informative PGx (not actionable)	Posaconazole EMA drug label.
	Voriconazole¥	CYP2C19	CYP2C19 genotype is associated with the pharmacokinetics variability of voriconazole.	Clinical annotation with evidence Level 1A	PMID:15,179,414; PMID:25,558,073; PMID:25,239,277; PMID:25,558,073; PMID:25,558,073; PMID:30,653,146;
				Precribing information available: -Clinical Guideline Annotations (CPIC *and DPWG**).	PMID: 27,981,572; Dutch Pharmacogenetics Working Group. Pharmacogenetic Recommendations. 2020. https://www.knmp.nl/@@searc h;
				Several variant annotations	PMID: 28,834,922; PMID:16,890,574; PMID:30,464,555;
				Drug Label Annotations available: -Annotation of FDA, Swissmedic, PMDA and HCSC Label for voriconazole and CYP2C19 with actionable PGx information and EMA Label for voriconazole and CYP2C19 with Informative PGx information.	Voriconazole (VFEND) EMA drug label.
		CYP3A4	CYP3A4 genotype may impact voriconazole plasma concentrations.	Clinical annotation with evidence Level 2 {This level of evidence has been described in patients with invasive fungal infections}	PMID:25,515,945
				Variant annotations	PMID: 28,786,228
				Drug Label Annotations available: Annotation of EMA Label for voriconazole and CYP3A4 with Informative PGx information.	Voriconazole (VFEND) EMA drug label.
				Other sources	PMID: 25,645,831; PMID: 27,937,048; PMID: 26,775,563; PMID: 28,786,218
		ABCB1	ABCB1 genotype may impact voriconazole clearance.	Clinical annotation with evidence Level 2	PMID:19,033,450
		CYP2C9	Some alleles are associated with decreased inhibition of CYP2C9 when exposed to fluconazole or voriconazole as compared to CYP2C9 *1 and CYP2C9 *3.	Variant annotation -One study	PMID: 26,886,310;
			-Association between genotype and likelihood of live toxicity.	Variant annotation - Cohort, retrospective study (n = 84)	PMID:17,827,141
				Drug Label Annotations available: Annotation of EMA Label for voriconazole and CYP2C9 with Informative PGx information.	Voriconazole (VFEND) EMA drug label.
				Other sources - Cohort study (n = 106)	PMID: 31,279,853
		CYP3A5	Association between genotype and likelihood of live toxicity.	Variant annotation - Cohort, retrospective study (n = 84)	PMID:17,827,141
		NR1I2	Association between NR1I2 genotype and concentrations of voriconazole .	Variant annotation - Cohort study (n = 172)	PMID: 31,932,875
				Other sources - Cohort study (n = 106)	PMID: 31,279,853
		FMO3	Association between FMO3 genotype and concentrations of voriconazole .	Variant annotation - Cohort study (n = 172)	PMID: 31,932,875
				Other sources - Cohort study (n = 106)	PMID: 31,279,853
		POR	Association between FMO3 genotype and concentrations of voriconazole .	Variant annotation - Cohort study (n = 172)	PMID: 31,932,875
				Other sources - Cohort study (n = 106)	PMID: 31,279,853
	Isavuconazole¥	Ø

Ø No information available.

¥Molecular structure source: PubChem

^α PharmGKB Level of Evidence.

* CPIC: Clinical Pharmacogenetics Implementation Consortium

**DPWG: Dutch Pharmacogenetics Working Group;

Abbreviations: FDA: U.S. Food and Drug Administration; EMA: European Medicines Agency; Swissmedic: Swiss Agency of Therapeutic Products; PMDA: Pharmaceuticals and Medical Devices Agency, Japan; HCSC: Health Canada (Santé Canada)

3. Expert opinion

Pharmacogenetics has emerged in the last years as a key discipline in the progress toward personalized medicine. Despite the fact that data available to this subject are experiencing exponential growth, pharmacogenomic testing are rapidly evolving and its costs progressively dropping, the clinical utility and cost-effectiveness of pharmacogenetic testing is sometimes restricted to certain gene–drug combinations and/or specific populations.

Herein, we have summarized the current status of Pharmacogenetics in the clinical use of azole antifungal drugs. Although several gene gene–drug associations are described for different azoles, for the majority of combinations, current levels of evidence are insufficient to support recommendations for changing clinical practice.

Evidence to support clinical recommendations is only available for voriconazole, albeit recommendations are not concurrent and large-scale implementation is not established.

It is difficult to predict how implementation of pharmacogenomics in the clinical use of azole will be carried out in the future. High-level evidence on genetic markers’ efficacy, effectivity and efficiency, needs to be obtained along with consensus on recommendations. To this aim, large population-based studies associating these biomarkers to drug response should be undertaken.

We suggest that, the ideal strategy to implement into clinical practice an efficacious strategy for dose individualization would be the development of prediction algorithms based on those factors contributing to drug variability in efficacy and safety, including pharmacogenetic and other clinical and/or physiological information. We also find that a preemptive genotyping strategy in risk populations (those who are susceptible to receive a drug) for which the pharmacogenetics information would be available at the moment of prescription allowing a precise initial dosing, is essential for bringing pharmacogenetics and clinics together.

In the specific case of voriconazole, there is sufficient evidence supporting implementation of pharmacogenetics in the clinical use (plenty of association studies and clinical guidelines are available), but it is necessary to evaluate the effectiveness and efficiency of the implementation. In this regard, our group is developing a multicenter, randomized clinical trial to evaluate the effectiveness and efficiency of a preemptive genotyping strategy for voriconazole, including an economic evaluation from the perspective of the Spanish National Health System [19]. In our view, the data available strongly support the hypothesis that voriconazole pharmacogenetics combined with close therapeutic drug monitoring of serum levels (which is already implemented in many hospitals), could improve the individualization of voriconazole dosing and we hope that the information obtained from this clinical trial will contribute to facilitate the implementation of pharmacogenetics into clinical practice [20].

Recent studies suggest that as drug exposure can be influenced by inflammation-induced phenoconversion, in addition to pharmacogenetics, the inflammatory status should be also taken into account when selecting voriconazole dose, especially in individuals with invasive fungal infections [21].

With respect to other azoles, further studies are needed to obtain evidence to support the influence of genomic biomarkers over dosing, safety, or efficacy.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.