Answers are grounded in this paper and cite their sources.
Five years have passed since the Food and Drug Administration (FDA) issued its directives restricting the use of ketoconazole as an antifungal agent in clinical practice, and discouraging the use of ketoconazole as an index inhibitor of cytochrome P450-3A (CYP3A) metabolic activity in drug-drug interaction studies involving healthy volunteers.1-3 The initial Drug Safety Communication dated July 25, 2013, stated that the FDA “conducted a comprehensive benefit-risk assessment of the safety and efficacy of Nizoral (ketoconazole) tablets in the context of the drug's labeled indications for the treatment of superficial and systemic fungal infections …” The communication goes on to state that data retrieved from the FDA Adverse Event Reporting System (AERS) were assessed independently by a hepatology expert in FDA. In a follow-up Drug Safety notice dated October 16 of the same year, the agency recommended that drug companies and researchers “avoid using oral ketoconazole in drug interaction studies.” Findings from the FDA's analysis of ketoconazole liver injury have not been published in the biomedical literature, but are available in documents11 The documents are available at: https://www.citizen.org/our-work/health-and-safety/petition-fda-ban-all-oral-forms-anti-fungal-drug-ketoconazole-0 and https://www.regulations.gov/docket?D=FDA-2015-P-0578. relating to the Public Citizen Health Research Group's petition to the FDA “ … to immediately require the removal from the market of the oral form of the antifungal agent ketoconazole,“ dated February 24, 2015. Appended to the petition was a document dated January 4, 2013, from the FDA's Office of Surveillance and Epidemiology, titled “Benefit-Risk Assessment of Oral Ketoconazole Tablets.” On September 1, 2017, the FDA issued a response denying the petition. In this Commentary, we update the current state of scientific knowledge on the topic of liver injury associated with ketoconazole and other azole antifungal agents. The only scientific publication cited by the FDA in the original 2013 safety communication was the study reported by Garcia Rodriguez et al.4 Among nearly 70,000 patients who received oral antifungal agents during the 5-year surveillance period, only 5 cases of liver injury were reported. Of these, 2 were associated with ketoconazole, 2 with itraconazole, and 1 with terbinafine. The calculated relative risk was highest with ketoconazole, but the very small number of cases makes it unrealistic to draw conclusions regarding comparative risk. A meta-analysis of available studies — both published and unpublished — reported by Yan et al5 found an overall incidence of ketoconazole-associated liver injury in the range of 3.6% to 4.2% of patients. Liver injury was identified solely on the basis of abnormal liver function tests, and there was no available comparison with other antifungals. Kao et al6 reported on azole antifungal-associated liver injury based on the National Health Insurance Program in Taiwan. In this study, the criteria for liver injury were confined to more serious hepatotoxicity, as opposed to liver function test abnormalities alone. Of nearly 91,000 patients who were exposed to azole antifungals, a total of 52 had liver injury. The overall incidence was highest with fluconazole (31.6 per 10,000 persons), and 6 of the 7 deaths from hepatotoxicity were associated with fluconazole. The incidence of liver injury with ketoconazole was 4.9 per 10,000 patients, and the risk appeared to increase with longer durations of treatment. Lo Re et al7 studied liver injury among more than 195,000 patients exposed to azole antifungals in the Kaiser Permanente Northern California system. Incidence rates were low for all antifungals, and could not be statistically distinguished among fluconazole, ketoconazole, and itraconazole. Rates were quantitatively higher — although not significantly so — for voriconazole and posaconazole, but the numbers of exposed patients were much smaller, and incidences may have been influenced by preexisting liver disease, and/or by greater severity of underlying illness. Two reports deal with data from the FDA AERS.8, 9 The drawbacks of spontaneous adverse event reporting systems are well-recognized, including the very uncertain validity of inferences regarding absolute or comparative event incidence.10, 11 Insofar as azole-associated liver injury could be compared among the agents in this class, the odds ratio confidence intervals overlap so much that the drugs cannot be realistically distinguished.9 In contrast to ketoconazole use to treat fungal infections in clinical practice, its use in drug-drug interaction (DDI) studies in healthy volunteers involves a short duration of exposure under monitored conditions in young adult subjects with no underlying liver disease or other risk factors. Two studies, conducted independently and published at about the same time, evaluated ketoconazole-associated liver injury in the context of healthy volunteer DDI pharmacokinetic studies.2, 12 Liver injury was defined based on biochemical criteria. The report of Banankhah et al12 evaluated 53 DDI studies involving 971 subjects who were exposed to systemic ketoconazole. The median dosage of ketoconazole was 400 mg per day, and the mediation duration of exposure was 7 days (maximum: 28 days). There were 6 reports of liver function test abnormalities among the 971 subjects, of which 4 were judged to be possibly or probably related to ketoconazole (0.4% incidence). All subjects remained clinically asymptomatic, and biochemical abnormalities returned to baseline after completion of the study. Outeiro et al2 evaluated 161 DDI studies involving 2355 subjects. The dosage and duration of ketoconazole exposure were similar to those reported by Banankhah et al,12 and it is likely that there was some redundancy in the DDI studies evaluated in the 2 reports. Outeiro et al2 found that 40 or the 2355 participants (1.7%) experienced liver function test abnormalities. These were asymptomatic and reversible. Ketoconazole has long served as the prototype CYP3A inhibitor in DDI studies. It has very high potency as a reversible CYP3A inhibitor,13, 14 is relatively selective for CYP3A, and requires only one day of pre-exposure to produce maximum CYP3A inhibition.15, 16 CYP3A metabolic phenotype returns rapidly to baseline after ketoconazole discontinuation, thereby facilitating crossover studies. The safety data discussed above indicate minimal risk associated with ketoconazole in the context of DDI studies. In the event that investigators and sponsors remain (unnecessarily) wary of the liver injury risk, ritonavir and cobicistat are the logical alternative CYP3A inhibitors.3, 17, 18 At low “boosting” doses, both are relatively selective for CYP3A,19, 20 produce in vivo CYP3A inhibition comparable to ketoconazole,2, 17, 18, 21 require only one day of pre-exposure to produce maximal inhibition, and allow reversion of CYP3A phenotype to baseline shortly after discontinuation. The probability of enzyme induction with boosting doses of ritonavir or cobicistat is minimal. The suggestion that itraconazole or clarithromycin could serve as alternatives to ketoconazole22, 23 is not supported by the scientific evidence.3, 17, 18 Neither itraconazole nor clarithromycin produces in vivo CYP3A inhibition comparable to ketoconazole, ritonavir, or cobicistat. Itraconazole is further complicated by a long half-life and pharmacologically active metabolites, which prolong the pretreatment period as well as the posttreatment washout period. Absorption of oral itraconazole is modified by a significant food effect. In addition to being a less potent CYP3A inhibitor, clarithromycin is complicated by its status as a mechanism-based (time-dependent) inhibitor. Five years after the regulatory restrictions and labeling changes imposed on ketoconazole, the scientific bases for these actions remain inconclusive. The available epidemiologic evidence fails to validate the proposition that the risk of liver injury during clinical use of ketoconazole is different from that associated with other azole antifungals. Prohibition of ketoconazole probably does not impact patient welfare in terms of antifungal treatment efficacy, but there is no evidence to indicate that the use of other systemic antifungals will reduce an already low risk of treatment-related liver injury. In the context of clinical DDI studies in which exposure duration is limited and predisposing risk factors are not present, the evidence strongly indicates that use of ketoconazole as an index CYP3A inhibitor carries minimal hazard. If an alternative inhibitor is to be used, ritonavir and cobicistat are logical choices, but itraconazole and clarithromycin are not.
No takes yet. Share an insight, caveat, or question.
Greenblatt et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: