Ketoconazole, the first broad-spectrum systemic imidazole antifungal agent, was approved by the US FDA in 1981 for use in humans.1 However, it was withdrawn from the human European and Australian markets and designated for restricted use in North America by 2013 because of rare hepatotoxicosis, leading to death or the need for liver transplantation.1 Although ketoconazole is not FDA approved for veterinary use, it is often prescribed for dogs with atopy complicated by Malassezia infection.2 Indeed, evidence indicates that ketoconazole administration at a dosage of 10 mg/kg/d (4.5 mg/lb/d), PO, for 3 weeks is effective in the treatment of atopic dogs with cutaneous Malassezia infection.3 In addition to its antifungal spectrum, ketoconazole has immunomodulatory and anti-inflammatory effects through suppression of T-cell proliferation and 5-lipoxygenase activity.4,5 Less frequently, ketoconazole is used for the treatment of hyperadrenocorticism or to reduce the required dose of cyclosporine owing to effects mediated through inhibition of cytochrome P450 oxidases.6,7 Because ketoconazole is considered a potent inhibitor of cytochrome P450 enzyme CYP3A12 and p-glycoprotein, its use can also cause adverse drug interactions.8–10
Ketoconazole toxicosis was first recognized in 1982 in a nonclinical experimental study involving Beagles. This study,11 reported in a monograph, investigated the usefulness and hazards of ketoconazole treatment for mycotic infections in dogs with escalated supratherapeutic dosages. Hyporexia, weight loss, variable emesis, and increases in serum ALT and ALP activities occurred when administered at 40 and 60 mg/kg/d (18 and 27 mg/lb/d), PO, for 1 year and 20 weeks, respectively.11 In addition to these signs, jaundice, gastritis, and lethal toxicosis were noted after 2 to 4 weeks of ketoconazole administration at 80 mg/kg/d (36 mg/lb/d), PO.
Adverse clinical reactions to ketoconazole (2.6 to 33.6 mg/kg/d [1.2 to 15.3 mg/lb/d], PO) have been retrospectively characterized in 15% (92/632) of dogs treated at 3 specialty dermatology practices.12 In that study,12 suspected adverse drug reactions were identified by use of the Naranjo adverse reaction probability scale.13 An adverse drug reaction was considered to have occurred if clinical signs, clinicopathologic markers, or hepatic histologic abnormalities developed after drug exposure and subsequently improved with drug discontinuation or dose reduction, in the absence of alternative plausible causes.13,14 Observed adverse effects included vomiting (7%), anorexia (5%), lethargy (2%), and diarrhea (1%) and, less frequently, cutaneous erythema, trembling, and weakness; some dogs had > 1 complication.12 Concurrent administration of cyclosporine or macrocyclic lactones was significantly associated with clinical toxicosis, and concurrent administration of ivermectin was associated with ataxia and trembling. Unfortunately, liver enzyme activities were infrequently monitored before (19/632 [3%]) or after (11/632 [1.7%]) ketoconazole treatment began.12 Among the monitored dogs, anicteric increases in serum liver enzyme activities were uncommonly encountered and no liver biopsies were performed. Although the nonclinical study11 involving ketoconazole administration to Beagles showed dose-related toxicosis at supratherapeutic dosages, findings of the retrospective study12 suggest that idiosyncratic toxicosis could develop in dogs at therapeutic dosages of the drug.
Ketoconazole-associated hepatotoxicosis in humans was characterized before the drug was withdrawn from the market.15–24 Increases in serum liver enzyme activities were associated with occasional nausea, inappetence, lethargy, and jaundice in humans with suspected DILI. Because clinical signs often resolved after drug discontinuation, liver biopsy was infrequently pursued. In more seriously ill humans who underwent liver biopsy, histopathologic features were characterized by variable injury patterns, including massive panlobular necrosis in patients with lethal DILI. In patients that recovered from the DILI, histopathologic liver features included mixed inflammatory infiltrates, variable canalicular cholestasis, and scattered single necrotic hepatocytes without consistent zonal localization. Aggregates of ceroid-lipofuscin–engorged macrophages (Kupffer cells) and marginating hepatocytes with cytosolic lipofuscin pigment affiliated with areas of parenchymal injury have also been reported.15,18–20,24 Similar injury patterns were observed in the nonclinical study11 of Beagles with ketoconazole-associated hepatotoxicosis. Occasionally, humans have developed chronic drug-initiated, immune-mediated hepatitis, and, in some situations, accidental drug reexposure–amplified liver injury consistent with immunosensitization or a hapten-mediated process.15,18,21 This DILI-related phenomenon is not unique to ketoconazole, and other drugs are also associated with immunoinjury patterns, as has been shown for diclofenac-treated dogs.25–29
Because no specific markers or tests exist that definitively confirm DILI, this diagnosis is typically designated as probable on exclusion of more common hepatobiliary disorders and consideration of environmental and therapeutic exposures and other risk factors.30 Such a diagnostic approach has also been used to characterize carprofen and lomustine hepatotoxicosis in dogs.31,32 The risk of ketoconazole-associated hepatotoxicosis in dogs has been anecdotally cited but, to the authors’ knowledge, never clinically documented. The purpose of the study reported here was to characterize the clinical and hepatic histopathologic features associated with probable ketoconazole-associated DILI in client-owned dogs treated with appropriate therapeutic dosages.
Materials and Methods
Case selection criteria
Dogs with probable ketoconazole-associated DILI were identified within a broad population of dogs with hepatotoxicosis suspected on histologic evaluation of liver biopsy samples submitted to the Animal Health Diagnostic Laboratory at Cornell University between January 2015 and December 2018. This broadly affected population was identified by use of proprietary software with a keyword relational database. Biopsy reports in the laboratory database were searched by use of search strings that included liver biopsy, canine, and one of the following terms: hepatotoxin, hepatotoxicity, drug-induced injury, drug-related hepatotoxicity or hepatotoxicosis, toxic hepatopathy, fulminant hepatic failure, panlobular necrosis, mushroom (Amanita), aflatoxin, mycotoxin, microcystin (blue-green algae and Rutland water toxicity), NSAID (carprofen,a meloxicam,b firocoxib,c deracoxib,d etodolac,e ibuprofen, and aspirin), anticonvulsant drugs (phenobarbital, zonisamide,g and levetiracetam), xylitol, jerky treats, sago palm (cycad), antimicrobials or antibiotics (tetracycline, doxycycline, ampicillin, cephalosporin, trimethoprim, sulfonamide, and rifampin), antimycotic drugs (ketoconazole, fluconazole, itraconazole, amphotericin B, and clotrimazole), chemotherapy drugs (lomustine,h vincristine, cyclophosphamide, chlorambucil,i streptozotocin,j and doxorubicin), and immunomodulatory drugs (azathioprine, mycophenolate, and cyclosporine).
For dogs to be considered for inclusion in the study, the classification of hepatotoxicosis required 2 or more histopathologic features consistent with hepatotoxicosis but distinct from commonly encountered hepatic abnormalities in dogs (eg, mechanical bile duct obstruction, gallbladder mucocele, suppurative cholangitis, cholangiohepatitis, chronic immune-mediated hepatitis, copper-associated hepatopathy, ductal plate malformation, or simple glycogen-type hepatocyte vacuolation). Histopathologic features considered consistent with hepatotoxicosis included hepatocyte necrosis (panlobular, lobular, centrilobular, or multifocal) or zonal parenchymal collapse associated with altered hepatocyte morphology (high mitotic activity, ring mitoses, anisocytosis, anisokaryosis, binucleation or polynucleation, megalohepatocyte transformation, hepatocyte disassociation, cytoskeletal injury, or hepatocyte lipid vacuolation) with accumulation of lipofuscin pigment in areas of injury (regional macrophages or cytosolic pigment accumulation in residual hepatocytes), canalicular cholestasis, or degenerative biliary epithelium. Suspected association with a drug or toxin was clarified by review of medical records and direct communication with attending clinicians. Dogs were deemed to have probable ketoconazole-associated DILI if they met these criteria and had a record of ketoconazole treatment that was temporally associated with illness and liver biopsy. Because this was a retrospective study, no approval was required from an institutional care and use committee or other oversight entity.
Medical record review
Data were extracted from medical records relevant to the time of ketoconazole administration and liver biopsy, including patient signalment (age, sex, body weight, and breed), reason for ketoconazole administration, ketoconazole dosage and treatment duration, concurrently administered medications or supplements, and other possible drug or toxin exposures. Case details were also obtained, including clinical signs, clinicopathologic test results (CBC, serum biochemical analysis, assays of pre- and postprandial serum total bile acids concentrations, urinalysis, and thyroid testing when available), results of abdominal ultrasonography performed by a board-certified veterinary radiologist or internist (assessment of hepatic size and hepatic parenchymal echogenicity; presence of hepatic nodules, smooth or irregular liver contour, or changes in architecture of other visceral and vascular structures; and evidence of abdominal effusion), liver histopathologic features, results of aerobic and anaerobic bacterial cultures of liver tissue or bile, and concurrent health problems. For the study, primary care veterinarians were contacted by telephone to obtain information regarding whether the liver injury had resolved (inferred from results of sequential liver enzyme assessments) and outcome.
Histologic evaluation of biopsy specimens
Liver biopsy specimens were fixed in neutral-buffered 10% formalin and sectioned at 5 μm. Sections were stained with H&E stain, reticulin stain (to evaluate hepatic cord width and organization, parenchymal collapse, and presence of proliferative or regenerative foci), Masson trichrome stain (to characterize presence, extent, and location of fibrillar collagen deposition), Prussian blue stain (to identify location and extent of iron sequestration in macrophages or hepatocytes and to differentiate ceroid-lipofuscin deposits from iron accumulation), and rhodanine stain (to identify hepatocyte cytosolic copper-protein aggregates). Biopsy specimens from 3 dogs were also stained with Ziehl-Neelsen stain to highlight ceroid-lipofuscin accumulation and Hall stain to inspect for canalicular cholestasis.
Specimens were evaluated for the presence and severity of distinct histopathologic features, including hepatocellular glycogen or microvesicular lipid vacuolation, ballooning degeneration, architectural remodeling (ie, parenchymal collapse or regenerative or proliferative foci), bile duct proliferation consistent with a ductular reaction, inflammatory infiltrates, fibrosis, portal or centrilobular bridging (with inflammatory infiltrates, ductular reaction, or fibrosis), hepatocyte necrosis, and pyogranulomatous inflammation. In addition, the accumulation and distribution of ceroid-lipofuscin pigment in macrophages or hepatocytes were characterized. Morphological features were assigned a qualitative grade ranging from 0 to 3 (0 = absent, 1 = mild, 2 = moderate, and 3 = severe) by 3 investigators (a veterinary internist with expertise in hepatic pathology [SAC] and 2 board-certified veterinary pathologists [SPM and JP]), with each evaluator unaware of specimen identity and working independently. Thereafter, scores from each evaluator were averaged together for each morphological feature and summed to generate a total histopathologic severity score.
In sections stained with rhodanine stain, hepatocyte copper accumulation was qualitatively graded by use of a scale from 0 (not detected) to 5 (numerous copper granules in ≥ 75% of cells), as described elsewhere.33 Quantification of copper was performed by digital scanning of rhodanine-stained liver sections or atomic absorption spectroscopy. Specimens with evidence of pyogranulomatous inflammation were additionally stained for infectious agents by use of Gram stain, modified Steiner, Gomori methenamine silver, and periodic acid–Schiff methods; the Ziehl-Neelsen and Fite-Faraco methods were used for detection of mycobacterial organisms. All stains applied to liver sections were routinely used for histologic evaluation at the laboratory.
Statistical analysis
Statistical analyses were completed by use of commercial software.k Clinical features (age, sex, body weight, breed, and clinical signs) were enumerated, and hepatic ultrasonographic findings were summated. Because hematologic and serum biochemical tests were completed in various laboratories with overlapping but different reference intervals, results were summarized as the percentage of dogs with results above, within, or below respective reference intervals. In accordance with methods in a prior study12 of adverse effects of ketoconazole in dogs, the fold increase in serum liver enzyme activities was also characterized relative to respective reference intervals.
Numeric data were largely nonparametric in distribution and are therefore reported as median (range). Pearson correlation coefficients (r) were computed to identify associations with serum liver enzyme activities for daily ketoconazole dose and treatment duration. Spearman rank correlation coefficients (ρ) were computed to identify correlations between total histopathologic severity score and dog age, daily ketoconazole dose, treatment duration, and serum liver enzyme activities. Values of P < 0.05 were considered significant.
Results
Animals
Of the 181 dogs that met the inclusion criteria for histopathologic features of hepatotoxicosis, the suspected causative agent was a drug in 87 (48%) dogs, a nondrug toxin in 58 (32%) dogs, and unidentified in 36 (20%) dogs. Among the 87 dogs with suspected DILI, liver biopsy specimens were received from 15 veterinary practices for 15 (17%) dogs with probable ketoconazole-associated DILI, and these 15 dogs were included in the study.
The biopsy specimens for the included dogs had been collected because of an unexplained increase in serum liver enzyme activities (n = 9) or suspected acute drug toxicosis (4) or during exploratory laparotomy for cholecystectomy (2). The 2 dogs that underwent cholecystectomy also had a long-term history of high serum liver enzyme activities that were monitored without liver biopsy. Exposure to ketoconazole was declared for 7 dogs before interpretation of the biopsy specimen and for 8 dogs during case discussion with the managing clinician after histopathologic features raised suspicion for DILI.
The 15 dogs included 9 spayed females and 6 castrated males with a median body weight of 13.0 kg (28.6 lb; range, 8.2 to 38.0 kg [18.0 to 83.6 lb]) and median age of 8.2 years (range, 5 to 15 years). Dogs were classified as Cocker Spaniel (n = 5), mixed breed (4), and Beagle, Brittany Spaniel, English Bulldog, Russell Terrier, Shiba Inu, and Shih Tzu (1 each).
Ketoconazole treatment
Ketoconazole had been prescribed to manage atopy complicated by Malassezia infection (n = 14) and for treatment of Malassezia-associated perivulvar dermatitis (1). Median daily ketoconazole dose was 7.8 mg/kg (3.5 mg/lb; range, 4.4 to 26.0 mg/kg [2.0 to 11.8 mg/lb]), PO. Relative to the recommended daily therapeutic dose of 10 mg/kg, 8 dogs received ≤ 10 mg/kg/d (4.5 mg/lb/d) and 7 dogs received 11 to 26 mg/kg/d (5 to 11.8 mg/lb/d). Treatment duration varied widely, ranging from 0.3 to 100 cumulative weeks. Six dogs received ketoconazole for ≤ 10 days, 2 dogs for 3 to 5 weeks, and 5 dogs intermittently (ie, for 3-to 6-week periods or 5 to 7 consecutive months) over a 2- to 5-year period.
Time to onset of clinical signs of hepatotoxicosis from the most recent ketoconazole administration ranged from 1 to 11 days for 7 dogs and 0.5 to 3 weeks for 6 dogs with cyclic or intermittent ketoconazole treatment and was 5 weeks for 1 dog. For 1 dog, the onset of clinical signs was insidious during 2 years of frequent 3-week treatment cycles. Time from ketoconazole discontinuation to liver biopsy was ≤ 1 week for 7 dogs; 2 weeks for 3 dogs; 4, 10, and 12 weeks for 1 dog each; and 20 weeks for 2 dogs.
Concurrent treatments
Thirteen dogs were fed hypoallergenic diets to control atopic dermatitis. Concurrent medications included macrocyclic lactones (n = 15), antimicrobials (8), antihistamines (6), oclacitinibl (4), carprofen (4), caninized monoclonal antibody against interleukin-31m (3), glucocorticoid medications (2), or cyclosporinen (2). Two dogs with hypothyroidism received daily thyroxine treatment, 1 dog with diabetes mellitus received daily insulin injections, 1 dog with urinary incontinence received daily phenylpropanolamine treatment, and 1 dog received phenobarbital treatment for epilepsy (with infrequent carprofen administration for arthritis). Three dogs were also treated with topical otic ketoconazole preparations for Malassezia-associated otitis externa. For 1 dog, ketoconazole administration was transitioned to fluconazole administration 5 days prior to liver biopsy. Another dog had previously been treated with fluconazole with no noted adverse clinical effect before subsequent ketoconazole treatment provoked clinical illness.
Clinical features
Common clinical features in dogs with suspected ketoconazole-associated DILI included signs of lethargy (n = 15) and anorexia (15), vomiting (8), jaundice (4), trembling (3), ataxia (1), and seizures (1). No abnormalities on CBC or discriminating features on urinalysis were recorded. One dog, which had received short-term carprofen treatment for postoperative pain associated with episioplasty, developed euglycemic glucosuria without cylindruria. Another dog had surface hemorrhages and prolonged bleeding at venipuncture sites that were associated with a 2.3-fold increase in prothrombin time and 1.2-fold increase in activated partial thromboplastin time. Serum biochemical abnormalities included mild to severe increases in liver enzyme activities. Median (range) fold increases in specific enzyme activities relative to reference intervals were as follows: ALT, 4.7 (1.3 to 21.2); AST, 3.3 (1.3 to 62.3); ALP, 2.9 (1.0 to 11.6); and GGT, 1.2 (1.0 to 14.1; Table 1). Of 7 dogs with hyperbilirubinemia, 3 had a mild 2-fold increase in serum total bilirubin concentration relative to the upper reference limit, whereas 4 dogs had jaundice with fold increases in serum total bilirubin concentration ranging from 8.0 to 14.8. Three of the 5 tested dogs without jaundice had a high serum total bile acids concentration (69 μmol/L, 116 μmol/L, and 219 μmol/L; reference interval, ≤ 25 μmol/L). For the dog with a total bile acids concentration of 219 μmol/L, mild hyperbilirubinemia corresponding to hepatocellular cholestasis was an obvious cause of this high value. Three dogs had clinicopathologic features consistent with failure of hepatic synthetic functions (ie, hypocholesterolemia and decreasing serum albumin and BUN concentrations as well as prolonged coagulation times for 1 dog). Four dogs developed ascitic abdominal effusion.
Summary data for selected serum biochemical analytes measured at the time of liver biopsy specimen collection for 15 dogs with probable ketoconazole-associated DILI.
| Analyte | No. of dogs | No. (%) with results within reference range | No. (%) with results lower than reference rage | No. (%) with results higher than reference range |
|---|---|---|---|---|
| ALT | 15 | 0 (0) | NA | 15 (100) |
| AST | 11 | 0 (0) | NA | 11 (100) |
| ALP | 15 | 2 (13) | NA | 13 (87) |
| GGT | 13 | 5 (38) | NA | 8 (62) |
| Total bilirubin | 15 | 8 (53) | NA | 7 (47) |
| Cholesterol | 13 | 6 (46) | 4 (31) | 3 (23) |
| Albumin | 15 | 7 (47) | 8 (53) | 0 (0) |
| Globulin | 14 | 13 (93) | 0 (0) | 1 (7) |
| BUN | 14 | 11 (79) | 2 (14) | 1 (7) |
| Creatinine | 14 | 14 (100) | 0 (0) | 0 (0) |
NA = Not applicable.
No reference ranges are provided because dogs were patients at 15 veterinary practices with different analyzers and associated reference ranges. Instead, dogs were classified as whether values were within or outside respective reference ranges.
Abdominal ultrasonography, which was performed for 14 dogs (the 1 dog with acute hepatic failure did not undergo abdominal ultrasonography), revealed a subjectively normal liver size in 13 dogs, microhepatia in 1 dog, and abdominal effusion in 4 dogs. The dog with microhepatia had heteroechoic hepatic parenchyma and an irregular liver contour. Five dogs had poorly defined parenchymal nodules that appeared hyperechoic (n = 3) or had hyperechoic margins surrounding heteroechoic parenchyma (2). Other visceral and vascular structures and mesenteric lymph nodes were considered unremarkable.
Liver biopsy specimens
Liver biopsy specimens (obtained a median of 1 week [range, 1 to 20 weeks] after ketoconazole discontinuation) were collected by laparoscopic (cup forceps; n = 9), surgical wedge (4), and 14-gauge ultrasound-guided biopsy needle (2) methods. One dog underwent an initial surgical wedge biopsy followed by a needle biopsy 6 months after ketoconazole discontinuation to investigate the cause of sustained high serum liver enzyme activities.
For the specimen from 1 dog, no rhodanine staining was completed because insufficient biopsy tissue was available following submission for copper quantification. Copper quantification in biopsy specimens was performed for all dogs by digital scanning of rhodanine-stained liver sections (n = 13) or atomic absorption spectroscopy (2). Three specimens in which pyogranulomatous inflammation was detected were additionally stained for identification of infectious agents. Results of aerobic and anaerobic bacterial cultures of liver specimens, bile samples, or both were negative for 13 dogs; the remaining 2 dogs had no bacterial culture performed.
Histopathologic features of liver biopsy specimens included portal tract inflammation for 12 dogs (classified as mild [n = 7], moderate [3], or severe [2]), glycogen-type vacuolar hepatopathy for 11 dogs (mild [3] or moderate [8]), ductular reaction for 11 dogs (mild [4], moderate [5], or severe [2]), and rare canalicular cholestasis for 6 dogs. Centrilobular inflammatory infiltrates were observed for 11 dogs (classified as mild [n = 4] or moderate [7]), centrilobular fibrosis for 11 dogs (mild [5], moderate [4], or severe [2]), portal tract fibrosis for 9 dogs (mild [5] or moderate [4]), portal tract bridging for 7 dogs (mild [1] or moderate [6]), and centrilobular bridging for 7 dogs (mild [1], moderate [4], or severe [2]). Moderate to severe central and midzonal hepatocyte necrosis was identified for 3 dogs, with panlobular necrosis for 1 dog. Regenerative nodules were identified for 5 dogs (classified as mild [n = 1], moderate [2], or severe [2]).
Ceroid-lipofuscin–engorged macrophages in liver biopsy specimens were identified for all 15 dogs (classified as mild [n = 5], moderate [6], or severe [4]), with a periportal distribution for 11 dogs (mild [5], moderate [4], or severe [2]) and centrilobular distribution for 13 dogs (moderate [10] or severe [3]; Figure 1). The specimens from 3 dogs had evidence of centrilobular parenchymal collapse that occasionally obscured the profiles of hepatic veins, consistent with a sinusoidal occlusion syndrome. For all dogs, the nature of the inflammatory infiltrates was mixed and included lymphocytes, macrophages, and fewer plasma cells and neutrophils. Lymphoplasmacytic lobular hepatitis was characterized in specimens from 2 dogs and multifocal effacing pyogranulomatous hepatitis in specimens from 3 dogs. Two dogs with pyogranulomatous inflammation had negative PCR assay results for mycobacteria, whereas the other dog was not tested. Results of infectious agent staining and bacterial culture were negative for each dog with pyogranulomatous inflammation. For one of these dogs, results of fluorescent in situ hybridization detection of eubacterial agents were also negative.
Photomicrographs of representative liver sections from dogs with probable ketoconazole-associated DILI. A—Dense aggregates of ceroid-lipofuscin–laden macrophages (golden-brown pigment) align along and bridge between portal tracts (PT). Light infiltrates of mixed inflammatory cells integrate at the interface between pigment-laden macrophages, hepatocytes, and bile duct profiles. Hepatocytes have variable mild glycogen-type vacuolation (asterisk) with rare ballooning degeneration (arrow). B—A fibrotic portal tract with periportal and intraportal lymphocytic infiltrates is visible, with scattered ceroid-lipofuscin–laden macrophages (golden-brown pigment). A ductular reaction (arrows), foci of aggregated lymphocytes, and pigment-laden macrophages are scattered within adjacent hepatic parenchyma where hepatocytes have glycogen-type vacuolation. C—Severe centrilobular collapse (arrows) causing regional congestion is visible, with numerous aggregates of ceroid-lipofuscin–laden macrophages (golden-brown pigment) obscuring the hepatic vein profile, consistent with development of sinusoidal occlusion syndrome. Parenchymal collapse has resulted in close proximity of portal and centrilobular regions; a marked ductular reaction (asterisks) is evident to the left of the hepatic venule. D—Centrilobular lymphoplasmacytic and histiocytic inflammation is visible in this specimen from a dog with chronic hepatitis subsequent to ketoconazole administration. Notice the scattered ceroid-lipofuscin–laden macrophages (golden-brown pigment), occasional necrotic hepatocytes (hyperchromatic hepatocytes; arrows), and cytosolic lipofuscin granules in hepatocytes (asterisk in inset). H&E stain; bar = 200 μm in panels A, B, and D (60 μm in inset) and 300 μm in panel C.
Citation: Journal of the American Veterinary Medical Association 256, 11; 10.2460/javma.256.11.1245
The Ziehl-Neelsen method stained cytosolic ceroid-lipofuscin granules an intense maroon color, consistent with glycoprotein content. Results of Prussian blue staining were negative for iron aggregates within pigment-engorged macrophages for 9 dogs and confirmed small amounts of iron colocalized with ceroid-lipofuscin pigment in macrophages and occasional hepatocytes in 6 dogs. Results of Hall staining confirmed the presence of rare canalicular bile plugs in specimens from the 3 dogs for which this stain was used.
Rhodanine staining revealed cytosolic copper-protein aggregates in the hepatocytes of 11 of 14 dogs; assigned qualitative copper scores were 0 (n = 3), 1 (4), 2 (2), 3 (1), 4 (1), and 5 (3). Seven dogs had liver copper concentrations ≤ 415 μg/g of DWL, whereas the remaining 8 dogs had high liver copper concentrations (ie, 627, 743, 936, 964, 1,520, 2,124, 2,161, and 9,840 μg/g of DWL; reference interval, ≤ 400 μg/g of DWL). In specimens from 5 dogs, hepatocytes with cytosolic copper granules closely abutted aggregates of ceroid-lipofuscin–laden macrophages that marked regions of DILI, suggesting copper-augmented parenchymal injury. However, in 4 dogs with markedly high liver copper concentrations, abundant aggregates of ceroid-lipofuscin–laden macrophages were not colocalized with copper-mediated injury, suggesting a separate 2-hit injury mechanism in those dogs.
Correlation analysis
Significant positive correlations were identified between ketoconazole dose and serum activities of the liver enzymes ALT (r = 0.71; P = 0.003), AST (r = 0.68; P = 0.02), ALP (r = 0.67; P = 0.006), and GGT (r = 0.60; P = 0.03). However, no significant (P > 0.10) correlations were found between duration of ketoconazole treatment and serum liver enzyme activities or between histopathologic severity score and dog age, ketoconazole dose, duration of ketoconazole treatment, or serum liver enzyme activities. There was also no significant (P = 0.06) correlation between histopathologic severity score and liver copper concentration.
Outcome
Seven dogs had decreases in serum liver enzyme activities after ketoconazole discontinuation. Five dogs remained alive at 4, 56, 64, 116, and 150 weeks after illness onset (the time of last follow-up). Of the 10 dogs reported to have died at last follow-up, survival time after onset of ketoconazole-associated illness ranged from 0.5 to 165 weeks. Death was associated with liver injury in 7 of the 10 nonsurvivors. Of these 7 dogs, 3 dogs with fulminant hepatic failure died within 1 week after diagnosis of probable ketoconazole-associated DILI, and 1 dog was euthanized because of surgical complications associated with liver biopsy and a projected poor prognosis (1.9 weeks after liver biopsy). The remaining 3 dogs died of progressive chronic hepatitis 36 to 54 weeks after liver biopsy. Of those 3 dogs, the hepatitis was classified as pyogranulomatous (n = 2) or lymphoplasmacytic (1). One dog with pyogranulomatous hepatitis was euthanized after it developed a portal vein thrombus.
Of the 4 dogs that died within 2 weeks after liver biopsy (ie, the 3 dogs with fulminant hepatic failure and the dog with biopsy-related complications), 1 (dog 1) had received long-term phenobarbital treatment for idiopathic epilepsy plus intermittent carprofen treatment for arthritis. This dog had a liver copper concentration of 2,124 μg/g of DWL. The second dog (dog 2) received intermittent, long-term carprofen treatment for arthritis and had a liver copper concentration of 2,161 μg/g of DWL. The third dog (dog 3) had received carprofen and ketoconazole concurrently for severe paw pain from pododermatitis, and the fourth dog (dog 4) had microhepatia, severe chronic hepatic remodeling, and ascitic abdominal effusion associated with a liver copper concentration of 9,840 μg/g of DWL. Dog 1 had histopathologic evidence of chronic liver injury (ie, diffuse regenerative nodules) and developed fulminant hepatic failure after only 3 days of ketoconazole treatment. Dog 2 became ill during 3 weeks of ketoconazole administration and acutely decompensated when carprofen was added as a postoperative analgesic. Dog 3 with histopathologic evidence of chronic liver injury developed panlobular hepatic necrosis after carprofen administration. Dog 4 with histopathologic evidence of chronic severe liver injury developed clinical illness after 3 weeks of ketoconazole treatment, 20 weeks before liver biopsy.
Overall, of the 4 dogs with ascitic abdominal effusion, 2 died of acute hepatic decompensation and 1 was euthanized (dog 4). The remaining dog recovered, and the abdominal effusion resolved. Of the 3 dogs that died of progressive chronic hepatitis, liver injury was unabated (as indicated via clinicopathologic monitoring) despite immunomodulation with glucocorticoid treatment (n = 2) or cyclosporine (1) and administration of ursodeoxycholic acid and bioavailable S-adenosylmethionine.o
Death was deemed unrelated to the ketoconazole-associated DILI in 3 dogs. These dogs included one that was euthanized after alimentary lymphoma was diagnosed, one that died of systemic sepsis secondary to severe chronic bacterial rhinosinusitis, and another with pyogranulomatous hepatitis that achieved apparent remission with long-term prednisone administration but was euthanized 2.5 years after diagnosis of probable ketoconazole-associated DILI because of continued signs of arthritic pain, steroid-associated adverse effects, and perceived poor life quality. In each of these cases, the dog recovered from ketoconazole-associated DILI (as suggested by improvements in serum liver enzyme activities on sequential assessments) before it developed non–liver-related fatal illness or was euthanized.
Discussion
No single test exists for the definitive diagnosis of DILI, and the associated injury patterns may resemble other primary liver disorders.26,34 Accordingly, a diagnosis of DILI is made by exclusion of other causes, with the most credible evidence being the coordinated finding of circumstantial drug exposure, clinical illness, and histologic evidence of liver injury. However, clinical signs of DILI in humans may have a delayed onset ranging from weeks to months and sometimes years after initial drug exposure, and not all patients with DILI experience resolution of liver injury after drug discontinuation.28,34,35 Indeed, progression of DILI to chronic hepatitis has been estimated to occur in as many as 10% of human cases and is a phenomenon rarely considered in veterinary medicine.34
The clinical signs observed in dogs with probable ketoconazole-associated DILI in the present study were similar to those reported for dogs with drug-related clinical illness in the previous retrospective study12 of dogs receiving ketoconazole treatment for Malassezia-complicated atopy (ie, lethargy, anorexia, vomiting, and rare trembling or ataxia). As reported for humans,36 clinical illness associated with such DILIs in dogs typically resolves with drug discontinuation.12 However, the frequency of ketoconazole-related increases in liver enzyme activities is unknown for dogs because serum biochemical monitoring is rarely undertaken during a conventional 4-week dermatologic treatment interval.12 Ketoconazole-associated DILI in humans is usually recognized within 2 to 6 months after initial drug exposure.18–20,24,36 A variable but generally shorter time frame was identified for the dogs of the present study. Because the risk for ketoconazole-associated hepatotoxicosis is marginalized in the veterinary literature, several clinicians who submitted liver biopsy specimens from these dogs had not prioritized this prospect.
Five of the 15 dogs in the study reported here were Cocker Spaniels. Although this breed was also common in the previously mentioned large dermatologic investigation of dogs treated with ketoconazole, no Cocker Spaniel in that study12 was reported to have ketoconazole-associated clinical illness. The Cocker Spaniels of the present study had no features consistent with a breed-specific hepatopathy, and the study design yielded no evidence that this particular breed was at greater risk than other breeds for ketoconazole toxicosis. Instead, we surmise that the relatively high number of Cocker Spaniels included in our data set reflected clinicians’ concerns that those dogs had Cocker Spaniel hepatopathy (rather than DILI).37,38
In the present study, dogs with suspected ketoconazole-associated DILI had variable increases in serum liver enzyme activities and occasional hyperbilirubinemia, similar to features characterized in humans.18,22 Although the degree of hyperbilirubinemia was sufficient to cause jaundice in only 4 dogs, high serum total bile acids concentrations were identified in 3 of the 5 dogs without jaundice for which this analyte was measured. The high bile acids values most plausibly reflected DILI-related intrahepatic cholestasis, although in 2 small-breed dogs, they might have also reflected microvascular dysplasia.39 Interpretation of high serum ALP activity was confounded by the enzyme induction phenomenon associated with coexistent systemic inflammation (moderate to severe dermatitis, otitis, and pododermatitis), occasional coadministration of glucocorticoid drugs (for control of atopy), and in 1 dog, long-term phenobarbital administration. Hepatocellular cholestasis undoubtedly contributed to high serum ALP and GGT activities in some dogs. High serum activities of ALT and AST reflected hepatic inflammation, hepatocyte necrosis, or reversible alterations in hepatocyte or mitochondrial membrane integrity. In some cases, hepatocyte injury was probably also potentiated by pathological copper accumulation and concurrent administration of NSAIDs or phenobarbital.
Few reports15–24 characterize histopathologic features of ketoconazole-associated DILI in humans because drug administration is simply discontinued and recovery is presumed on the basis of serum biochemical findings. A similar management strategy contributes to the dearth of information for dogs with possible ketoconazole-associated DILI. Abundant ceroid-lipofuscin–engorged macrophages aligning areas of parenchymal injury were identified in every dog of the present study, and this feature has also been reported for humans with ketoconazole-associated DILI15,18–20,24 and Beagles receiving supratherapeutic ketoconazole doses.11 Aggregates of pigment-laden macrophages (recruited or resident macrophages [Kupffer cells]) reflect phagocytosis of necrotic cellular debris (including oxidized membrane fragments).40,41 Recognition of these sentinel aggregates raised the suspicion of toxin- or drug-related liver injury in each dog of the present study, prompting investigation to identify a probable cause. However, it is important to clarify that aggregates of ceroid-lipofuscin–laden macrophages are not pathognomonic for ketoconazole-associated DILI because they are nonspecific remnants of historic parenchymal injury attributable to multiple causes (eg, drugs or toxins causing oxidative injury and necrosis).27,42–48 Similar to histopathologic findings in humans with ketoconazole-associated and other types of DILIs,27,42–48 no consistent lobular injury pattern was noted in the dogs. Acute necrotizing injury in 3 dogs had a central and midzonal distribution. A single dog with histologic evidence of prior liver injury developed fatal massive panlobular necrosis, as has been described in some humans with lethal ketoconazole-associated DILI.18,20 Otherwise, the most consistent histopathologic features included sentinel aggregates of ceroid-lipofuscin–laden macrophages and variable severities of mixed inflammatory infiltrates, patchy bile duct proliferation (ductular reaction), and individual hepatocyte necrosis, as similarly described in humans with ketoconazole-associated DILI.18,20,27
Chronic progressive lymphoplasmacytic hepatitis and pyogranulomatous hepatitis in 4 dogs of the present study raised the possibility of a drug-initiated hepatitis. Although the possibility remains that response patterns consistent with hepatitis preceded ketoconazole exposure, we posit that this was unlikely because serum liver enzyme activities were within reference intervals in 3 of these dogs at 1 to 2.5 weeks before ketoconazole treatment began. The fourth dog, which lacked pretreatment laboratory assessments, was clinically well days before ketoconazole treatment began but developed signs of severe illness within 72 hours after treatment began. Development of chronic hepatitis subsequent to DILI is not unprecedented. In humans, severe progressive hepatitis with autoimmune phenotypes or chronic granulomatous injury has been attributed to DILI despite cessation of treatment with the responsible drug.18–20,27,28,34–36,49 It is speculated that acquired autoimmune or granulomatous inflammation reflects an awakened latent autoimmunity or development of neoepitopes that trigger adaptive immune responses, leading to chronic sustained inflammation.27,36,48–53 Research in dogs with diclofenac-associated DILI has elucidated relevant pathophysiologic mechanisms that include a diverse interplay of gene transcriptional responses and metabolic pathways integrated with innate and adaptive immune reactions.25 Importantly, these complex interactions may culminate in histopathologic features naïvely classified as idiosyncratic when they actually represent acquired immunologic hypersensitivity.25,27,36
A prolonged gradual decrease in serum liver enzyme activity was documented in 7 dogs of the present study after ketoconazole discontinuation. In humans, high serum ALT and AST activities persist for a mean of 7 weeks (range, 1 to 16 weeks), and sometimes as long as 6 months, after ketoconazole discontinuation.17,18,22,34 Even so, some humans with ketoconazole-associated DILI may also have a delayed (weeks to months) onset of high serum liver enzyme activities after ketoconazole treatment begins, despite drug discontinuation.34 Thus, prolonged hepatocyte membrane injury (including mitochondrial injury) and inflammation may persist regardless of whether chronic hepatitis follows.18,34 The pathomechanism of this phenomenon likely involves the impact of residual bioactivated drug metabolites (ie, bound to macromolecules, mitochondrial enzymes, or structural proteins).
In humans, no antecedent or combination drug treatments have been implicated as risk factors for ketoconazole-associated DILI.18–20,24 Results of the retrospective study12 of ketoconazole-related clinical illness in dogs with dermatologic disorders suggest a possible association with concurrent cyclosporine or macrocyclic lactone administration. The dogs of the present study were exposed to a spectrum of medications, most commonly including macrocyclic lactones, NSAIDs, and antimicrobials. Given the design and small sample size of our study, one cannot conclude that macrocyclic lactone treatment increased the risk of ketoconazole-associated DILI because these drugs are popularly used to prevent heartworm infection. However, concurrent treatment with carprofen in 4 dogs and phenobarbital in 1 dog as well as concurrent hepatocyte copper-accumulation in 8 dogs may have potentiated the observed centrilobular parenchymal injury. Histopathologic changes attributable to carprofeninduced injury and copper hepatotoxicosis impact centrilobular hepatocytes by oxidative membrane injury. Phenobarbital accelerates certain cytochrome P450 oxidases, which can influence formation and accumulation of bioactivated or toxic metabolites from numerous xenobiotics. Unfortunately, the limited number of dogs with probable ketoconazole-associated DILI precluded rigorous evaluation of risk factors that might have influenced or augmented drug toxicosis.
No evidence exists of a dose- or duration-related risk for ketoconazole-associated DILI in humans. Rather, ketoconazole-associated DILI is considered an idiosyncratic hepatotoxicosis. An acquired hypersensitivity response has been suspected in some individuals with high serum liver enzyme activities (and surmised liver injury) after accidental drug reexposure.15,21 As previously mentioned, research in Beagles has shown hepatotoxicosis with supratherapeutic ketoconazole administration.11 Although no dog in the case series reported here received ketoconazole at a supratherapeutic dose, a significant linear association between drug dose and serum liver enzyme activities was demonstrated. Nevertheless, in humans with DILI, high serum liver enzyme activities do not reliably predict histopathologic injury,36 consistent with the lack of an observed correlation between ketoconazole dose and histopathologic severity score in the present study.36 Also, simply deducing that ketoconazole-associated DILI is idiosyncratic may be erroneous given that adverse reactions may reflect pharmacogenetic differences between individuals that influence drug metabolism or elimination or age-related changes impacting drug distribution or metabolism (ie, higher realized drug dose owing to body condition status or an altered rate of metabolite formation or elimination).36,54–57 The propensity for DILI may also be influenced by drug lipophilicity and cytochrome P450 activity, both of which are relevant to ketoconazole.54,58–62 Ketoconazole is highly lipophilic, has widely variable bioavailability among oral formulations (evaluated in dogs63), and is metabolized by and influences cytochrome P450 oxidase activity. In the present study, formulations administered to individual dogs were not specified, and we could not determine whether some dogs received greater-than-expected doses as a result of different bioavailabilities. The risk of DILI may be related to variability (expression and function) in cytochrome P450 activity among dogs and the influence of age, breed, or disease-related metabolic adaptations as well as the impact of concurrently administered xenobiotics on cytochrome P450 enzymes.55–57,59,64 Although no significant correlations were identified between severity of histopathologic liver injury and patient age, drug dose, concurrent treatments, or serum liver enzyme activities, the number of dogs evaluated was small and liver biopsy specimens from dogs receiving ketoconazole without adverse effects were not available for comparison.
In humans and rodents, ketoconazole undergoes efficient enteric absorption, hepatic uptake, and metabolism.65 Although dogs metabolize ketoconazole to numerous products, no specific investigation of hepatotoxic moieties has been reported.11 Metabolism in rats involves various enzymes, including flavin-containing monooxygenases and arylacetamide deacetylase.66–70 In vitro studies68–71 of rat and human hepatocytes and microsomes have characterized a reactive drug adduct (DAK) that imparts oxidative injury and provokes hepatocyte ALT release in a dose- and concentration-related response. Cellular injury is affiliated with covalent binding of injurious ketoconazole metabolites with macromolecules and associated with hepatocyte glutathione depletion.72,73 Intracellular concentrations of ketoconazole metabolites achieved through in vitro studies have been deemed relevant to therapeutic drug exposures. Experimentally, DAK accumulation depletes hepatocyte and mitochondrial ATP, causes oxidative membrane injury, provokes mitochondrial-initiated hepatocyte apoptosis, and damages mitochondrial DNA. The damage to mitochondrial DNA is a serious pathomechanism of hepatotoxicosis because mitochondria are largely devoid of nucleoprotein repair mechanisms.69,72,74 Consequently, ketoconazole administration at clinical doses can be deduced to functionally influence mitochondrial enzyme complexes essential to the electron transport chain.72 In vitro findings also suggest the possibility of an increased risk of ketoconazole-associated DILI in patients with antecedent liver injury owing to preexistent hepatocyte ATP depletion and DAK cytotoxic effects accentuated by mitochondrial injury. Although this supposition is contrary to conclusions made in observational reports18–20,24,70 of ketoconazole-associated DILI in humans, it corresponds to the observations of the present study whereby hepatocyte injury may have been provoked by concurrent copper accumulation, NSAID administration, or phenobarbital treatment.
Eight dogs in the study reported here had pathological hepatocyte copper accumulation. Because copper functions as a transition metal facilitating formation of reactive superoxide and hydroxyl radicals (ie, Haber-Weiss and Fenton reactions), it can escalate oxidative injury in hepatocytes damaged by other pathomechanisms.75 Indeed, 2 dogs with liver copper concentrations > 2,000 μg/g of DWL died suddenly after ketoconazole treatment began.
The frequency of ketoconazole-associated DILI in dogs remains unknown given that most ketoconazole-treated dogs are not clinicopathologically monitored. Nevertheless, in the source population of dogs with histopathologic evidence of hepatotoxicosis, ketoconazole treatment accounted for 17% of suspected DILIs, which was not a trivial prevalence. Although a spectrum of liver injuries is associated with ketoconazole-associated DILI in dogs and humans, the sentinel injury pattern described herein (ie, aggregates of ceroid-lipofuscin–laden macrophages and hepatocytes) suggested possible DILI or toxin exposure, warranting more thorough investigation of historical events and xenobiotic or toxin exposures. Progressive hepatitis despite drug discontinuation was documented for 3 of 4 dogs, as reported in some humans.18 Although the gold standard test for diagnosis of DILI requires reexposure to the drug with demonstration of recurrent or recrudescent liver injury, that process imposes unethical hazards and rarely occurs.27,36 Intentional reexposure of dogs with suspected ketoconazole-associated DILI was not undertaken in the present case series. Thus, the diagnosis of probable ketoconazole-associated DILI was supported by documentation of circumstantial drug exposure, biochemical evidence of liver injury, and DILI-compatible histopathologic features, combined with meticulous review of historical environmental, drug, and toxin exposures as well as the absence of alternative reasonable underlying causes.
Given the findings of the present study, we suggest that liver enzyme monitoring be performed before ketoconazole treatment begins and then at 1 week and 1 month during initial or cyclic treatment, as recommended for humans who are prescribed ketoconazole.1 Intermittent biochemical assessments should be considered for dogs with sporadic or seasonal treatment. These strategies are proposed to recognize liver injury that would indicate the need for drug discontinuation. However, such strategies are imperfect because some humans with short-term ketoconazole exposure develop only biochemical evidence of hepatic injury days to weeks after ketoconazole discontinuation, ostensibly owing to the prolonged impact of DAK or other metabolites.18,76 Because ketoconazole-associated DILI is linked with oxidative injury, it seems prudent to add a glutathione donor (eg, bioavailable S-adenosylmethionine) to the regimen of ketoconazole-treated dogs, to administer N-acetylcysteine to dogs hospitalized for suspected ketoconazole-associated DILI, and to avoid concurrent NSAID administration. However, whether concurrent administration of a glutathione donor or membranoprotective agent such as S-adenosylmethionine can protect against ketoconazole-associated DILI is unknown. Ursodeoxycholic acid has been advocated for humans with various types of DILIs because of its choleretic influence that may hasten drug or metabolite elimination and its hepatoprotective effects.77 Glucocorticoid administration has also been advocated for humans with severe ketoconazole-associated DILI, in whom a beneficial effect on clinical recovery has been described.22 In the present study, long-term glucocorticoid drug administration to 1 dog with chronic pyogranulomatous hepatitis seemingly resulted in clinical remission during a 2.5-year interval.22 Nevertheless, despite interventions including glucocorticoid and other immunomodulatory drugs, ursodeoxycholic acid, and bioavailable S-adenosylmethionine, unremitting severe progressive hepatitis led to eventual death in 3 of 15 dogs (2 with pyogranulomatous hepatitis and 1 with lymphoplasmacytic hepatitis) in our study.
ABBREVIATIONS
| ALP | Alkaline phosphatase |
| ALT | Alanine aminotransferase |
| AST | Aspartate aminotransferase |
| DAK | N-deacetyl-ketoconazole |
| DILI | Drug-induced liver injury |
| DWL | Dry-weight liver |
| GGT | γ-Glutamyltransferase |
Footnotes
Rimadyl, Zoetis Inc, Kalamazoo, Mich.
Metacam, Boehringer Ingelheim, Vetmedica Inc, St Joseph, Mo.
Previcox, Merial Inc, Duluth, Ga.
Deramaxx, Elanco Animal Health, Greenfield, Ind.
EtoGesic, Fort Dodge Animal Health, Overland Park, Kan.
Advil, Pfizer Inc, New York, NY.
Zonegran, Dainippon Pharmaceutical Co Ltd, Osaka, Japan.
CCNU, NextSource Biotechnology LLC, Miami, Fla.
Leukeran, Aspen Global Inc, Grand Baie, Mauritius.
Zanosar, Teva Pharmaceuticals, Parsippany, NJ.
Statistix, version 9, Analytical Software, Tallahassee, Fla.
Apoquel, Zoetis Inc, Kalamazoo, Mich.
Cytopoint, Zoetis Inc, Kalamazoo, Mich.
Atopica, Elanco Animal Health, Greenfield, Ind.
Denosyl, Nutramax Laboratories Veterinary Science, Edgewood, Md.
References
1. Gupta AK, Lyons DCA. The rise and fall of ketoconazole. J Cutan Med Surg 2015;19:352–357.
2. Moriello KA. Ketoconazole: clinical pharmacology and therapeutic recommendations. J Am Vet Med Assoc 1986;188:303–306.
3. Negre A, Bensignor E, Guillot J. Evidence-based veterinary dermatology: a systematic review of interventions for malassezia dermatitis in dogs. Vet Dermatol 2009;20:1–12.
4. Rowan-Kelly B, Ferrante A, Thong YH. Modification of polymorphonuclear leucocyte function by imidazoles. Int J Immunopharmacol 1984;6:389–393.
5. Van Cutsem J, Van Gerven F, Cauwenbergh G, et al. The antiinflammatory effects of ketoconazole. A comparative study with hydrocortisone acetate in a model using living and killed Staphylococcus aureus on the skin of guinea pigs. J Am Acad Dermatol 1991;25:257–261.
6. Peterson ME. Medical treatment of canine pituitary-dependent hyperadrenocorticism (Cushing's disease). Vet Clin North Am Small Anim Pract 2001;31:1005–1014.
7. Dahlinger J, Gregory C, Bea J. Effect of ketoconazole on cyclosporine dose in healthy dogs. Vet Surg 1998;27:64–68.
8. Lu P, Singh SB, Carr BA, et al. Selective inhibition of dog hepatic CYP2B11 and CYP3A12. J Pharmacol Exp Ther 2005;313:518–528.
9. Kuroha M, Kuze Y, Shimoda M, et al. In vitro characterization of the inhibitory effects of ketoconazole on metabolic activities of cytochrome P-450 in canine hepatic microsomes. Am J Vet Res 2002;63:900–905.
10. Foy DS, Trepanier LA. Antifungal treatment of small animal veterinary patients. Vet Clin North Am Small Anim 2010;40:1171–1188.
11. Heel RC. Toxicology and safety studies. In: Levine HB, ed. Ketoconazole in the management of fungal disease. New York: ADIS Press, 1982;74–76.
12. Mayer UK, Glos K, Schmid M, et al. Adverse effects of ketoconazole in dogs—a retrospective study (Erratum published in Vet Dermatol 2008;19:319). Vet Dermatol 2008;19:199–208.
13. Alderman C, Naranjo ADR. Probability scale. RGH Pharacym E-Bulletin 2006;20:11.
14. Arimone Y, Begaud B, Miremont-Salame G, et al. A new method for assessing drug causation provided agreement with experts’ judgment. J Clin Epidemiol 2006;59:308–314.
15. Heiberg JK, Svejgaard E. Toxic hepatitis during ketoconazole treatment (lett). Br Med J (Clin Res Ed) 1981;283:825–826.
16. Janssen PAJ, Cauwenbergh G, Symoens J. Hepatic reactions during ketoconazole treatment: a 1 year update. Beerse, Belgium: Janssen Pharmaceutica, 1983.
17. Janssen PAJ, Symoens J. Hepatic reactions during ketoconazole treatment. Am J Med 1983;74:80–85.
18. Lewis JH, Zimmerman HJ, Benson GD, et al. Hepatic injury associated with ketoconazole therapy: analysis of 33 cases. Gastroenterology 1984; 86:503–513.
19. Stricker BHCH, Blok APR, Bronkhorst FB, et al. Ketoconazole-associated hepatic injury a clinicopathological study of 55 cases. J Hepatol 1986;3:399–406.
20. Lake-Bakaar G, Scheueer PJ, Sherlock S. Hepatic reactions associated with ketoconazole in the United Kingdom. Br Med J (Clin Res Ed) 1987;294:419–422.
21. van Parys G, Evenepoel C, Van Damme B, et al. Ketoconazole-induced hepatitis: a case with a definite cause-effect relationship. Liver 1987;7:27–30.
22. Benson GD, Anderson PK, Combes B, et al. Prolonged jaundice following ketoconazole-induced hepatic injury. Dig Dis Sci 1988;33:240–246.
23. Daneshmend TK, Warnock DW. Clinical pharmacokinetics of ketoconazole. Clin Pharmacokinet 1988;14:13–34.
24. Findor JA, Sorda JA, Igartua EB, et al. Ketoconazole-induced liver damage. Medicina (B Aires) 1998;58:277–281.
25. Selvaraj S, Oh JH, Spanel R, et al. The pathogenesis of diclofenac induced immunoallergic hepatitis in a canine model of liver injury. Oncotarget 2017;8:107763–107824.
26. Bénichou C. Criteria of drug-induced liver disorders. Report of an international consensus meeting. J Hepatol 1990;11:272–276.
27. deLemos AS, Foureau DM, Jacobs C, et al. Drug-induced liver injury with autoimmune features. Semin Liver Dis 2014;34:194–204.
28. Stine JG, Chalasani N. Chronic liver injury induced by drugs: a systematic review. Liver Int 2015;35:2343–2353.
29. Licata A. Adverse drug reactions and organ damage: the liver. Eur J Intern Med 2016;28:9–16.
30. Chalasani N, Fontana RJ, Bonkovsky HL, et al. Causes, clinical features, and outcomes from a prospective study of drug-induced liver injury in the United States. Gastroenterology 2008;135:1924–1934.
31. Kristal O, Rassnick KM, Gliatto JM, et al. Hepatotoxicity associated with CCNU (lomustine) chemotherapy in dogs. J Vet Intern Med 2004;18:75–80.
32. MacPhail CM, Lappin MR, Meyer DJ, et al. Hepatocellular toxicosis associated with administration of carprofen in 21 dogs. J Am Vet Med Assoc 1998;212:1895–1901.
33. Center SA, McDonough SP, Bogdanovic L. Digital image analysis of rhodanine-stained liver biopsy specimens for calculation of hepatic copper concentrations in dogs. Am J Vet Res 2013;74:1474–1480.
34. Ramachandran R, Kakar S. Histological patterns in drug-induced liver disease. J Clin Pathol 2009;62:481–492.
35. Kleiner DE, Chalasani NP, Lee WM, et al and the drug-induced liver injury network (DILIN). Hepatic histological findings in suspected drug-induced liver injury: systematic evaluation and clinical associations. Hepatology 2014;59:661–670.
36. Mosedale M, Watkins PB. Drug-induced liver injury: advances in mechanistic understanding that will inform risk management. Clin Pharmacol Ther 2017;101:469–480.
37. Mizooku H, Kagawa Y, Matsuda K, et al. Histological and immunohistochemical evaluations of lobular dissecting hepatitis in American Cocker Spaniel dogs. J Vet Med Sci 2013;75:597–603.
38. Kanemoto H, Sakai M, Sakamoto Y, et al. American Cocker Spaniel chronic hepatitis in Japan. J Vet Intern Med 2013;27:1041–1048.
39. Schermerhorn T, Center SA, Dykes NL, et al. Characterization of hepatoportal microvascular dysplasia in a kindred of Cairn Terriers. J Vet Intern Med 1996;10:219–230.
40. Jung T, Bader N, Grune T. Lipofuscin formation, distribution, and metabolic consequences. Ann N Y Acad Sci 2007;1119:97–111.
41. Seehafer SS, Pearce DA. Review: you say lipofuscin, we say ceroid: defining autofluorescent storage material. Neurobiol Aging 2007;26:576–588.
42. Davies MH, Harrison RF, Elias E, et al. Antibiotic-associated acute vanishing bile duct syndrome: a pattern associated with severe, prolonged, intrahepatic cholestasis. J Hepatol 1994;20:112–116.
43. O'Donohue J, Oien KA, Donaldson P, et al. Co-amoxiclav jaundice: clinical and histological features and HLA class II association. Gut 2000;47:717–720.
44. Craig CEH, Quaglia A, Seldon C, et al. The histopathology of regeneration in massive hepatic necrosis. Semin Liver Dis 2004;24:49–64.
45. Puls F, Agne C, Klein F, et al. Pathology of flupirtine-induced liver injury: a histological and clinical study of six cases. Virchows Arch 2011;458:709–716.
46. Masia R, Pratt DS, Misdraji J. A histopathologic pattern of centrilobular hepatocyte injury suggests 6-mercaptopurine-induced hepatotoxicity in patients with inflammatory bowel disease. Arch Pathol Lab Med 2012;136:618–622.
47. Johncilla M, Misdraji J, Pratt DS, et al. Ipilimumab-associated hepatitis clinicopathologic characterization in a series of 11 cases. Am J Surg Pathol 2015;39:1075–1084.
48. Takeda K, Oda M, Okada T, et al. Over-the-counter drug induced autoimmune hepatitis. Intern Med 2016;55:1293–1297.
49. Buechter M, Manka P, Heinemann FM, et al. Potential triggering factors of acute liver failure as a first manifestation of autoimmune hepatitis-a single center experience of 52 adult patients. World J Gastroenterol 2018;24:1410–1418.
50. Zhou S, Chan E, Duan W, et al. Drug bioactivation covalent binding to target proteins and toxicity relevance. Drug Metab Rev 2005;37:41–213.
51. Benvenuti F. The dendritic cell synapse: a life dedicated to T cell activation. Front Immunol 2016;7:70.
52. Verboogen DR, Dingjan I, Revelo NH, et al. The dendritic cell side of the immunological synapse. Biomol Concepts 2016;7:17–28.
53. Hivroz C, Chemin K, Tourret M, et al. Crosstalk between T lymphocytes and dendritic cells. Crit Rev Immunol 2012;32:139–155.
54. Lauschke VM, Ingelman-Sundberg M. The importance of patient-specific factors for hepatic drug response and toxicity. Int J Mol Sci 2016;17:E1714.
55. Seripa D, Pilotto A, Panza F, et al. Pharmacogenetics of cytochrome P450 (CYP) in the elderly. Ageing Res Rev 2010;9:457–474.
56. Akhtar S, Ramani R. Geriatric pharmacology. Anesthesiol Clin 2015;33:457–469.
57. Seripa D, Panza F, Daragjati J, et al. Measuring pharmacogenetics in special groups: geriatrics. Expert Opin Drug Metab Toxicol 2015;11:1073–1088.
58. Chen M, Borlak J, Tong W. High lipophilicity and high daily dose of oral medications are associated with significant risk for drug-induced liver injury. Hepatology 2013;58:388–396.
59. Backman JT, Filppula AM, Niemi M, et al. Role of cytochrome p450 2c8 in drug metabolism and interactions. Pharmacol Rev 2016;68:168–241.
60. Daneshmend TK, Warnock DW. Clinical pharmacokinetics of systemic antifungal drugs. Clin Pharmacokinet 1983;8:17–42.
61. Yu K, Geng X, Chen M, et al. High daily dose and being a substrate of cytochrome P450 enzymes are two important predictors of drug-induced liver injury. Drug Metab Dispos 2014;42:744–750.
62. McEuen K, Borlak J, Tong W, et al. Associations of drug lipophilicity and extent of metabolism with drug-induced liver injury. Int J Mol Sci 2017;18:E1335.
63. Baxter JG, Brass C, Schentag JJ, et al. Pharmacokinetics of ketoconazole administered intravenously to dogs and orally as tablet and solution to humans and dogs. J Pharm Sci 1986;75:443–447.
64. Court MH. Canine cytochrome P-450 pharmacogenetics. Vet Clin North Am Small Anim Pract 2013;43:1027–1038.
65. Heel RC, Brogden RN, Carmine A, et al. Ketoconazole: a review of its therapeutic efficacy in superficial and systemic fungal infections. Drugs 1982;23:1–36.
66. Buchi KN, Gray PD, Tolman KG. Ketoconazole hepatotoxicity: an in vitro model. Biochem Pharmacol 1986;35:2845–2847.
67. Rodriguez RJ, Acosta D. Comparison of ketoconazole- and fluconazole-induced hepatotoxicity in a primary culture system of rat hepatocytes. Toxicology 1995;96:83–92.
68. Fukami T, Iida A, Konishi K, et al. Human arylacetamide deacetylase hydrolyzes ketoconazole to trigger hepatocellular toxicity. Biochem Pharmacol 2016;116:153–161.
69. Rodriguez RJ, Acosta D. N-deacetyl ketoconazole-induced hepatotoxicity in a primary culture system of rat hepatocytes. Toxicology 1997;117:123–131.
70. Rodriguez RJ, Proteau PJ, Marquez BL, et al. Flavin-containing monooxygenase mediated metabolism of N-deacetyl ketoconazole by rat hepatic microsomes. Drug Metab Dispos 1999;27:880–886.
71. Haegler P, Joerin L, Krahenbuhl S. Hepatocellular toxicity of imidazole and triazole antimycotic agents. Toxicol Sci 2017;157:183–195.
72. Rodriguez RJ, Buckholz CJ. Hepatotoxicity of ketoconazole in Sprague-Dawley rats: glutathione depletion, flavin-containing monooxygenases-mediated bioactivation and hepatic covalent binding. Xenobiotica 2003;33:429–441.
73. US FDA. FDA Drug Safety Communication: FDA limits usage of Nizoral (ketoconazole) oral tablets due to potentially fatal liver injury and risk of drug interactions and adrenal gland problems. Available at: www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-limits-usage-nizoral-ketoconazole-oral-tablets-due-potentially. Accessed Sep 3, 2019.
74. Pessayre D, Fromenty B, Berson A, et al. Central role of mitochondria in drug-induced liver injury. Drug Metab Rev 2012;44:34–87.
75. Gaetke LM, Chow-Johnson HS. Copper: toxicological relevance and mechanisms. Arch Toxicol 2014; 88:1929–1938.
76. Greenblatt HK, Greenblatt DJ. Liver injury associated with ketoconazole: review of the published evidence. J Clin Pharmacol 2014;54:1321–1329.
77. Wree A, Dechěne A, Herzer K, et al. Steroid and ursodesoxycholic acid combination therapy in severe drug-induced liver injury. Digestion 2011;84:54–59.
