Relationship between daily dose of oral medications and idiosyncratic drug-induced liver injury: Search for signals†
Potential conflict of interest: Dr. Chalasani is a consultant for Takeda, Pfizer, Advanced Life Sciences, Atherogenics, Metabasis, and Lilly. He also received grants from Sanofi-Roche. Dr. Bjornsson is a consultant for AstraZeneca and Astellas.
Idiosyncratic drug-induced liver injury (DILI) is traditionally thought not to be dose-related. However, it has been pointed out that most medicines that were withdrawn from marketing or received a black-box warning because of hepatotoxicity were prescribed at daily doses greater than 50 mg/day. To examine the relationship between daily dose of medications and idiosyncratic DILI, we conducted a study with two aims. First, using two pharmaceutical databases, we examined the relationship between daily dose of commonly prescribed medicines in the United States and reported frequency of their selected hepatic adverse events. Second, we examined serious DILI cases reported to the Swedish Adverse Drug Reactions Advisory Committee (1970-2004) for any signals supporting the relationship between daily dose and idiosyncratic DILI. Medications were categorized into ≤10 mg/day, 11-49 mg/day, and ≥50 mg/day groups. Among US prescription medicines, a statistically significant relationship was observed between daily dose of oral medicines and reports of liver failure (P = 0.009), liver transplantation (P < 0.001), and death caused by DILI (P = 0.004) but not alanine aminotransferase (ALT) > 3 × upper limit of normal (P = 0.10) or jaundice (P = 0.16). Of 598 eligible Swedish DILI cases, 9% belonged to the ≤10 mg/day group, 14.2% to the 11-49 mg/day group, and 77% of cases were caused by medications given at dose ≥50 mg/day. A statistically significant relationship was noted between daily dose and poor outcome (death or liver transplantation) of Swedish DILI cases (2%, 9.4%, and 13.2% in ≤10, 11-49, and ≥50 mg/day groups, respectively, P = 0.03). Conclusion: These data suggest a relationship between daily doses of oral prescription medications and idiosyncratic DILI. More studies are needed to validate these observations and to explore their implications. (HEPATOLOGY 2008.)
Provide feedback or get help You are viewing our new enhanced HTML article.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Idiosyncratic drug-induced liver injury (DILI) is rare, but when it occurs it may have serious consequences.1 It is one of the most frequent causes of acute liver failure in the United States1, 2 and in fact is one of the most common reasons for not receiving approval or for withdrawal from marketing by the United States Food and Drug Administration.3 The pathogenesis of idiosyncratic DILI is not well understood but is generally thought to be unpredictable and not dose-dependent. However, it has been pointed out that most drugs that either have been withdrawn from the market or have received a black box warning due to hepatotoxicity were prescribed at daily doses greater than 50 mg, suggesting some dose relationship4 (Table 1). There is some relationship between daily dose of a medication (for example, statins, bosentan) and the reported frequency of elevated aminotransferases, suggesting that in some instances DILI may be dose-dependent.5, 6 Anecdotally, we have observed many instances in which individuals developed clear-cut hepatotoxicity on increasing the dose of a medication that they have received at stable doses for a lengthy period. For example, a middle-aged woman without known underlying liver disease or alcoholism developed pronounced hepatotoxicity caused by duloxetine very soon after increasing its dose to 60 mg/day, although she received 30 mg/day duloxetine for many weeks with no complications (unpublished data). These observations raise the possibility that some relationship may exist between daily dose of a medication and its propensity to exhibit hepatotoxicity. However, this has not been previously studied in a formal fashion.
Table 1. Daily Doses of Medications That Either Have Been Withdrawn from the United States Market or Have Received Hepatotoxicity Warning by the FDA
|Strong Warning of Serious Hepatotoxicity||Moderate Warning of Serious Hepatotoxicity|
|Compound||Daily Dosage||Compound||Daily Dosage|
|Valproic acid||250–1250 mg||Zileuton||1600–2400 mg|
|Ketoconazole||200–800 mg||Tacrine||40–160 mg|
|Nicotinic acid||100–4500 mg||Labetalol||200–400 mg|
|Rifampin||300–1200 mg||Diclofenac||75–200 mg|
|Chlorzoxazone||750–3000 mg||Dantrolene||25–400 mg|
|Isoniazid||300 mg|| || |
|Dantrolene||25–400 mg|| || |
|Nefazadone||200–600 mg|| || |
|Telethromycin||800 mg|| || |
|Nevirapine||200–400 mg|| || |
|Atomoxetine||40–100 mg|| || |
|Infliximab||5 mg/kg|| || |
|Assigned as Second-line Agents Because of Hepatotoxicity|
|Pemoline||mean 56-75 mg, max 112.5 mg|
To explore for an association between daily dose of medications and idiosyncratic DILI, we conducted a study with two specific aims. First, using two comprehensive pharmaceutical databases, we examined the relationship between daily dose of oral medications that are commonly prescribed in the United States and reported frequency of their hepatic adverse events. Second, we examined the daily doses of oral medications that have been implicated in the reports of suspected hepatic adverse drug reactions submitted to the Swedish Adverse Drug Reactions Advisory Committee during the period 1970 to 2004.
Materials and Methods
We used a publicly available pharmacy database (www.drugtopics.com) to extract the names of the top 200 brand and top 200 generic medications by prescription volume in the United States during the year 2005.7, 8 Duplicate compounds were reduced to a single entry, and nonoral medications were removed. List compilation yielded 230 medications available for further consideration. These compounds were further categorized into dosage groups of 10 mg or less, 11 to 49 mg, and 50 mg or greater based on daily recommended doses (Table 2). The compounds with broad ranges of recommended daily dose were placed within dosage groups based on the average of maximum and minimum recommended daily dose. For example, atorvastatin, with a daily recommended dose ranging between 10 and 80 mg (average, 45 mg) was placed in the 10-mg to 50-mg group. We subsequently reviewed each of these medications to assess whether they were ever reported to have caused selected hepatic adverse events, using the Thompson's Micromedex Drugdex System. Drugdex is one of the most comprehensive pharmacy databases, consisting of package insert data and published literature.9 Hepatic adverse events selected for this specific aim were alanine aminotransferase (ALT) greater than 3 times the upper limit of normal, cholestatic jaundice, liver failure, liver transplantation, and death caused by hepatic injury. To ensure completeness, each listed compound was cross-checked in the PubMed, Adverse Event Reporting System database, and the Physicians' Desk Reference.10–12 For each listed compound, we extracted whether these selected hepatic adverse events were reported rather than the number of events reported.
Table 2. Most Commonly Prescribed Medicines in the United States Categorized into Three Dose Groups
|Alprazolain||Accupril||Allegra-D 12 hour|
|Bisoprolol/HCTZ||Benicar HCT||Atenolol chlorthal|
|Clannex||Citalopram HBR||Bupropion hydrochloride|
|EE/ethynodiol di||Diphenoxylate w/Atro||Cimetidine|
|Felodipine ER||Doxepin||Ciprofloxacin HCl|
|Fosinopril sod||Farnotidine||Clopidrogel hydrogen sulfate|
|Glipizide XL||Flexeril||Diclofenac Sodium|
|Indapamide||Furoscmide oral||Diltiazem hydrochloride|
|Levonorgestrel||Geodon oral||Docusate sodium|
|Lunesta||Imitrex Oral||Fexofenadine hydrochloride|
|Premarin||Meclizine HCl||Imipramine HCl|
|Tizanidine HCl||Nadolol||Losartan potassium-hydrochlorothiazide|
| ||Paxil CR||Metronidazole Tabs|
| ||Pravachol||Moxifloxacin hydrochloride|
| ||prednisolone||Mytussin AC|
| ||Prochlorperaz mal||Naproxen|
| ||Promethazine tabs||niacin|
| ||Ritalin-LA||Oseltamivir phosphate|
| ||Spironolactone||Penicillin VK|
| ||Thyroid, Armour||Phenazopyridine HCl|
| ||Timolol Maleate GFS||Phenobarbital|
| ||Tuprol XL||Phenytoin|
| ||Triamterene w/HCTZ||Propranolol hydrochloride|
| ||Tussionex||Quetiapine furnarate|
| ||Viagra||Quinine sulfate|
| ||Vytorin||Raloxifene hydrochloride|
| ||Zelnorm||Ranitidine hydrochloride|
| || ||Terbinafine hydrochloride|
| || ||Tetracycline|
| || ||Theophylline SR|
| || ||Topiramate|
| || ||Trandolapril/verapamil|
| || ||Trazodone HCl|
| || ||Trimethoprim/Sulfa|
| || ||Valacyclovir hydrochloride|
| || ||Valproic Acid (divalproex sodium)|
| || ||Verapamil SR|
All reports of suspected drug-induced liver injury received by the Swedish Adverse Drug Advisory Committee since 1970 have been computerized. The reporting of fatal, otherwise serious, and new reactions has been compulsory since 1975. In a previous paper, the number and nature of suspected adverse drug-induced liver disease associated with fatalities or leading to liver transplantation between 1966 and 2002 were reported (Swedish DILI death/transplant cases).13 In a subsequent paper, all cases of suspected DILI with concomitant jaundice reported between 1970 and 2004 were reviewed for assessing outcome and prognostic markers of severe DILI (Swedish DILI jaundice cases).14 Cases contained within these two data sets were retrieved, and duplicate entries were excluded from further consideration. The causality was rigorously assessed according to the International Consensus Criteria,15–17 and all cases had least “possible” causality relationship. Apart from the daily dose of the implicated medication, the following information was retrieved from the reports: drug(s) suspected, results of peak aspartate aminotransferase, ALT, alkaline phosphatase, and bilirubin values, and outcome (recovery, death, or liver transplantation). Cases with more than one drug suspected to have caused DILI and acetaminophen alone or in combination with other drugs were excluded. Only oral medications were included in the analysis. The implicated medications were categorized into 10 mg or less, 11 to 49 mg, and 50 mg or greater based on the daily dosage. The objectives were to examine how many Swedish DILI cases belonged to each of these dosage groups and to explore the relationship between daily dose and outcome of Swedish DILI cases.
For aim 2, as a supplement to the Swedish DILI data, we reanalyzed the paper published by Russo et al.,18 which consisted of 270 DILI cases that required liver transplantation in the United States between 1990 and 2002.18 Of these 270 cases, 133 were acetaminophen (APAP) related, and 137 were non–APAP related. We examined how many of these non-APAP cases belonged to 10 mg/day or less, 11 to 49 mg/day, and 50 mg/day or greater groups.
For aim 1, proportions were used to describe the clinical dichotomous outcomes. The Kruskal-Wallis test compared the distribution of the number of times a compound was prescribed in 2005 among the three groups of compounds. The Cochran-Armitage trend test was applied for association between clinical outcomes and compounds. The overall logistic regression model linked the odds of ever reporting an incidence of clinical outcome to the covariate, three groups of compounds. For the validity of the model fitting for the transplant outcome, we used two groups of compounds, less than 50 mg versus 50 mg or greater, because none of the compounds in the less than 10 mg group had undergone transplantation. Estimated odds ratios and 95% confidence intervals were reported. For aim 2, descriptive statistics and Fisher's exact test were used to analyze the data. A P value of less than 5% was considered statistically significant. SAS (version 9.1, Cary, NC) was used for statistical analyses.
There were 54 compounds in the 10 mg/day or less group, 83 in the 11-mg to 49-mg/day group, and 93 in the 50 mg/day or greater group (Table 2). In 2005, the total number of prescriptions written ranged from 1,258,000 to 34,230,000 for the 10 mg/day or less group, 1,260,000 to 47,829,000 for the 11-mg to 49-mg/day group, and 1,286,000 to 101,639,000 for the 50 mg/day or greater group. There was no statistically significant difference in the median number of prescriptions written in 2005 among these three groups (P = 0.37; Table 3).
Table 3. Association Between Daily Doses of Oral Prescription Medications and Hepatic Adverse Events
|ALT > 3 × ULN (n, %)||10 (19)||22 (27)||29 (31)||0.10|
|Jaundice (n, %)||18 (33)||33 (40)||42 (45)||0.16|
|Liver Failure (n, %)||9 (17)||10 (12)||30 (32)||0.009|
|Death (n, %)||6 (11)||9 (11)||26 (28)||0.004|
|Transplant (n, %)||0 (0)||2 (2)||12 (13)||<0.001|
|Prescriptions (median in 2005)||4,746,500||4,938,000||3,733,000||0.3|
There was a statistically significant relationship between recommended daily dose and reports of liver failure, liver transplantation, or liver-related death (Tables 3 and 4). Seventeen percent of compounds belonging to the 10 mg/day or less group had been reported to cause liver failure, in comparison with 12% in the 11-mg to 49-mg/day group and 32% in the 50 mg/day or greater group (P = 0.009). No compounds belonging to the 10 mg/day or less group caused liver transplantation, in comparison with 2% in the 11-mg to 49-mg/day group and 14% in the 50 mg/day or greater group (P < 0.001). Eleven percent of compounds belonging to the 10 mg/day or less group and the 11-mg to 49-mg/day group were reported to have caused DILI-related deaths, in comparison with 28% of compounds belonging to the 50 mg/day or greater group (P = 0.007). There was no statistically significant association between daily recommended dose of oral prescription medications and reports of ALT more than 3× the upper limit of normal or jaundice (Tables 3 and 4).
Table 4. Relationship Between Daily Dose and Different Hepatic Events: Results from the Logistic Regression Analysis
|ALT > 3 × ULN|| || || |
|Compound|| || || |
| ≤ 10 mg||1.00|| || |
| 11–490 mg||1.59||0.68–3.68||0.28|
| ≥50 mg||1.99||0.88–4.50||0.10|
|Jaundice|| || || |
|Compound|| || || |
| ≤10 mg||1.00|| || |
| 11–49 mg||1.32||0.65–2.70||0.45|
| ≥50 mg||1.65||0.82–3.31||0.16|
|Liver failure|| || || |
|Compound|| || || |
| ≤10 mg||1.00|| || |
| 11–49 mg||0.69||0.26–1.81||0.446|
| ≥50 mg||2.38||1.03–5.50||0.042|
|Death|| || || |
|Compound|| || || |
| ≤10 mg||1.00|| || |
| 11–49 mg||0.97||0.33–2.91||0.961|
| ≥50 mg||3.10||1.19–8.12||0.021|
|Transplant|| || || |
|Compound|| || || |
| <50 mg||1.00|| || |
| ≥ 50 mg||10.00||2.18–45.8||0.003|
The logistic regression analyses showed that a significantly higher proportion of medications belonging to the 50 mg/day or greater group had liver failure (odds ratio, 2.38; 95% confidence interval, 1.03-5.50), liver transplantation (odds ratio, 10.00; 95% confidence interval, 2.18-45.81), or liver-related death (odds ratio, 3.10; 95% confidence interval, 1.19-8.12), whereas medications belonging to the 10 mg/day or less and 11-mg to 49-mg/day groups had no differences (Table 4).
After applying exclusion criteria to Swedish DILI death/transplant and jaundice cases, a total of 598 cases were eligible for further analyses (58% females, median age of 59 years (interquartile range, 42-74); 51% had hepatocellular type and 49% had cholestatic/mixed type of liver injury. Seventy-seven percent of Swedish DILI cases belonged to the 50 mg/day or greater group, whereas only 9% belonged to the 10 mg/day or less group, and 14% belonged to the 11-mg to 49-mg/day group (Table 5). The pattern of liver injury and patient outcomes in different dose categories are shown in Table 5. Only one patient treated with 10 mg/day or less died, and only very few other patients with a lower daily dose than 50 mg/day either died or underwent liver transplantation (Table 5). A statistically significant relationship was noted between daily dose and poor outcome (death or liver transplantation) of Swedish DILI cases (13.2%, 9.4%, and 2% in ≥50, 11-49, and ≤10 mg/day groups, respectively; P = 0.03). Selected details of patients belonging to the 10 mg/day or less group and the 11-mg to 49-mg/day group who had poor outcome after DILI event are shown in Table 6. As a control measure, we assessed the proportion of prescription medications and their prescription volume in Sweden that belong to each of these dosage groups. After excluding nonoral formulations from 200 most prescribed medicines in Sweden in 2005, 37.5% belonged to the 50 mg/day or greater group, 27.5% belonged to the 11-mg to 49-mg/day group, and 22.5% belonged to the 10 mg/day or less group. The median number of prescriptions written in 2005 in Sweden were 249,197 (range, 101,094-1,847,843) for the 10 mg/day or less group, 360,149 (range, 99,986-2,647,547) for the 11-mg to 49-mg/day group, and 215,760 (range, 104,073-3,601,864) for the 50 mg/day or greater group.
Table 5. Types of Liver Injury and Outcome Stratified According to Daily Dose: The Swedish Hepatic ADR Data (Total Eligible Cases = 598)
|Number of DILI cases||53 (8.9%)||85 (14.2%)||460 (76.9%)|
|Pattern of injury|| || || |
| Hepatocellular||34 (64.2%)||43 (50.6%)||231 (50.2%)|
| Cholestatic||13 (24.5%)||22 (25.9%)||124 (26.9%)|
| Mixed||6 (11.3%)||20 (37.8%)||104 (22.6%)|
|Outcome|| || || |
| Death/liver transplantation*||1 (2%)||8 (9.4%)||61 (13.2%)*|
| Survived||52 (98%)||77 (90.6%)||399 (86.8%)|
|Top 200 prescribed medicines in Sweden in 2005|| || || |
| Proportion belonging to each dose group||22.5%||27.5%||37.5%|
| Median # of prescriptions||249,197||360,149||215,760|
Table 6. Selected Details of Patients with Poor Outcome DILI (Death or Transplantation) Caused by Compounds Given at <50 mg Daily Dose: The Swedish Hepatic ADR Dataset
Of 137 non-APAP DILI cases that required liver transplantation in the United States between 1990 and 2002, after excluding cases caused by inhalation agents (n = 4), intravenous agents (n = 3), herbal agents (n = 7), amanita mushrooms (n = 9), and combination of agents (n = 3), 111 cases were further analyzed. Of these 111 cases, two cases were caused by compounds belonging to the 10 mg/day or less group, eight belonged to the 11-mg to 49-mg/day group, and 101 cases were reportedly caused by compounds with a daily recommended dose of 50 mg or greater (Table 7).
Table 7. DILI Requiring Liver Transplantation in the United States From 1990 to 2002 According to the UNOS Database
|Individual agents (number of cases)||Cerivastatin (2)||Fialuridine (3)||Isoniazid (24)|
| ||Lisinopril (1)||Propylthiouracil (13)|
| ||Simvastatin (1)||Phenytoin (10)|
| || ||Pemoline (1)||Valproate (10)|
| || ||Zafirkulast (1)||Nitrofurantoin (7)|
| || ||Paroxctine (1)||Ketoconazole (6)|
| || || ||Disulfiram (6)|
| || || ||Troglitazone (4)|
| || || ||Sulfasalazine (3)|
| || || ||Methyldopa (3)|
| || || ||Nefazadone (2)|
| || || ||Labetalol (2)|
| || || ||Amoxicillin/Clavulan (1)|
| || || ||Bromfenac (1)|
| || || ||Ibuprofen (1)|
| || || ||Hydrocodone (1)|
| || || ||6-Mercaptopurine|
| || || ||Itraconazole (1)|
| || || ||Carbamazepine (1)|
| || || ||Trimethoprim/Sulfametho|
| || || ||Buproprion (1)|
| || || ||Iron (1)|
| || || ||Naproxen (1)|
Although idiosyncratic DILI is traditionally thought to be not dose-related, some experts have noted that most medications that have either been withdrawn or received black box warning because of hepatotoxicity were prescribed at daily doses greater than 50 mg (Table 1).4 This observation in conjunction with our own anecdotal experience prompted us to search for any epidemiological signals supporting such a relationship in the existing literature. Our study consisting of different data sets indeed supports a relationship between the daily dose of an oral medication and its propensity to cause serious hepatotoxicity.
Using two comprehensive pharmacy databases, we examined the relationship between recommended daily doses of 230 oral medications commonly prescribed in the United States and reports of selected hepatic adverse events. A significantly greater proportion of oral medications belonging to the 50 mg/day or greater group were reported to cause liver failure and liver failure leading to death or liver transplantation than medications belonging to other two groups. There was a trend in the relationship between daily dose and reports of ALT greater than 3× ULN and jaundice, but they did not reach statistical significance. Because we had difficulty in distinguishing cholestatic and hepatocellular jaundice from the reports contained within the DRUGDEX, we included both cholestatic and hepatocellular cases in our jaundice category. This may explain in part why we failed to observe a statistically significant relationship between daily dose and reports of jaundice in the US prescription medication data set. The results from the Swedish ADR data set are noteworthy in that 77% of all serious DILI cases belonged to the 50 mg/day or greater group. Furthermore, DILI events belonging to the 50 mg/day or greater group were more likely to lead to death or liver transplantation when compared with the other two groups combined. Furthermore, US liver transplantation data also revealed that there is significant enrichment of DILI cases requiring liver transplantation in the 50 mg/day or greater group.
It has been hypothesized that idiosyncratic DILI could either be attributable to hypersensitivity or to metabolic idiosyncracy.19 Our epidemiological study is not able to address whether the dose of a compound is more relevant for one particular mechanism of idiosyncratic DILI; however, intuitively one may suggest that daily dose may be more relevant for metabolic idiosyncrasy rather than hepatotoxicity caused by hypersensitivity. The categories of compounds known to cause idiosyncratic DILI at a higher frequency (for example, antibacterials, nonsteroidal anti-inflammatory drugs) were overrepresented in the 50 mg/day or greater group. However, it is not known whether these categories of compounds are more hepatotoxic because their daily dose was often 50 mg or greater or whether the noted association between daily dose and DILI is driven by overpopulation of these categories in the 50 mg/day or greater group. If confirmed by other studies, the noted association between daily dose and idiosyncratic DILI would support a reactive metabolite hypothesis for the pathogenesis of idiosyncratic DILI.
Some aspects of our study deserve further discussion. First, this is a retrospective analysis of existing literature and thus its findings should be viewed as epidemiological signals rather than factual observations. However, questions of this nature cannot easily be addressed by prospective randomized studies because of the very rare occurrence of clinically significant hepatotoxicity. Second, because our study dealt exclusively with oral medications, it is not possible to extrapolate our observations to parenteral, topical, or inhalation formulations. Finally, although 50 mg/day or greater appears to be the threshold daily dose beyond which the risk of DILI may be significantly higher compared with lower doses, it is not an absolute threshold. In other words, although serious DILI is rarely caused by compounds belonging to the 10 mg/day or less group, it is not that rare to exclude DILI as a possibility when an individual consuming a medication with a daily dose of 10 mg/day or less develops acute liver injury.
We believe that our observations are worthy of further investigation. Proprietary data sets possessed by the pharmaceutical companies, the Adverse Event Reporting System database, and prospective studies such as that conducted by the Drug Induced Liver Injury Network (DILIN) may provide additional opportunities to further explore this relationship. If these results are supported by other investigations, these observations are of importance in drug development to avoid problems with DILI in the future.