Are you treating the patient or killing the liver?

A wide variety of drugs are prescribed to patients every day. Some are needed because the damage or discomfort from the disease justifies the use of the drug. However a wide variety of herbs, vitamin supplements, metal supplements are used and promoted on the basis of very doubtful scientific or medical research or information. Add to this the number of drugs which are considered safe but can cause a totally unpredictable idiosyncratic reaction which is based on an individual’s reaction which can be toxic although the majority of patients do not have this reaction. As I emphasize a doctor cannot tell in advance if an idiosyncrasy exists. Let me show you a slide predicting some of the reactions.

How would you know that the drug has caused the liver injury? Note the time between the administration of the drug and the onset of liver injury. You can do this more efficiently if you frequently check the liver enzymes and maintain accurate records. Note if the liver injury subsides after withdrawing the drug. If you are very brave or foolhardy then re-administering the drug and watching out for more injury is an option. Keep in mind that the incidence is estimated to be 14 to 19 cases per 100,000 persons, with jaundice accompanying 30% of cases. Remember that the progress on the liver injury fronts has been modest in cases related to drug induced injury. If you are aware of the drugs’ capacity for liver injury it helps to make a diagnosis. The symptoms of liver failure are common to all forms of liver disease. The tests done like hepatic enzymes, levels of bilirubin and liver biopsy and forms of imaging do not point at the drug but help to rule out other causes. Keep in mind that the incidence is estimated to be 14 to 19 cases per 100,000 persons, with jaundice accompanying 30% of cases. Remember that the progress on the liver injury fronts has been modest in cases related to drug induced injury. If you are aware of the drugs’ capacity for liver injury it helps to make a diagnosis.

LiverTox, the National Institutes of Health–sponsored website on hepatotoxicity, has descriptions of more than 1200 agents (prescription and over-the-counter medications, herbal products, nutritional supplements, metals, and toxins), along with their potential to cause liver injury. Liver injury can be direct if the drug or one of its metabolites is directly toxic to the liver cells. The injury is predictable, dose related and the onset is usually within 1-5 days of taking the drug.

This is a list of drugs which frequently cause idiosyncratic reactions taken from a review article in the NEJM.. These reactions are unpredictable, not dose related or reproducible in animals. The reaction may be cholestatic or hepatocellular. The drugs listed here are commonly used drugs generally considered safe for the liver.


Serum enzyme elevations without jaundice constitute the most common pattern of direct drug-induced liver injury, with elevations of alanine aminotransferase or alkaline phosphatase levels but without hyperbilirubinemia and with minimal or no symptoms. The elevations resolve when the drug is stopped or the dose is lowered but can also resolve spontaneously, a phenomenon referred to as adaptation. In some cases, adaptation does not occur, and enzyme elevations worsen and jaundice and symptoms arise.  Acute hepatic necrosis is the most common form of clinically apparent direct hepatotoxicity. The injury occurs abruptly, soon after the medication has been started, often after exposure to a single high dose or a dose increase.

High doses of acetaminophen, aspirin, niacin, amiodarone, and many antineoplastic agents can cause acute hepatic necrosis. Typically, these drugs can be restarted at lower doses without a recurrence of injury. Poisonous mushrooms (Amanita phalloides) and other environmental toxins can cause a similar syndrome of acute hepatic necrosis.

Sinusoidal obstruction syndrome, previously known as veno-occlusive disease, is due to acute injury and loss of intrasinusoidal endothelial cells, resulting in obstruction of sinusoidal blood flow and liver injury. Drugs are the usual cause, the most common being myeloablative agents administered in preparation for hematopoietic cell transplantation. Symptoms of abdominal pain, increase in liver size, and weight gain, followed by jaundice, appear 1 to 3 weeks after exposure and may progress rapidly to hepatic failure. Drugs that cause sinusoidal obstruction syndrome include alkylating agents such as busulfan or cyclophosphamide and monoclonal antibody–cytotoxic conjugates such as gemtuzumab ozogamicin. The syndrome can also be caused by botanicals (pyrrolizidine alkaloids). Defibrotide, an antithrombotic agent, has recently been approved as therapy for severe sinusoidal obstruction syndrome with organ failure, but its use is controversial.

Nodular regenerative hyperplasia is usually manifested as unexplained, noncirrhotic portal hypertension with esophageal varices or ascites. Nodular regeneration can be caused by cancer chemotherapeutic agents given over a long period or in multiple courses (azathioprine, mercaptopurine, or thioguanine) or by first-generation nucleoside antiretroviral agents (zidovudine, stavudine, or didanosine). Nodular regenerative hyperplasia with resultant portal hypertension has also been linked to oxaliplatin infusions for metastatic colon cancer. 

Lactic acidosis with microvesicular steatosis and hepatic dysfunction typically occurs with nonspecific symptoms of abdominal discomfort, fatigue, and weakness, with subsequent confusion, stupor, and coma accompanied by liver injury. Lactic acidosis or hyperammonemia may be prominent. Jaundice arises late, and enzyme elevations are variable, sometimes markedly hepatocellular (with Reye’s syndrome triggered by aspirin) and sometimes with milder, mixed patterns. The time to onset can be days (with aspirin or intravenous tetracycline), weeks (with linezolid) or months (with didanosine). Liver biopsy shows microvesicular steatosis with minimal inflammation and necrosis. The pathogenesis of the injury is mitochondrial toxicity and failure of aerobic metabolism. Therapy should focus on withdrawal of the responsible agent, administration of glucose infusions, and correction of acidosis.


Acute hepatocellular hepatitis is the most common manifestation of idiosyncratic liver injury. The latency period generally ranges from 5 to 90 days. The symptoms and course resemble those of acute viral hepatitis, with prominent alanine aminotransferase elevations (increased by a factor of 5 to 50), whereas alkaline phosphatase levels are only modestly increased. Liver histologic studies show changes suggestive of acute viral hepatitis, the major disorder in the differential diagnosis, but eosinophils may be prominent. The rate of death from icteric hepatocellular injury due to medications is high, usually 10% or higher, a feature first stressed by the late Hyman J. Zimmerman, for which reason it is called Hy’s law. A key feature of Hy’s law is jaundice with hepatocellular rather than cholestatic injury. Drug-induced idiosyncratic acute hepatocellular injury is an important cause of acute liver failure, accounting for 11 to 15% of cases in series from the United States and Europe. Common causes of drug-induced idiosyncratic acute hepatocellular injury are isoniazid, nitrofurantoin, and diclofenac.

Chronic hepatitis is an uncommon form of drug-induced liver injury; the chronicity occurs if the agent is continued and typically resolves slowly once the agent has been stopped. Many agents that cause acute hepatocellular injury can also cause a chronic hepatocellular pattern. The injury arises after months or years of exposure. Autoantibodies are frequently present, and the differential diagnosis often focuses on ruling out spontaneous autoimmune hepatitis. Common causes of drug-induced, autoimmune-like chronic liver injury are nitrofurantoin, minocycline, hydralazine, methyldopa, statins, and fenofibrate. Glucocorticoids, which are frequently used to manage chronic hepatitis (starting dose, 20 to 60 mg of prednisone or its equivalent daily), may alleviate symptoms and speed recovery, but the injury will often resolve without intervention. If prednisone is used, the dose and duration should be kept to a minimum. Monitoring for evidence of relapse should be performed for at least 6 months after the withdrawal of glucocorticoids.

Cholestatic hepatitis is characterized by prominent symptoms of pruritus and jaundice accompanied by moderate-to-marked elevations in alkaline phosphatase levels. Drug-induced cholestatic liver injury is usually self-limited, and although often protracted, it ultimately resolves. Liver histologic studies show bile duct injury and cholestasis in small bile canaliculi. Exceptions to the usual benign course occur when there is bile duct loss, which is associated with delayed resolution of jaundice and elevated enzyme levels. Some cases evolve into vanishing bile duct syndrome, with prolonged jaundice, liver failure, need for liver transplantation, or death. Common causes of drug-induced cholestatic hepatitis are amoxicillin–clavulanate, cephalosporins, terbinafine, azathioprine, and temozolomide.

Drug-induced mixed hepatitis is caused by many agents, some of which also cause hepatocellular or cholestatic hepatitis. The mixed forms of drug-induced liver injury tend to have the most benign outcomes, rarely leading to liver failure. Common causes of drug-induced mixed hepatitis include the fluoroquinolone and macrolide antibiotics, phenytoin, and sulfonamides.

All forms of idiosyncratic drug-induced hepatitis can be accompanied by immunoallergic features, such as rash, fever, and eosinophilia — signs of drug hypersensitivity. More extreme examples include drug reaction with eosinophilia and systemic symptoms (DRESS) syndrome, toxic epidermal necrolysis, and the Stevens–Johnson syndrome. Prominent causes of idiosyncratic drug-induced hepatitis with immunoallergic features include allopurinol, carbamazepine, phenytoin, sulfonamides, and macrolide antibiotics. Immunoallergic hepatitis is more common among people of black American lineage than among white Americans.

Bland cholestasis represents a distinctive phenotype of drug-induced liver injury, characterized by marked and prolonged jaundice with pruritus. In women, bland cholestasis is typically caused by estrogens or oral contraceptives, and in men it is typically caused by anabolic steroids, usually obtained illicitly for bodybuilding or improving athletic performance. Jaundice and pruritus arise within 30 to 90 days, and elevations in enzyme levels are minimal or modest, despite marked and prolonged jaundice

Indirect drug-induced liver injury results from the medication’s actions rather than from its inherent hepatotoxic effects or immunogenicity; the injury represents induction or exacerbation of a liver disease. Fatty liver disease can be an indirect effect of drugs that cause weight gain (risperidone and haloperidol) or that alter triglyceride disposition (lomitapide) or insulin sensitivity (glucocorticoids). Acute hepatitis can be the indirect effect of anticancer chemotherapeutic agents that cause a reactivation of hepatitis B or of antiretroviral agents that cause immune reconstitution and exacerbation of hepatitis C. An increasingly common form of indirect injury is immune-mediated liver injury due to various immunomodulatory agents,tumor necrosis factor antagonists, and most dramatically, antineoplastic checkpoint inhibitors. Many of these agents are monoclonal antibodies and are thus unlikely to cause direct or idiosyncratic liver injury. Hepatocellular or mixed hepatitis with immune features usually arises within 2 to 12 weeks after the start of therapy (or after one to three courses) and is often detected during routine monitoring at the time of each infusion. Many cases are anicteric and asymptomatic, but without intervention, the hepatitis can worsen and become life-threatening. Therapy with glucocorticoids is usually recommended.62 If the injury resolves promptly, the agent can be restarted or another agent can be substituted (infliximab can be switched to etanercept, or ipilimumab to nivolumab).

Indirect liver injury is a new and not completely accepted category of hepatotoxicity. Indirect injury is much more frequent than idiosyncratic forms and is a common reaction to a whole class of medications (e.g., tumor necrosis factor antagonists and checkpoint inhibitors) rather than a rare and idiosyncratic reaction to a random, specific agent (e.g., nitrofurantoin or atorvastatin). Indirect drug-induced liver injury represents an expanded concept of hepatotoxicity and provides insights into liver conditions that are worsened (e.g., the types of immunomodulation that cause reactivation of hepatitis B) or into predispositions to liver conditions. There are plausible explanations for the pathogenesis of indirect injury, and in most instances, this type of drug-induced liver injury can be prevented or treated.

The 25 most commonly listed listed drugs causing liver injury do so because they are used very frequently, or for a long time i.e. 1-2 years or intravenously. These agents are amoxicillin–clavulanate, isoniazid, nitrofurantoin, trimethoprim–sulfamethoxazole, atorvastatin and minocycline. These medications might be the most common causes of idiosyncratic drug-induced liver injury, but liver injury in persons taking these drugs is rare.

The role of herbal and dietary supplements in causing acute liver injury is a growing and perplexing problem. In studies from the United States, the proportions of cases of liver injury caused by herbal or dietary supplements increased from 7 to 9% in 2004–2007 to 19 to 20% in 2010–2014. The specific products implicated are generally not single herbs (aloe vera, saw palmetto, or black cohosh) or single nutritional substances (creatine, omega fatty acids, or vitamins) but rather are typically multiple-ingredient dietary supplements marketed for weight loss, bodybuilding, or improvements in sexual function, general well-being, or mental acuity. These products often have 5 to 20 ingredients, including vitamins, minerals, proteins, and herbs or botanicals of uncertain quality and concentration, often referred to as a “proprietary blend.” The specific chemical component (or components) responsible for the liver injury is rarely obvious. Strikingly, the clinical phenotype of liver injury in most cases associated with herbal and dietary supplements is acute hepatocellular hepatitis, which is often severe, with a high rate of fulminant hepatic failure and need for liver transplantation. Commonly implicated components are green tea extracts (Camellia sinensis). The suspected active molecular constituents are catechins, which at high doses cause liver injury in animal models . Green tea extracts are often used to prevent breast cancer.

So how can we protect our patients? Use drugs only if absolutely necessary. Check renal and kidney functions frequently if you are prescribing a drug for a prolonged period. Take a history of what supplements and minerals and vitamins the patient is buying on their own, or what they are taking to prevent cancer, aging and dementia which are the commonly feared diseases of today. Keep alert to whether your patient is getting better or worse on treatment or if unusual symptoms are developing. pruritus, a rash, jaundice, a waning appetite should alert a wary physician to do do more tests specially for the liver. If the patient already has Hep B or C or steatosis then think twice about using a drug which may either damage the liver or put more fat in the organ. The regular use of alcohol by the patient should also make a physician careful about the medicines which affect the liver. Awareness of the likelihood of liver injury, being quick off the mark in doing tests for liver injury and withdrawing the drug as soon as possible are the best ways to safeguard your patients.

Published by


I am a Professor of Medicine and a Nephrologist. Having served in the Army Medical College, Pakistan Army for 27 years I eventually became the Dean and Principal of the Bahria University Medical and Dental College Karachi from where I retired in 2016. My passion is teaching and mentoring young doctors. I am associated with the College of Physicians and Surgeons Pakistan as a Fellow and an examiner. I find that many young doctors make mistakes because they do not understand how they should answer questions; basically they do not understand why a question is being asked. My aim is to help them process the information they acquire as part of their education to answer questions, pass examinations and to best take care of patients without supervision of a consultant. Read my blog, interact and ask questions so that I can help you more.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s