What Drugs Can Trigger Autoimmune Hepatitis?

Dozens of medications have been linked to autoimmune hepatitis, but a handful stand out as the most frequent and best-documented culprits. Antibiotics, biologic therapies used for inflammatory diseases, cholesterol-lowering statins, and newer cancer immunotherapies top the list. A large pharmacovigilance analysis of adverse-event reports identified 22 drugs with strong statistical signals for triggering the condition, spanning antibiotics, antivirals, cardiovascular drugs, cancer treatments, and even a few anti-inflammatory painkillers. The good news is that drug-induced autoimmune hepatitis generally carries a better outlook than the spontaneous form of the disease, often resolving once the offending medication is stopped.

The Two Classic Culprits Among Antibiotics

If you spend time reading about drug-induced autoimmune hepatitis, two drug names appear over and over: minocycline and nitrofurantoin. These have the longest track record and the most documented cases. Minocycline is a tetracycline antibiotic prescribed widely for acne, and nitrofurantoin is commonly used to prevent or treat urinary tract infections. Both tend to cause trouble after prolonged use rather than short courses.

In the case of minocycline, a systematic review of published cases and pharmacovigilance data found two distinct patterns of liver damage. One pattern resembled autoimmune hepatitis with lupus-like symptoms, appearing after a median of about a year of use in women and about two years in men. The other was a faster hypersensitivity reaction within roughly five weeks, involving skin rashes and elevated immune cells called eosinophils.¹ The autoimmune pattern is the one that genuinely mimics classic autoimmune hepatitis, presenting with the same antibodies and biopsy features you would expect in a patient who developed the disease on their own.²

Nitrofurantoin follows a similar story. Case series describe women on long-term nitrofurantoin prophylaxis for recurrent urinary infections who developed liver disease with autoimmune markers and biopsy findings indistinguishable from idiopathic autoimmune hepatitis. Some patients presented with jaundice, and at least one required prolonged hospitalization for liver failure.³ The common thread with both drugs is duration of exposure. Short antibiotic courses carry far less risk than months or years of continuous use.

Beyond minocycline and nitrofurantoin, a real-world pharmacovigilance study of the FDA’s adverse event reporting system flagged two other antibiotics with significant signals: doxycycline and azithromycin.⁴ Doxycycline belongs to the same tetracycline family as minocycline, so its presence on the list is not entirely surprising. Azithromycin is prescribed far more commonly and briefly, so the absolute risk appears to be very low, but it is worth knowing it has surfaced in reporting data.

Biologic Drugs Used for Inflammatory Diseases

A newer and increasingly recognized category of triggers involves biologic drugs, particularly the anti-TNF agents used to treat conditions like Crohn’s disease, rheumatoid arthritis, and psoriasis. Infliximab carries the strongest signal. An analysis of 389 cases of anti-TNF-associated autoimmune hepatitis in a large international pharmacovigilance database confirmed statistically significant associations for infliximab, adalimumab, and etanercept.⁵

A separate systematic review noted 16 cases linked to infliximab and four to adalimumab among published reports. Encouragingly, all of those cases improved once the biologic agent was stopped.⁶ One study tracking autoimmune hepatitis cases over time found that drug-induced cases were suspected in roughly 18% of diagnoses, with infliximab accounting for the majority.⁷

There is an irony here that is hard to miss: these drugs suppress part of the immune system to control one autoimmune condition, yet they can set off a different autoimmune attack on the liver. The prevailing theory is that blocking TNF-alpha shifts the balance of immune signaling in ways that can unmask latent autoimmunity in genetically susceptible people. This is a theme that runs through drug-induced autoimmune hepatitis in general, and we will come back to it.

Statins and Other Cardiovascular Medications

Statins are among the most widely prescribed drugs on the planet, and their appearance on the list of triggers gets attention. The same FDA adverse-event analysis that flagged antibiotics and biologics also identified atorvastatin, simvastatin, and rosuvastatin as having significant signals for drug-induced autoimmune hepatitis.⁴ Case reports have detailed patients who developed autoimmune-like hepatitis while taking atorvastatin, with compatible biopsy findings and a clear temporal link to the medication.⁸

It is worth putting this in perspective. Tens of millions of people take statins, and the absolute number of autoimmune hepatitis cases linked to them remains small. The risk is better described as rare than common, but it matters for doctors and patients to keep it on the radar, especially if unexplained liver enzyme elevations appear during statin therapy. Olmesartan, a blood pressure medication, also appeared in the pharmacovigilance data, and methyldopa, an older antihypertensive sometimes still used in pregnancy, has long been known to cause a hepatitis that mimics autoimmune hepatitis both clinically and on biopsy.⁹

Cancer Immunotherapy and Checkpoint Inhibitors

Immune checkpoint inhibitors have transformed cancer treatment over the past decade, but they work by unleashing the immune system, and the liver can get caught in the crossfire. Drugs like nivolumab, pembrolizumab, and atezolizumab all showed significant associations with autoimmune hepatitis in pharmacovigilance data.⁴

Checkpoint inhibitor liver injury has some distinctive features that set it apart from the classic drug-induced autoimmune hepatitis caused by antibiotics or biologics. In a histological study of seven cases, liver dysfunction appeared within about three to seventeen weeks of starting immunotherapy. All patients had elevated liver enzymes, but none tested positive for the antinuclear antibodies that typically characterize autoimmune hepatitis. Elevated IgG, another hallmark of classical autoimmune hepatitis, was also absent in all tested patients.¹⁰ A larger comparison study found that while over half of checkpoint inhibitor liver biopsies showed a hepatitis-like injury pattern, a substantial minority showed bile duct injury patterns that are uncommon in traditional autoimmune hepatitis.¹¹

This distinction matters clinically. Some researchers argue that checkpoint inhibitor hepatitis is better thought of as a related but separate entity rather than a true drug-induced autoimmune hepatitis. Regardless of the label, management usually involves stopping the immunotherapy and starting corticosteroids.

Why These Drugs and Not Others

The leading explanation involves the way certain drugs are processed in the liver. A reactive byproduct of the drug’s metabolism can bind to a protein on the surface of liver cells, creating what the immune system perceives as a foreign target. T cells that would normally ignore liver tissue get activated against this modified protein, and in people whose immune regulation is not quite robust enough to shut down the attack, an autoimmune-like response against the liver takes hold.¹² Regulatory immune cells that normally keep these responses in check play a critical role; when their function is deficient, the temporary confusion escalates into ongoing liver inflammation.¹³

This mechanism helps explain why duration matters for drugs like minocycline and nitrofurantoin. The longer the liver is exposed to the drug and its metabolites, the more opportunities exist for this abnormal immune recognition to develop. It also helps explain why the condition does not strike everyone, and why genetic susceptibility plays a significant role.

The Genetic Angle

Not everyone who takes minocycline for years or starts an anti-TNF biologic develops liver problems. Genetics explains a large part of this variability. Certain versions of immune-system genes, particularly the HLA genes that control how the body presents proteins to immune cells, dramatically raise the risk of drug-induced liver injury.

For minocycline specifically, research identified the HLA-B*35:02 allele as a significant risk factor. Roughly 16% of patients who developed minocycline-related liver injury carried this allele, compared to less than 1% of the general population, translating to about a 30-fold increased risk.¹⁴ Computer modeling suggests that minocycline may bind directly to this particular HLA molecule, which could be the initial spark that triggers the immune attack on the liver.¹⁵

Similar HLA associations exist for other drugs. The same review noted an 81-fold increased risk of flucloxacillin-related liver injury in people carrying the HLA-B*57:01 allele, and associations between HLA-DRB1*15:01 and liver injury from amoxicillin-clavulanate.¹⁵ These are not drugs you would get genetically tested for before a prescription in most clinical settings today, but the findings underscore that drug-induced autoimmune hepatitis is not random. It reflects a mismatch between a specific drug and a specific person’s immune hardware.

How Doctors Tell Drug-Induced From Spontaneous Autoimmune Hepatitis

This is one of the genuinely hard diagnostic problems in liver medicine. Drug-induced autoimmune hepatitis can look almost identical to the spontaneous form, with the same antibodies circulating in the blood and the same patterns of inflammation on biopsy. At the onset of drug-induced liver injury, studies have found that around 72% of patients have elevated antinuclear antibodies, about 60% have elevated smooth muscle antibodies, and roughly 39% have elevated IgG levels, all hallmarks you would also see in idiopathic autoimmune hepatitis.¹⁶

Liver biopsy can help somewhat. A study comparing biopsies from confirmed drug-induced liver injury with biopsies from autoimmune hepatitis found that certain features tip the balance. Clusters of plasma cells in the liver’s portal areas and a pattern called rosette formation (liver cells arranging themselves in ring-like clusters) favor a diagnosis of true autoimmune hepatitis, while the presence of certain white blood cells called neutrophils and signs of bile buildup inside liver cells lean toward drug-induced injury. A combination of these biopsy features could distinguish between the two with good accuracy.¹⁷

In practice, clinicians rely on validated scoring tools. The Roussel Uclaf Causality Assessment Method helps evaluate whether a drug is likely responsible for the liver injury, while the Simplified Autoimmune Hepatitis Score assesses how much the presentation resembles autoimmune hepatitis. Using both in tandem gives the clearest picture.¹⁸ The critical clue remains the timeline: did symptoms begin after starting a known trigger drug, and do they improve after stopping it?

What Happens After You Stop the Drug

This is where the news is largely reassuring. Drug-induced autoimmune hepatitis generally has a better prognosis than the spontaneous form. In one well-characterized study, immunosuppressive treatment was successfully stopped in 14 patients with drug-induced autoimmune hepatitis, with none of them relapsing afterward. By contrast, 65% of patients with spontaneous autoimmune hepatitis relapsed when their immunosuppression was withdrawn.¹⁹ A separate analysis echoed this finding, concluding that patients with drug-induced disease have a greater chance of having treatment suspended with a low risk of relapse, progression to cirrhosis, or need for liver transplant.²⁰

That said, some cases do not resolve simply by removing the drug. There is evidence that certain drugs may awaken a latent autoimmune tendency rather than directly causing the disease, meaning the liver inflammation persists even after the trigger is gone.²¹ These patients end up requiring long-term immunosuppressive therapy just like someone with spontaneous autoimmune hepatitis. Distinguishing between a drug that caused the disease and a drug that merely unmasked a disease that was going to happen anyway is one of the unresolved puzzles in this area.

Supplements and Vaccines as Unexpected Triggers

Prescription drugs are not the only things that can set off drug-induced autoimmune hepatitis. Dietary supplements have appeared in case reports with increasing frequency, and this is a blind spot for many patients and clinicians alike. One report described a 71-year-old woman who developed autoimmune hepatitis while taking turmeric supplements for cardiovascular health. Her condition resolved rapidly after she stopped taking them.²² Another case documented autoimmune hepatitis triggered by skullcap, a herbal supplement sometimes used for anxiety and sleep.²³

Because supplements are not regulated with the same rigor as prescription medications, contamination and mislabeling add another layer of uncertainty. When someone develops liver injury while taking a supplement, it can be difficult to determine whether the active ingredient, a contaminant, or an interaction with another substance is responsible. The takeaway is practical: if you are experiencing unexplained liver problems, your doctor needs to know about every supplement you take, not just prescription medications.

COVID-19 vaccines were also linked to autoimmune hepatitis in a small number of reports. A review summarized 27 cases of autoimmune hepatitis developing after COVID-19 vaccination, spanning different vaccine types.²⁴ A systematic review of the literature identified 32 such cases total and described the association as uncommon.²⁵ Given that billions of vaccine doses were administered worldwide, these numbers are extremely small in absolute terms. The cases are documented because they inform clinical awareness, not because they indicate a common risk.

A Growing List of Drugs Under Scrutiny

The roster of implicated medications continues to expand as pharmacovigilance systems collect more data. The comprehensive FDA database analysis categorized the 22 drugs with the strongest signals across several therapeutic classes. Beyond the antibiotics, biologics, statins, and checkpoint inhibitors already discussed, the list includes antivirals used for hepatitis C treatment (pegylated interferon and ribavirin), the multiple sclerosis drugs glatiramer acetate and alemtuzumab, the anti-inflammatory painkillers naproxen and meloxicam, the seizure medication levetiracetam, and bosentan, a drug for pulmonary arterial hypertension.⁴

A minireview published in the World Journal of Gastroenterology organized these drugs by strength of evidence, classifying minocycline, nitrofurantoin, and infliximab as having a definite association with drug-induced autoimmune hepatitis, while drugs like diclofenac, atorvastatin, rosuvastatin, and etanercept were classified as probable.²⁶ The hierarchy matters because some of these drugs are used by enormous numbers of people, and calling something a “trigger” without context can cause unnecessary alarm. A definite association with strong case series behind it, like minocycline, warrants different clinical vigilance than a possible association based on a handful of adverse-event reports.

What makes this field particularly challenging is that new biologic therapies, immunotherapies, and targeted drugs are entering the market at a rapid pace. Each one potentially introduces a novel way to perturb the immune system, and some fraction of those perturbations will inevitably affect the liver. The practical implication for anyone on long-term medication, especially the drug classes outlined above, is straightforward: routine liver function monitoring catches problems early, and unexplained fatigue, nausea, abdominal pain, or yellowing skin warrants prompt medical evaluation and a careful review of everything you are taking.