This document provides expert recommendations for healthcare professionals on how to diagnose, manage, and understand liver injury caused by prescription drugs, herbal products, and dietary supplements (HDS). It is called a “guidance” instead of a “guideline” because there are very few high-quality randomized controlled trials on this topic; the recommendations are based on the consensus of experts reviewing available data from registries and case series.
1. What is DILI? Definition and Classification
DILI is liver injury caused by medications or supplements. It is classified into three main types based on how it happens:
| Type of DILI | Description | Key Features | Examples |
|---|---|---|---|
| Direct (Intrinsic) | Predictable injury that happens in most people if they take enough of the drug. | Dose-dependent, short latency (days), reproducible in animals. | Acetaminophen (APAP), high-dose intravenous methotrexate |
| Idiosyncratic | Unpredictable injury that occurs only in a small number of susceptible people. | Not dose-dependent, variable latency (days to years), not reproducible in animals. | Amoxicillin-clavulanate, isoniazid, nitrofurantoin |
| Indirect | Injury that occurs because the drug affects the body’s immune system, which then indirectly harms the liver. | Variable latency and manifestations, related to the drug’s biological effect. | Immune checkpoint inhibitors (ICIs), drugs that reactivate HBV (e.g., rituximab) |
2. How Common is DILI? (Epidemiology)
- Incidence: Idiosyncratic DILI is rare. The estimated annual incidence is 14-19 cases per 100,000 people in the general population. It is more common in hospitalized patients.
- Geographic Differences: The most common drugs causing DILI differ by country.
- In Western countries (US, Spain), antibiotics (especially amoxicillin-clavulanate) and HDS are top causes.
- In many Asian countries (China, Korea), HDS and traditional medicines are the leading cause.
- In India, anti-tuberculosis (TB) drugs are the most common cause.
Table: Top Causes and Outcomes of DILI in Different Countries (from Table 2)
| Country | Top Implicated Agents | % Female | % Liver-Related Death | % Liver Transplant |
|---|---|---|---|---|
| United States | HDS, Amoxicillin-clavulanate, Isoniazid | 59% | 3.0% | 3.7% |
| Spain | Amoxicillin-clavulanate, Anti-TB drugs, HDS | 48% | 2.1% | 1.5% |
| China | Traditional Chinese Medicine, Anti-TB drugs | 49% | 0.28% | 0.01% |
| India | Anti-TB drugs, HDS, Anti-epileptics | ~45% | ~17% | 0% |
3. Who is at Risk? (Risk Factors)
A person’s risk of developing DILI depends on a mix of factors:
| Risk Factor Category | Details from the Guidance |
|---|---|
| Drug Properties | Drugs given at a daily dose >50-100 mg, those that are lipophilic, and those that are extensively metabolized by the liver have a higher risk. |
| Host Factors (Age, Sex, Race) | Older age increases risk for some drugs (amoxicillin-clavulanate, isoniazid). Women are more susceptible to DILI in general and from specific drugs. Race/Ethnicity can influence both the drugs that cause DILI and the outcomes (e.g., African Americans had worse outcomes in one study). |
| Medical Comorbidities | Obesity, diabetes, and pre-existing liver disease can increase the risk and severity of DILI from certain drugs (e.g., methotrexate, anti-TB therapy). The role of alcohol and smoking is less clear. |
| Genetics | A variant in the PTPN22 gene is a general risk factor for all-cause DILI. Specific HLA alleles are strongly linked to DILI from specific drugs (see table below). |
Table: Examples of Genetic Risk Factors for Specific Drugs (from Table 3)
| Drug | Associated Genetic Marker | Increased Risk (Odds Ratio) |
|---|---|---|
| Flucloxacillin | HLA-B*57:01 | 36.6x – 79.2x |
| Amoxicillin-clavulanate | HLA-A*30:02, DRB1*15:01 | 2.3x – 6.7x |
| Green Tea Extract | HLA-B*35:01 | 6.8x |
| Isoniazid (Anti-TB) | NAT2 slow acetylator status | ~2.0x |
4. How to Diagnose DILI
DILI is a diagnosis of exclusion, meaning other common causes of liver injury must be ruled out first.
A. Definition of Clinically Significant DILI:
A patient must meet one of the following criteria:
- ALT or AST > 5x ULN (upper limit of normal) on two separate occasions, OR
- ALP > 2x ULN on two separate occasions, OR
- Total Bilirubin > 2.5 mg/dL with any elevated ALT, AST, or ALP, OR
- INR > 1.5 with any elevated ALT, AST, or ALP.
B. Steps in Diagnosis (See Figure 1 Algorithm):
- Take a detailed history: List ALL prescription drugs, over-the-counter products, and HDS used in the last 6 months, including start/stop dates.
- Discontinue the suspect agent(s).
- Calculate the “R value” to classify the pattern of injury:
- R = (ALT / ULN) / (ALP / ULN)
- Hepatocellular: R ≥ 5 (e.g., ALT very high, ALP normal)
- Mixed: 2 < R < 5
- Cholestatic: R ≤ 2 (e.g., ALP very high, ALT mildly high)
- Rule out other causes based on the injury pattern:
- For Hepatocellular/Mixed: Test for viral hepatitis (HAV, HBV, HCV, HEV), autoimmune hepatitis (AIH), ischemia, Wilson disease, etc.
- For Cholestatic: Get an ultrasound to rule out biliary obstruction, then consider Primary Biliary Cholangitis (AMA test), etc.
- Check databases like LiverTox (https://www.ncbi.nlm.nih.gov/books/NBK547852/) to see if the clinical picture matches previously reported cases for the suspect drug.
- Consider a liver biopsy if the diagnosis is unclear, the injury is severe or not improving, or autoimmune features are suspected.
5. Causality Assessment: Is This Drug Really the Cause?
Since there is no definitive test, structured methods help estimate the likelihood.
| Method | Description | Pros & Cons |
|---|---|---|
| RUCAM (Roussel Uclaf Causality Assessment Method) | A scored questionnaire based on timing, dechallenge, risk factors, and exclusion of other causes. | Most widely used structured method. Has been updated to improve usability. |
| RECAM (Revised Electronic Causality Assessment Method) | A newer, online, computerized version of RUCAM. | May be more reproducible and reliable, but needs more validation. |
| DILIN Expert Opinion | A panel of experts reviews the entire case and assigns a probability score (e.g., Definite >95%, Probable 50-74%). | Considered the best method but is not practical for routine clinical use. |
6. Special Topic: Herbal and Dietary Supplement (HDS) Injury
- Use is Very Common: Over 50% of US adults use dietary supplements.
- HDS represents ~20% of all DILI cases in the US and is the leading cause in some Asian countries.
- Diagnosis is Challenging because:
- Products can be mislabeled, contaminated, or adulterated (e.g., with hidden prescription drugs).
- They often contain multiple ingredients, making it hard to pinpoint the culprit.
- Some High-Risk Ingredients: The guidance lists specific ingredients (see Table 7 in doc), including:
- Green Tea Extract (associated with HLA-B*35:01)
- Polygonum multiflorum (He Shou Wu)
- Anabolic steroids (in bodybuilding supplements)
- Outcomes can be severe: HDS-induced liver failure is more likely to lead to death or transplant than drug-induced liver failure, possibly due to delayed recognition.
7. Management and Outcomes of DILI
| Aspect | Key Points from the Guidance |
|---|---|
| General Management | 1. Immediately stop the suspect drug. 2. Provide supportive care (e.g., anti-nausea meds, antipruritics for itching, hydration). 3. Hospitalize patients with signs of severe injury (e.g., coagulopathy, encephalopathy). |
| Specific Treatments | • N-acetylcysteine (NAC): Recommended for adults with DILI-related acute liver failure (ALF), as it can improve survival. • Corticosteroids: May be considered for severe cases with autoimmune features or hypersensitivity reactions (like DRESS syndrome). The optimal dose/duration is unknown. • Ursodeoxycholic Acid: May help with itching but is not a proven treatment for DILI itself. • Drug-specific: Examples include L-carnitine for valproate toxicity, cholestyramine for leflunomide. |
| Outcomes | • ~80% of patients recover fully. • ~10% die or require a liver transplant within 6 months. • ~10-20% develop chronic DILI (liver test abnormalities lasting >6 months). Risk is higher with cholestatic patterns. • Poor prognosis is associated with high bilirubin, high INR, low albumin, and severe necrosis on biopsy. |
8. Special Topic: Acetaminophen (APAP) Hepatotoxicity
- Mechanism: Direct (dose-dependent) hepatotoxin. The leading cause of ALF in the US.
- Diagnosis: Based on history of ingestion (>4g/day), very high ALT/AST (>1000 U/L), and low bilirubin early on. A serum APAP level is useful for acute overdose.
- Management:
- Activated charcoal if presentation is within 4 hours of ingestion.
- N-acetylcysteine (NAC) is the antidote. It is highly effective if given within 12 hours but should still be given even after 12 hours if there is signs of injury.
- Prognosis: Can be fatal, but with prompt NAC treatment, most patients recover completely.
9. Monitoring for DILI in Clinical Practice
- Patient Education: The most important step is to educate patients to recognize and report symptoms (nausea, vomiting, abdominal pain, jaundice, dark urine, itching) immediately.
- Lab Monitoring: Recommendations vary by drug. It is not required for all drugs.
- Isoniazid: Monitor for symptoms. Routine lab monitoring is not proven to prevent injury but is often used in high-risk patients.
- Methotrexate: Regular lab monitoring (every 4-12 weeks) is standard. Non-invasive tests for fibrosis (like elastography) are recommended for long-term users.
- Statins: Routine monitoring of liver enzymes is no longer recommended due to the very low risk of serious injury. Baseline testing can be considered.
- Immune Checkpoint Inhibitors (ICIs): Lab monitoring before each treatment cycle is standard. Treatment is held and steroids are started based on the grade of liver enzyme elevation.
10. Future Directions & Biomarkers
- Unmet Needs: Better diagnostic and prognostic biomarkers are needed. Current tests (ALT, AST, ALP) are not specific.
- Promising Biomarkers: The document mentions several under investigation, including:
- MicroRNA-122 and Keratin-18 (for detecting liver cell death)
- HMGB1 and Osteopontin (for predicting prognosis)
- Goal: To eventually have tests that can confirm which drug caused the injury and predict who will recover and who will need a transplant.
