Hepatotoxicity Biomarkers
Traditional liver enzymes remain the standard for detecting hepatotoxicity, but emerging biomarkers offer earlier detection and better prognostic accuracy.
## Overview
Detecting drug-induced liver injury early and accurately is critical because delayed recognition leads to continued exposure, progressive injury, and potentially irreversible damage. Current biomarkers range from established serum enzymes to novel molecular markers that promise greater sensitivity and specificity.
## Traditional Biomarkers
**Alanine aminotransferase (ALT)** is the most widely used marker of hepatocellular injury. It is relatively liver-specific (also found in muscle) and rises within hours of hepatocyte damage. ALT >3x the upper limit of normal (ULN) is the standard threshold for potential DILI in clinical trials.
**Aspartate aminotransferase (AST)** is found in liver, heart, muscle, and red blood cells. AST/ALT ratio >2 suggests alcoholic liver disease, while ratios <1 are typical of viral or drug-induced hepatitis.
**Alkaline phosphatase (ALP)** indicates cholestatic injury (bile duct damage). Elevated ALP with normal ALT points to cholestatic DILI; elevated ALT with normal ALP indicates hepatocellular DILI. Mixed patterns occur frequently.
**Total bilirubin** elevation with hepatocellular injury is the most ominous combination (Hy's Law). Bilirubin rises later than aminotransferases because the liver has substantial functional reserve — by the time bilirubin elevates, significant hepatocyte mass has been lost.
**International Normalized Ratio (INR)** reflects synthetic function (coagulation factor production). INR prolongation in DILI indicates severe injury and poor prognosis.
## Hy's Law and Temple's Corollary
**Hy's Law** states that drug-caused hepatocellular injury (ALT >3x ULN) combined with jaundice (bilirubin >2x ULN) without biliary obstruction carries 10-50% mortality. **Temple's Corollary** posits that if a drug causes ALT >3x ULN in clinical trials, severe DILI cases will eventually emerge in larger populations.
The **eDISH** (evaluation of Drug-Induced Serious Hepatotoxicity) plot — a log-scale scatter plot of peak ALT vs. peak bilirubin for all trial participants — is now a standard FDA submission requirement.
## Emerging Biomarkers
**Glutamate dehydrogenase (GLDH)** is highly liver-specific (mitochondrial matrix enzyme). Unlike ALT, it is not elevated by muscle damage, making it superior for distinguishing hepatic from extrahepatic injury.
**MicroRNA-122 (miR-122)** is the most abundant hepatocyte microRNA. It is released earlier than ALT after injury, is more liver-specific, and correlates with histological damage severity. It is the most promising next-generation DILI biomarker.
**High Mobility Group Box 1 (HMGB1)** differentiates immune-mediated from direct hepatotoxicity. Acetylated HMGB1 indicates immune activation; total HMGB1 correlates with necrosis extent.
**Keratin-18 (K18)** fragments distinguish apoptotic (caspase-cleaved, M30) from necrotic (full-length, M65) cell death, potentially guiding prognosis and treatment decisions.
## Key Takeaways
- ALT remains the primary screening biomarker for hepatocellular DILI
- Hy's Law (ALT >3x + bilirubin >2x) predicts serious or fatal outcomes
- miR-122 is more sensitive and liver-specific than ALT and rises earlier
- GLDH eliminates false-positive ALT elevations from muscle sources
- Biomarker panels combining traditional and novel markers improve DILI detection accuracy