What is the most commonly used tumor marker for hepatocellular carcinoma (liver cancer)?

Tumor Marker for Hepatocellular Carcinoma (Liver Cancer)

Alpha-fetoprotein (AFP) is the most commonly used tumor marker for hepatocellular carcinoma, though it has significant limitations with normal levels in up to 35% of small HCC cases. 1

Primary Tumor Marker: Alpha-Fetoprotein (AFP)

AFP remains the most widely tested and utilized biomarker for HCC worldwide, despite its suboptimal performance characteristics. 1

Performance Characteristics

• Sensitivity varies dramatically by cut-off value:

  • At 20 ng/mL: sensitivity 60%, specificity 76-97% 1, 2
  • At 200 ng/mL: sensitivity drops to only 22%, but specificity approaches 100% 1

• Critical limitation: Only 10-20% of early-stage HCC tumors present with abnormal AFP serum levels 1

• Normal AFP is found in 35-46% of HCC cases, even with large tumors 1, 2

Diagnostic Thresholds

• AFP >200 ng/mL with typical imaging features (arterial hyperenhancement with portal/delayed phase washout) allows HCC diagnosis without biopsy in cirrhotic patients 1

• AFP >400 ng/mL has nearly 100% specificity for HCC in the appropriate clinical context 1

• Rising AFP over time, even below diagnostic thresholds, is highly suspicious for HCC and warrants intensified surveillance 2

Guideline Recommendations on AFP Use

Western Guidelines (More Conservative)

AASLD, EASL-EORTC, and ESMO-ESDO do not recommend combining AFP with ultrasound for surveillance, as the 6-8% improvement in detection rate does not offset the 80% increase in false-positive results and cost. 1

These guidelines recommend ultrasound-based surveillance alone for at-risk populations. 1

Asian Guidelines and NCCN (More Inclusive)

Most Asian guidelines recommend surveillance combining ultrasound with AFP every 6 months for high-risk populations. 1

The NCCN states AFP may enhance detection when combined with ultrasound in screening at-risk individuals. 1

Additional Tumor Markers

AFP-L3 (Lens Culinaris Agglutinin-Reactive AFP Fraction)

• Japanese guidelines (JSH) recommend combined use of AFP >200 ng/mL, AFP-L3 >15%, or DCP >40 mAU/mL for HCC diagnosis 1

• AFP-L3 has been correlated with aggressive HCC molecular subtypes (S2 class, EpCAM positive) 1

• Limited validation in surveillance settings; primarily studied in diagnostic contexts 1

DCP/PIVKA-II (Des-Gamma-Carboxy Prothrombin)

• DCP at 40 mAU/mL cut-off is recommended by Japanese guidelines for screening 2

• Major limitation: DCP levels are associated with portal vein invasion and advanced tumoral stage, potentially detecting later-stage rather than early disease 1

• Not widely validated or available in Western countries 1, 2

Critical Pitfalls to Avoid

False-Positive AFP Elevations

AFP can be nonspecifically elevated in: 1, 2

• Active hepatitis B or C flares
• Active hepatocyte regeneration in cirrhosis
• Pregnancy
• Intrahepatic cholangiocarcinoma
• Colon cancer metastases
• Lymphoma and germ cell tumors

False-Negative Results

• Up to 35% of small HCC cases have normal AFP 1
• Two-thirds of HCCs <4 cm have AFP <200 ng/mL 2
• 20% of HCC patients never produce AFP, even with very large tumors 2

Practical Clinical Algorithm

For surveillance in high-risk patients (cirrhosis, chronic hepatitis B/C):

1. Ultrasound every 6 months is the cornerstone 1

2. Consider adding AFP measurement in:

   • Asian populations with hepatitis B 1, 2
   • Settings where ultrasound quality is suboptimal 1
   • Patients with hepatitis B regardless of cirrhosis status 2

3. If AFP is elevated or rising:

   • Perform dynamic contrast-enhanced CT or MRI 2
   • Look for arterial hyperenhancement with portal/delayed washout 2
   • If imaging shows typical HCC features and AFP >200 ng/mL, diagnosis can be made without biopsy 1

4. If AFP is normal but nodule detected on ultrasound:

   • Proceed to definitive imaging (CT/MRI) 2
   • Consider biopsy if imaging is atypical 2

The key takeaway: AFP should never be used alone for HCC diagnosis or surveillance due to inadequate sensitivity and specificity, but remains valuable when combined with imaging in appropriate clinical contexts. 1, 2