How do you assess for portal hypertension?

Assessment of Portal Hypertension

Hepatic venous pressure gradient (HVPG) measurement via hepatic vein catheterization remains the gold standard for diagnosing and quantifying portal hypertension, with HVPG ≥10 mm Hg defining clinically significant portal hypertension (CSPH), though noninvasive imaging with transient elastography (TE) liver stiffness measurement (LSM) is the best available alternative when invasive testing is not feasible. 1

Invasive Assessment: The Reference Standard

HVPG measurement is the definitive diagnostic test for portal hypertension: 1

• Portal hypertension is defined as HVPG >5 mm Hg 2, 1
• CSPH is defined as HVPG ≥10 mm Hg, which predicts risk of clinical decompensation and development of varices 2, 1
• HVPG ≥16 mm Hg is strongly associated with death 1
• This invasive procedure is limited by availability, technical requirements, and is impractical for routine screening 1

Noninvasive Assessment: Practical Clinical Approach

First-Line Imaging: Ultrasound with Doppler

Ultrasound with Doppler should be the initial noninvasive approach, looking for direct signs of portal hypertension: 1

• Portosystemic collaterals (recanalized umbilical vein, splenorenal shunts, gastric varices) are 100% specific for CSPH 1
• Portal vein flow reversal (hepatofugal flow) is 100% specific for CSPH 1
• Splenomegaly is sensitive but nonspecific when used alone; valuable when combined with other parameters 1

Best Noninvasive Test: Transient Elastography (TE)

When direct ultrasound signs are absent, TE with LSM is the most accurate noninvasive test for detecting CSPH: 2, 1

Algorithmic interpretation of LSM values:

• LSM ≤15 kPa: CSPH is highly unlikely (sensitivity 90-96%, NPV 89-96%) 2, 1

  • At 13.6-15 kPa cutoff: sensitivity 93-97%, specificity 32-50% 2

• LSM 15-25 kPa: Intermediate zone requiring additional assessment 1

  • At 15-18.8 kPa: sensitivity 92-100%, specificity 48-71% 2
  • At 19-21.8 kPa: sensitivity 72-89%, specificity 70-73% 2

• LSM ≥25 kPa: CSPH is likely present (specificity 82-93%, PPV 96%) 2, 1

  • At 25 kPa cutoff: sensitivity 85%, specificity 82% 2

Alternative elastography modalities:

• 2D-SWE at 14 kPa cutoff provides 91% sensitivity and 37% specificity to rule out CSPH 2
• 2D-SWE at 32 kPa cutoff provides 47% sensitivity and 89% specificity to rule in CSPH 2
• MRE-based LSM shows similar correlation to ELF score but limited comparative data exist 2

Blood-Based Tests: Limited Utility

Blood-based tests have poor accuracy and should NOT be used alone to diagnose portal hypertension: 1

• Platelet count <100,000/mcL: only 78% sensitivity for CSPH, specificity <50% 2, 1
• APRI score: 56% sensitivity, 68% specificity for CSPH 2, 1
• FIB-4 score: 54% sensitivity, 73% specificity for CSPH 2, 1
• ELF score: AUROC 0.79 for severe portal hypertension, but no significant correlation with HVPG in CSPH subgroup 2

Combined Scoring Systems: Enhanced Accuracy

Combining LSM with clinical parameters improves diagnostic accuracy:

• LSM ≤15 kPa + platelets ≥150,000/mcL: NPV 97% for excluding CSPH 2
• LSM <21.3 kPa + FIB-4 <1.85: NPV 93% for excluding CSPH 2
• LSPS (Liver stiffness-spleen diameter to platelet ratio score): 1
  • At cutoff 0.927 kPa: 99% sensitivity, 54% specificity 1
  • At cutoff 3.75 kPa: 52% sensitivity, 97% specificity 1
  • At cutoff >2.06: 90% specificity for CSPH 1

Practical Clinical Algorithm

Step 1: Perform ultrasound with Doppler to identify direct signs (collaterals, flow reversal) 1

Step 2: If direct signs absent, obtain TE-LSM 1

• LSM ≤15 kPa: CSPH unlikely, especially if platelets ≥150,000/mcL 2, 1
• LSM 15-25 kPa: Consider HVPG measurement or close monitoring 1
• LSM ≥25 kPa: CSPH likely present, proceed with variceal screening 1

Step 3: If diagnostic uncertainty remains or therapeutic decisions require precise quantification, proceed to HVPG measurement 1

Important Caveats

Limitations of LSM accuracy:

• Decompensated cirrhosis reduces LSM accuracy: 78% of false negatives with 2D-SWE <14 kPa were decompensated patients 2
• Disease etiology affects cutoff values: Most data derived from HCV patients; cutoffs may differ for alcohol-associated liver disease 2
• LSM correlates with fibrosis AND portal pressure: Cannot distinguish between the two contributors 2

Factors that can falsely elevate LSM:

• Acute hepatitis or hepatic inflammation 2
• Extrahepatic cholestasis 2
• Hepatic congestion 2
• Recent food intake 2

Blood-based tests are inadequate markers for CSPH despite their utility in detecting cirrhosis, as they show substantially lower AUROC values (0.64-0.65) compared to LSM (0.85) 2