Steatosis Score on Liver Biopsy
The steatosis score is a semi-quantitative histological grading system that assesses the percentage of hepatocytes containing fat vacuoles on liver biopsy, graded on a 0-to-3 scale: Grade 0 (<5% of hepatocytes), Grade 1 (5-33%), Grade 2 (34-66%), and Grade 3 (>66% of hepatocytes with fat). 1, 2
Histological Grading System
The standardized scoring system for hepatic steatosis assessment uses the following thresholds:
- Grade 0 (None): <5% of hepatocytes containing fat vacuoles
- Grade 1 (Mild): 5-33% of hepatocytes affected
- Grade 2 (Moderate): 34-66% of hepatocytes affected
- Grade 3 (Severe): >66% of hepatocytes affected
This grading system was validated in the NASH Clinical Research Network study 2 and has been consistently applied across multiple guideline documents 1.
Clinical Context and Application
The steatosis score is one component of the NAFLD Activity Score (NAS), which combines steatosis (0-3), lobular inflammation (0-2), and hepatocellular ballooning (0-2) to provide a comprehensive assessment of disease activity 2. The NAS was specifically designed for use in clinical trials, with scores ≥5 correlating with NASH diagnosis and scores <3 indicating "not NASH."
Important Limitations
Visual semi-quantitative assessment significantly overestimates the actual fat content in hepatocytes. Research demonstrates that what pathologists grade as:
- Grade 1 (mild) corresponds to only ~2.2% actual fat content
- Grade 2 (moderate) corresponds to ~9.2% actual fat content
- Grade 3 (severe) corresponds to ~23.1% actual fat content 3
This represents a substantial overestimation compared to the stated thresholds of 5-33%, 34-66%, and >66% respectively. Stereological point counting methods provide more accurate quantification (ICC 0.99) compared to visual assessment (weighted kappa 0.87) 3.
Inter-rater Reliability
The steatosis score demonstrates good inter-rater agreement among pathologists (weighted kappa 0.79-0.84 for steatosis grading) 2, making it reasonably reproducible despite being subjective. However, this reproducibility is lower than quantitative methods and should be considered when precise measurements are critical, such as in liver transplant donor evaluation 4.
Relationship to Non-Invasive Assessment
Ultrasound grading (also 0-3 scale) correlates moderately with histological steatosis (Spearman correlation 0.80) but frequently misclassifies individual cases 1. Nearly half of patients with mild steatosis by ultrasound had moderate steatosis on biopsy, and 83% of patients with negative ultrasound still showed steatosis histologically 1.
MRI-PDFF shows stronger correlation with histological steatosis grade (Fisher's Z 0.67, AUROC 0.99 for detecting any steatosis) 5 and represents the most accurate non-invasive method for steatosis quantification.
When Biopsy is Indicated
Liver biopsy should be reserved for patients at increased risk of steatohepatitis and advanced fibrosis (F2-F3), particularly when non-invasive tests suggest significant disease or when competing etiologies need exclusion 6. The presence of metabolic syndrome and elevated NAFLD Fibrosis Score identify patients who benefit most from histological assessment 6.
For routine steatosis detection and monitoring, non-invasive methods (MRI-PDFF, controlled attenuation parameter) are preferred given biopsy's invasiveness, cost, and sampling variability 6, 5.