Rosemont Classification for Chronic Pancreatitis
The Rosemont classification is a consensus-based EUS (endoscopic ultrasound) diagnostic system that categorizes chronic pancreatitis findings into major and minor criteria, classifying patients as normal, indeterminate, suggestive of, or consistent with chronic pancreatitis 1.
Classification Structure
The Rosemont criteria divide EUS findings into two tiers:
Major Criteria (A Features)
- Hyperechoic foci with shadowing (indicating calcifications)
- Main pancreatic duct calculi
- Lobularity with honeycombing (contiguous lobules)
Minor Criteria (B Features)
- Cysts
- Dilated main pancreatic duct ≥3.5 mm
- Irregular main pancreatic duct contour
- Dilated side branches ≥1 mm
- Hyperechoic duct wall
- Strands (hyperechoic lines in parenchyma)
- Non-shadowing hyperechoic foci
- Lobularity with non-contiguous lobules
Diagnostic Categories
The classification assigns patients to four groups 1:
- Normal: No major or minor criteria
- Indeterminate: ≥3 minor criteria without major criteria
- Suggestive of CP: 1 major criterion + ≥3 minor criteria, OR ≥5 minor criteria alone
- Consistent with CP: ≥2 major criteria, OR 1 major + ≥3 minor criteria
Clinical Application
EUS is the preferred initial diagnostic test for suspected chronic pancreatitis after unrevealing standard workup 2. The examination should be delayed 2-6 weeks after acute pancreatitis resolution to avoid inflammatory changes that obscure findings 2.
Predictive Value
The Rosemont classification has clinical utility beyond diagnosis:
- Lithiasis in the main pancreatic duct is the strongest predictor of pancreatic exocrine insufficiency (OR 2.92) 3
- Patients classified as "consistent with" or "suggestive of" CP show 62% response to pancreatic enzyme supplementation versus 24% in normal/indeterminate groups 4
- Specific features (hyperechoic foci with shadowing, lobularity, stranding) correlate with pain improvement after enzyme therapy 4
Important Caveats
Interobserver agreement is only moderate (κ = 0.46-0.53), meaning different endosonographers may classify the same pancreas differently 5. The Rosemont system does not substantially improve agreement over older classification systems 5.
The correlation between Rosemont classification and exocrine function is weak overall (Spearman's rho = -0.02), with moderate diagnostic accuracy (AUROC 0.62) for predicting pancreatic insufficiency 3. Only main duct lithiasis reliably predicts functional impairment 3.
Approximately 27.7% of patients with <4 traditional criteria would be reclassified as "suggestive" under Rosemont, potentially identifying early disease 6. However, this increased sensitivity comes with uncertain specificity, as the classification was developed by expert consensus rather than validated against a gold standard 1.