What is the role of the Paris classification in managing colonic polyps?

Paris Classification in Colonic Polyps

Overview and Purpose

The Paris classification should be used as standard nomenclature to describe the surface morphology of colorectal polyps, as it provides a common language for stratifying advanced pathology risk and informing removal strategy. 1

The Paris classification, proposed in 2002, is an endoscopic classification system for superficial colorectal lesion morphology, where "superficial" indicates the depth of penetration does not extend beyond the submucosa (no muscularis propria infiltration). 1

Classification System Structure

The Paris classification divides lesions into three major morphologic categories: 1

Type 0-I (Polypoid Lesions)

• 0-Ip (Pedunculated): Lesion attached by a stalk with narrow base 1
• 0-Is (Sessile): Lesion without stalk; base and top have same diameter 1

Type 0-II (Non-Polypoid Lesions)

• 0-IIa (Superficially elevated): Height <2.5 mm above normal mucosa 1
• 0-IIb (Flat): No protrusion above mucosa 1
• 0-IIc (Slightly depressed): Base lower than normal mucosa 1

Type 0-III (Excavated)

• Rarely seen in the colon 1

Clinical Significance for Risk Stratification

Submucosal Invasion Risk by Morphology

Depressed (0-IIc) lesions carry the highest malignancy risk and require heightened vigilance: 1

• Overall submucosal invasion risk: 27%-35.9% 1
• Small (6-10 mm) depressed lesions: >40% contain submucosal invasive cancer 1
• Large (>20 mm) depressed lesions: virtually all have submucosal invasion 1

Flat elevated (0-IIa) lesions have substantially lower risk: 1

• Submucosal invasion risk: 0.7%-2.4% 1

Pedunculated (0-Ip) malignant polyps at T1 stage are considered cured with polypectomy if: 2

• Complete resection achieved
• Well or moderately differentiated
• Resection margin tumor-free
• No lymphovascular invasion

Sessile (0-Is) malignant polyps at T1 are generally not considered cured with polypectomy alone. 2

Integration with Laterally Spreading Tumor (LST) Classification

For non-pedunculated adenomatous lesions (Paris 0-II and 0-Is) ≥10 mm, surface morphology should also be described as granular or non-granular LST. 1

LST-Granular (LST-G) Subtypes

• LST-G with homogeneous even-sized nodules: Lowest submucosal invasion risk (<2%), can grow to very large diameters 1
• LST-G with mixed-sized nodules: Higher invasion risk (7.1% for <20 mm; 38% for >20 mm), with invasion typically under largest nodule 1

LST-Non-Granular (LST-NG) Subtypes

• LST-NG pseudodepressed: Highest invasion risk (27.8% for 10-19 mm; 41.4% for 20-29 mm) 1
• LST-NG flat elevated: Moderate invasion risk (6.4% for 10-19 mm; 10.4% for 20-29 mm) 1
• LST-NG lesions often have submucosal fibrosis making standard EMR technically challenging 1

Documentation Requirements

Photo documentation of all lesions ≥10 mm before removal is strongly recommended, with suggested documentation of post-resection defect. 1

Endoscopic descriptors must be documented in the procedure report, including: 1

• Location
• Size in millimeters
• Morphology using Paris classification

Important Caveats and Limitations

Interobserver Agreement Issues

Interobserver agreement of Paris classification among expert endoscopists is only moderate (kappa=0.42). 1, 3 This represents a significant limitation for both research and routine practice, though the classification remains the most widely used international standard. 1

Practical Application Challenges

• Despite moderate interobserver variability, the Paris classification provides the essential first step in stratifying advanced pathology 1
• The classification informs removal strategy by identifying high-risk morphologies requiring en bloc resection or surgical referral 1
• Combined Paris classification with rectosigmoid location and non-granular surface morphology significantly increases covert malignancy risk 4

Quality of Evidence

The recommendation to use Paris classification is conditional with low-quality evidence, reflecting the moderate interobserver agreement and limited prospective validation studies. 1 However, it remains the standard nomenclature due to lack of superior alternatives and widespread international adoption. 1