Interobserver agreement in measuring submucosal invasion depth in pT1b-SM esophageal squamous cell carcinoma: practical implications for post-endoscopic resection treatment decision.
Accurate assessment of submucosal invasion depth in pT1b-SM esophageal squamous cell carcinoma (ESCC) is critical to determine the need for additional treatment after endoscopic resection (ER). However, the reliability of submucosal invasion depth measurement has not been sufficiently validated. This study aimed to evaluate interobserver agreement in measuring invasion depth and to identify a reliable and reproducible assessment method.
Thirty-four pT1b-SM ESCC specimens obtained by ER were independently assessed by three pathologists. Invasion depth was measured using three methods on hematoxylin and eosin (H&E)- and desmin-stained slides: Method 1, from an imaginary line representing the original level of the muscularis mucosae; Method 2, from a reconstructed trajectory based on disorganized muscularis mucosae; and Method 3, from a line connecting the lowest visible ends of the disrupted muscularis mucosae, with each method measuring perpendicular to the deepest point of invasion. The lesions were classified as pT1a-MM, pT1b-SM1 (≤200 μm), or pT1b-SM2 (>200 μm). Interobserver agreement was assessed using Fleiss' κ and intraclass correlation coefficients (ICC).
Moderate to substantial agreement was observed across all methods. Method 3 (κ = 0.678, ICC = 0.789) achieved the highest agreement on H&E-stained slides, whereas Method 2 (κ = 0.615, ICC = 0.801) on desmin-stained slides demonstrated comparable reliability.
A stepwise approach using Method 3 on H&E-stained slides as a first-line method and Method 2 on desmin-stained slides as a complementary option when the muscularis mucosae is disrupted or ambiguous demonstrated high interobserver agreement and may enhance reproducibility in routine pathological practice.
Thirty-four pT1b-SM ESCC specimens obtained by ER were independently assessed by three pathologists. Invasion depth was measured using three methods on hematoxylin and eosin (H&E)- and desmin-stained slides: Method 1, from an imaginary line representing the original level of the muscularis mucosae; Method 2, from a reconstructed trajectory based on disorganized muscularis mucosae; and Method 3, from a line connecting the lowest visible ends of the disrupted muscularis mucosae, with each method measuring perpendicular to the deepest point of invasion. The lesions were classified as pT1a-MM, pT1b-SM1 (≤200 μm), or pT1b-SM2 (>200 μm). Interobserver agreement was assessed using Fleiss' κ and intraclass correlation coefficients (ICC).
Moderate to substantial agreement was observed across all methods. Method 3 (κ = 0.678, ICC = 0.789) achieved the highest agreement on H&E-stained slides, whereas Method 2 (κ = 0.615, ICC = 0.801) on desmin-stained slides demonstrated comparable reliability.
A stepwise approach using Method 3 on H&E-stained slides as a first-line method and Method 2 on desmin-stained slides as a complementary option when the muscularis mucosae is disrupted or ambiguous demonstrated high interobserver agreement and may enhance reproducibility in routine pathological practice.
Authors
Tateishi Tateishi, Nakano Nakano, Fujishima Fujishima, Nakanishi Nakanishi, Takamatsu Takamatsu, Fujii Fujii, Kushima Kushima, Sakashita Sakashita, Watanabe Watanabe, Fujiwara Fujiwara, Nimura Nimura, Nemoto Nemoto, Yao Yao, Kawachi Kawachi
View on Pubmed