Endoscopy 2022; 54(S 01): S43-S44
DOI: 10.1055/s-0042-1744648
Abstracts | ESGE Days 2022
ESGE Days 2022 Oral presentations
16:30-17:30 Thursday, 28 April 2022 Club E. Post polypectomy and IBD surveillance

ENDOSCOPIST ESTIMATE OF POLYP SIZE IS DOUBLE THE ACTUAL SIZE OF NEOPLASIA WITHIN A POLYP

R. Heda
1   Tulane University School of Medicine, Internal Medicine, New Orleans, United States
,
F. Ghamsari
1   Tulane University School of Medicine, Internal Medicine, New Orleans, United States
,
J. Anderson
2   Geisel School of Medicine at Dartmouth, Department of Gastroenterology, Hanover, United States
3   White River Junction VA Medical Center, Section of Gastroenterology, White River Junction, United States
,
S. Hirschman
4   White River Junction VA Medical Center, Pathology, White River Junction, United States
,
H. Pohl
2   Geisel School of Medicine at Dartmouth, Department of Gastroenterology, Hanover, United States
3   White River Junction VA Medical Center, Section of Gastroenterology, White River Junction, United States
› Author Affiliations
› Further Information
Also available at
 

Aims Polyp size is a major determinant for post-colonoscopy surveillance. It is unknown how endoscopist size estimates (ES) compares to actual neoplasia size (NS) within a resected polyp. We aimed to compare ES with NS.

Methods This prospective study included patients undergoing elective colonoscopy with en bloc polypectomy. Each polyp had three size determinations: 1) by the endoscopist, 2) ex-vivo as the polyp specimen size, and 3) as the maximum extent of neoplasia within the polyp (pathologist). The primary outcome was the mean absolute (AD) and relative difference (RD) between ES and NS with 95% confidence interval (CI).

Results 2356 polyps from 1239 patients were included. 97% were adenomatous. Median ES was 4 mm (IQR 3-5). ES estimates were smaller than specimen sizes ([Table 1]), and both were greater than NS. ES was on average twice the NS (RD 1.95 (95% CI 1.86-2.04) with a mean AD of 1.3 mm (95% CI 1.21 – 1.37). In multivariable analysis the difference increased with size (per endoscopist) and flat morphology and was lower with trainee involvement. Of all=10 mm polyps (n=44 in 44 patients) as assessed by the endoscopist, only 23% (95% CI 11.5-37.8) (10 polyps in 10 patients) had an actual neoplasia size of at least 10 mm.

Table 1

Endoscopist size, mean mm (SD)*(ES)

Specimen size, mean mm (SD)*(SS)

Neoplasia size, mean mm (SD)*(NS)

Mean relative difference (95% CI) (ES/NS)

All polyps (N=2356)

4.0 (2.1)

5.0 (2.9)

2.7 (1.8)

1.95 (1.86–2.04)

Size groups (per endoscopist):

1-5 mm

3.4 (1.1)

4.7 (2.7)

2.4 (1.3)

1.86 (1.78–1.94)

6-9 mm

6.8 (0.95)

6.4 (3.1)

4.2 (2.5)

2.23 (2.04–2.42)

≥10 mm

13 (3.9)

9.0 (4.4)

6.4 (4.4)

4.39 (1.65–7.14)

Conclusions This is the first study to compare endoscopist polyp size estimates, polyp specimen size, and actual neoplasia size within a polyp. These findings question current practice, call for objective neoplasia size measurements, and may have implications for assignment of patient risk and surveillance interval.



Publication History

Article published online:
14 April 2022

© 2022. European Society of Gastrointestinal Endoscopy. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany