Original article
Alimentary tract
Endoscopic Detection of Proximal Serrated Lesions and Pathologic Identification of Sessile Serrated Adenomas/Polyps Vary on the Basis of Center

https://doi.org/10.1016/j.cgh.2013.11.034Get rights and content

Background & Aims

We investigated rates of detection of proximal serrated lesions in a cohort of average-risk patients undergoing screening colonoscopies.

Methods

We reviewed results from screening colonoscopies performed by attending gastroenterologists at 32 endoscopy centers from 2008–2010. Pathology slides were interpreted at the individual centers. For this analysis, serrated lesions included hyperplastic polyps larger than 10 mm, those interpreted as sessile serrated adenomas (or sessile serrated polyp), and traditional serrated adenomas. Rates of detection for conventional adenomas and serrated lesions were compared among centers.

Results

A total of 5778 lesions were detected in 7215 screening colonoscopies. Of the 5548 lesions with pathology results, 3008 (54.2%) were conventional adenomas, 350 (6.3%) were serrated, and 232 (4.2%) were proximal serrated. The proportion of colonoscopies with at least 1 proximal serrated lesion was 2.8% (range among centers, 0%–9.8%). The number of serrated lesions per colonoscopy ranged from 0.00–0.11 (average, 0.05 ± 0.25). Overall lesion detection rates correlated with proximal serrated lesion detection rates (R = 0.91, P < .0001); conventional adenoma and proximal serrated lesion detection rates also correlated (R = .43, P = .025). The detection rate of proximal serrated lesions differed significantly among centers (P < .0001); odds ratios for detection ranged from 0–0.79. Some centers' pathologists never identified proximal serrated lesions as sessile serrated adenomas/polyps.

Conclusions

In an average-risk screening cohort, detection of proximal serrated lesions varied greatly among endoscopy centers. There was also substantial variation among pathologists in identification of sessile serrated adenomas/polyps. Nationally, a significant proportion of proximal serrated lesions may be missed during colonoscopy examination or incorrectly identified during pathology assessment. ClinicalTrials.gov Number: NCT00855348.

Section snippets

Setting

The current study represents a secondary analysis of the prospective, multicenter clinical trial evaluating methylated Septin9 for the detection of CRC in an average-risk screening population (NCT00855348 Number: ClinicalTrials.gov). All authors had access to the study data and reviewed and approved the final manuscript. The study, Prospective Evaluation of Septin 9 (PRESEPT), prospectively enrolled subjects in 22 sites in the United States and 10 in Germany.20 There were 3 academic centers in

Results

A total of 7215 screening colonoscopies were performed at 32 endoscopy centers during the course of the nearly 2-year study. There were 22 U.S. centers that recruited 76% of all subjects, and 10 German centers recruited 24% of subjects. Overall, mean age was 60.5 ± 7.6 years, and 45% of subjects were male. A total of 5778 lesions were detected in 3247 subjects, of which 3008 (52.1%) were conventional adenomas, and 350 (6.1%) were serrated lesions as defined for this study (SSA/P, SSA with

Discussion

This multicenter study shows that in a screening cohort, the proportion of subjects with at least 1 proximal serrated lesion (defined as SSA/P or a hyperplastic polyp ≥10 mm in size) was 2.8%, with a range of 0%–9.8% between centers. These findings are in agreement with prior studies demonstrating variable detection between individual examiners,18, 19 although the definition of serrated lesion in earlier studies also included hyperplastic lesions <10 mm. This is a demonstration of variable

Acknowledgments

Theo DeVos and Thomas Konig of Epigenomics, Inc, participated in statistical analysis and revision of the manuscript.

References (24)

  • S. Ogino et al.

    CpG island methylator phenotype, microsatellite instability, BRAF mutation and clinical outcome in colon cancer

    Gut

    (2009)
  • K. Nosho et al.

    A case of colorectal carcinoma in adenoma analyzed by a cDNA array

    Int J Colorectal Dis

    (2005)
  • Cited by (0)

    Conflicts of interest These authors disclose the following: S. Payne was an employee of Epigenomics when the study was initiated. T. Church's institution received from Epigenomics for the original PRESEPT study a grant, travel support, and fees for participation in review activities and payment for writing or reviewing the original PRESEPT manuscript. M. Wandell was an employee of Epigenomics when the study was initiated. D. Snover received travel support and fees for slide review in the original PRESEPT study. R. Day is a paid consultant to Epigenomics. D. Rex is a member of the Epigenomics Scientific Advisory Board. The remaining authors disclose no conflicts.

    Epigenomics initiated and conducted the PRESEPT study. The colonoscopies performed in that study are the basis for the current study. The study sponsor provided statistical support, and Theo DeVos participated in revision of the manuscript. Shannon Payne was an employee of Epigenomics, supervised the initial analyses, and wrote the first draft of the manuscript during her employment by Epigenomics.

    View full text