Prevalence of nonpolypoid (flat and depressed) colorectal neoplasms in asymptomatic and symptomatic adults

JAMA. 2008 Mar 5;299(9):1027-35. doi: 10.1001/jama.299.9.1027.


Context: Colorectal cancer is the second leading cause of cancer death in the United States. Prevention has focused on the detection and removal of polypoid neoplasms. Data are limited on the significance of nonpolypoid colorectal neoplasms (NP-CRNs).

Objectives: To determine the prevalence of NP-CRNs in a veterans hospital population and to characterize their association with colorectal cancer.

Design, setting, and patients: Cross-sectional study at a veterans hospital in California with 1819 patients undergoing elective colonoscopy from July 2003 to June 2004.

Main outcome measures: Endoscopic appearance, location, size, histology, and depth of invasion of neoplasms.

Results: The overall prevalence of NP-CRNs was 9.35% (95% confidence interval [95% CI], 8.05%-10.78%; n = 170). The prevalence of NP-CRNs in the subpopulations for screening, surveillance, and symptoms was 5.84% (95% CI, 4.13%-8.00%; n = 36), 15.44% (95% CI, 12.76%-18.44%; n = 101), and 6.01% (95% CI, 4.17%-8.34%; n = 33), respectively. The overall prevalence of NP-CRNs with in situ or submucosal invasive carcinoma was 0.82% (95% CI, 0.46%-1.36%; n = 15); in the screening population, the prevalence was 0.32% (95% CI, 0.04%-1.17%; n = 2). Overall, NP-CRNs were more likely to contain carcinoma (odds ratio, 9.78; 95% CI, 3.93-24.4) than polypoid lesions, irrespective of the size. The positive size-adjusted association of NP-CRNs with in situ or submucosal invasive carcinoma was also observed in subpopulations for screening (odds ratio, 2.01; 95% CI, 0.27-15.3) and surveillance (odds ratio, 63.7; 95% CI, 9.41-431). The depressed type had the highest risk (33%). Nonpolypoid colorectal neoplasms containing carcinoma were smaller in diameter as compared with the polypoid ones (mean [SD] diameter, 15.9 [10.2] mm vs 19.2 [9.6] mm, respectively). The procedure times did not change appreciably as compared with historical controls.

Conclusion: In this group of veteran patients, NP-CRNs were relatively common lesions diagnosed during routine colonoscopy and had a greater association with carcinoma compared with polypoid neoplasms, irrespective of size.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Cohort Studies
  • Colonoscopy
  • Colorectal Neoplasms / epidemiology*
  • Colorectal Neoplasms / pathology*
  • Cross-Sectional Studies
  • Female
  • Humans
  • Male
  • Middle Aged
  • Prevalence