Risk factors for sessile serrated adenomas

J Clin Gastroenterol. 2011 Sep;45(8):694-9. doi: 10.1097/MCG.0b013e318207f3cf.

Abstract

Background: Although sessile serrated adenomas (SSAs) may represent a separate and important pathway for colorectal cancer (CRC), little is known about the risk factors for these lesions. Molecular abnormalities such as BRAF have been observed in SSA and smokers. Our hypothesis is that smoking may be associated with these lesions.

Methods: All patients diagnosed with an SSA from January 2007 to September 2010 were identified retrospectively based on a pathology database query. There were 2 sets of controls. One group had no adenomas, whereas another group had tubular adenomas. These groups were randomly identified from 2007 to 2010. Data collected included age, sex, ethnicity, height, weight, family history of CRC, diabetes mellitus, use of aspirin, statins, and calcium, and serum trigylcerides and cholesterol. We defined smokers as those patients who smoked at least 20 pack-years.

Results: We identified 90 patients with an SSA of any size, 90 patients with tubular adenomas, and 200 controls with no adenomas. Of the 90 SSAs, 42 were 6 mm or larger and 19 of them were ≥1 cm. Most of the SSAs was flat (76/90; 84.4%). After multivariate analyses, smokers with at least 20 pack-year exposure were found to have an increased risk [adjusted odds ratio (OR)=7.31; 95% confidence interval (CI), 3.92-13.63] of having any SSAs, SSAs ≥6 mm (adjusted OR=7.77; 95% CI, 3.48-17.35), and large SSAs (adjusted OR=10.20; 95% CI, 3.31-31.41) compared with nonsmokers. We also observed this relationship when comparing patients with SSAs to those with tubular adenomas.

Conclusions: Our data suggest that smoking at least 20 pack-years is strongly associated with any and large SSAs. In addition, diabetes mellitus and obesity seem to be associated with SSAs as well. Our data has implications for CRC screening.

MeSH terms

  • Adenoma / etiology*
  • Adenoma / pathology
  • Case-Control Studies
  • Chi-Square Distribution
  • Colonic Polyps / etiology*
  • Colonic Polyps / pathology
  • Colonoscopy
  • Colorectal Neoplasms / etiology*
  • Colorectal Neoplasms / pathology
  • Connecticut
  • Diabetes Complications / etiology*
  • Diabetes Complications / pathology
  • Female
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Obesity / complications*
  • Odds Ratio
  • Retrospective Studies
  • Risk Assessment
  • Risk Factors
  • Smoking / adverse effects*
  • Time Factors