Endoscopy 2010; 42(4): 320-323
DOI: 10.1055/s-0029-1244067
Editorial

© Georg Thieme Verlag KG Stuttgart · New York

Preventing colorectal cancer and cancer mortality with colonoscopy: what we know and what we don’t know

D.  K.  Rex1
  • 1Division of Gastroenterology/Hepatology, Indiana University Hospital, Indianapolis, Indiana, USA
Further Information

Publication History

Publication Date:
30 March 2010 (online)

Preview

Introduction

The goal of most colonoscopies is the prevention of incident colorectal cancer and especially death from colorectal cancer. The effectiveness of colonoscopy is now justly under attack following the publication of studies showing that colonoscopy reduces the risk of distal colorectal cancer and advanced neoplasms but has no [1] [2] or limited [3] effect on the incidence of proximal neoplasms. The reasons for no or limited proximal colon protection by colonoscopy are uncertain, but there are two broad categories of possible explanations. One category is altered biology in the proximal colon. Interval cancers in the proximal colon are more likely to be microsatellite unstable (MSI) and to bear the CpG Island Methylator Phenotype (CIMP) [4]. MSI is known to push tumors through the adenoma–carcinoma sequence rapidly, and CIMP tumors are believed to be the end product of serrated polyps, a group of lesions that are endoscopically subtle and about which little is known compared with adenomas.

A second category of explanations is technical failure, including poor preparation (which preferentially affects the right colon), incomplete cecal intubation, incomplete polypectomy, and suboptimal examination techniques that fail to detect precancerous lesions and early cancers [5]. The two categories of problems could be related, in that nonpolypoid lesions and serrated lesions are more common in the proximal colon and because both types of lesions are endoscopically subtle, so that their detection may escape a larger fraction of colonoscopists and a wider range of examination techniques. Indeed, consistent detection of flat and serrated lesions may yet be shown to require special technology. A key piece of missing information that will suggest which group of explanations underlies poor right colon protection is whether interval cancers cluster among individual examiners [6]. If clustering is present, then we should be able to raise performance and correct the problem. If clustering is not present, it suggests that either we should give up on colonoscopy for screening and return to flexible sigmoidoscopy, or that we need new technology to enhance detection during colonoscopy.

This editorial will discuss the potential causes of interval cancers, with an emphasis on where existing data suggest we should change clinical practice, where data suggest we should change terminology, and areas where more data are needed to guide us to best practice.

References

D. K. RexMD 

Department of Medicine
Division of Gastroenterology
Indiana University School of Medicine

550 N. University Boulevard UH 4100
Indianapolis
IN 46202
USA

Fax: +1-317-274-5449

Email: drex@iupui.edu