2012-02-07 21:24:50 UTC

Should We Diagnose and Discard Small Polyps?

Sept. 14, 2010

Robert D. Odze, MD, FRCPC

Robert D. Odze, MD, FRCPC

Professor and Chief, Gastrointestinal Pathology, Brigham & Women's Hospital, Boston, MA  


A provocative paper by Ignjatovic et al.1 brought to light an emerging and highly controversial subject regarding diagnosis and management of small (<10 mm) colonic polyps: that endoscopic diagnosis of small polyps is an acceptable diagnostic strategy, and that dispensing of formal histopathology could improve the efficiency of the procedure and lead to substantial savings in time and cost. Several other multicentric trials have shown that white-light and high-definition endoscopy, with or without contrast enhancement, can accurately distinguish neoplastic (tubular adenoma) from non-neoplastic (hyperplastic) polyps (HP). This has provided support for the diagnosis-and-discard strategy of small colorectal polyps.

The current debate on the diagnosis-and-discard management of polyps, without pathologic examination, refers mainly to distinction between tubular adenomas and HPs. Unfortunately, this rather simple classification of polyps is outdated in 2010. With the advent of more sophisticated molecular and pathology techniques, there has been an expansion of our knowledge regarding premalignant lesions of the GI tract.

Recent clinical, pathologic and molecular genetic studies have lead to a major reappraisal of the classification of HPs and HP-like lesions.2 With the discovery of the serrated pathway of carcinogenesis of the colon, it is now recognized that some cancers develop via a sequence of molecular changes that begin in polyps that resemble HPs histologically.3 These HP-like lesions are termed sessile serrated polyps (SSP, or sessile serrated adenoma). SSPs occur anywhere in the colorectum (most commonly in the right colon), are typically larger in size and have malignant potential. SSPs are non-dysplastic sessile lesions that contain exaggerated luminal serration, branching and distortion of crypts, and hyperproliferative characteristics. When SSPs progress, they acquire molecular abnormalities that ultimately lead to microsatellite instability (MSI) colorectal cancer.3, 4 Acquisition of MSI often correlates with the morphologic appearance of dysplasia, which can be difficult to diagnose pathologically because of its resemblance to HP. These dysplastic polyps have been termed mixed hyperplastic/adenomatous polyps, advanced hyperplastic polyp or SSP with dysplasia, the latter of which is the preferred term among pathologists. Many authorities believe that removal of these lesions may explain the protective effect of colonoscopy for right-sided colon cancer.5 Unfortunately, distinction of traditional, presumably non-premalignant, HPs from SSPs is difficult, due to significant overlap in their histologic features.2 In one study, detection and classification of SSPs by endoscopists was quite variable, indicating that many premalignant, apparently hyperplastic-appearing polyps are not being detected by individual endoscopists and pathologists.6

The key pathologic features that distinguish conventional HPs from SSPs occur in the bases of the crypts, in the deep portion of the mucosa, presumably beyond the endoscopist’s view. Pathologic distinction of HPs from SSPs shows a high degree of interobserver variability, even among experts.7, 8 In reality, pathologic distinction of a large HP from a small SSP is difficult, arbitrary and often unreliable. No objective criteria have been developed, and there are no histo-chemical or molecular methods that can help in this differential. To complicate matters further, some data suggests that HPs represent a precursor to SSP and, thus, represent lesions in the same carcinogenic pathway.3 Interestingly, a recent study examined how proximal colon polyps interpreted as HPs in 2001 would be interpreted by an expert pathologist in 2007.9 In that study, one GI pathologist changed the diagnosis from HP to SSP in 85 percent of cases, and the overall Kappa value among the three pathologists was extremely poor (K=0.16). This data has significant implications with regard to the study by Ignjatovic et al. in which important premalignant serrated lesions were not evaluated. Since 64 percent of the polyps in that study occurred proximal to the descending colon, surely the results would have been quite different if non-HP serrated polyps were taken into account.

There are other reasons why pathology is vital in the management of patients with colonic polyps. Parallel to recent advancements in our understanding of the timing, type and sequence of molecular events that lead to neoplastic transformation, is our knowledge that clonal molecular abnormalities that lead to dysregulation of cell proliferation and differentiation, and an increasing risk for subsequent neoplastic progression, may occur without morphologic characteristics of neoplasia commonly referred to as dysplasia.10 For instance, in Barrett’s esophagus and IBD, alterations in DNA content and cell-cycle control, as well as mutations in tumor suppressor genes, occur at an early stage of neoplastic transformation prior to the development of morphologic features of dysplasia. As such, these lesions would be difficult to recognize and distinguish from non-premalignant lesions, both pathologically and endoscopically. For instance, some forms of dysplasia in Barrett’s esophagus and IBD have been shown to effect only the bases of the crypts early in its development, without surface (luminal) involvement, which is only epithelium exposed to the eye of the endoscopist.11 Thus, the endoscopic appearance of the surface of the lesion would appear normal, while the unrecognized dysplastic (pre-malignant) crypts are allowed to grow unchecked.

The story is no less confusing when one critically evaluates the diagnosis and classification of conventional adenomas that develop through the chromosomal instability pathway. Much effort has been placed by clinicians on classifying polyps as small and apparently clinically insignificant (<10 mm) compared to those that are large (>10 mm), have villous structures or high-grade dysplasia, the latter referred to as advanced adenoma. Unfortunately, the definition and criteria for advanced adenoma and its components vary considerably among different studies.12 There is a distinct lack of uniformity and consistency in the literature regarding pathologic categorization of adenomas as villous, tubulovillous or tubular, and for the criteria of high-grade dysplasia. There are no uniform or universally accepted pathologic criteria to define a polyp as villous or highly dysplastic! In one reproducibility study in which 21 adenomas were evaluated by 30 pathologists in six clinical practices (general and academic), the reproducibility for determination of villous contour or high-grade dysplasia was abysmal.13 If pathologists, even those with so-called expert GI practices, cannot agree reliably on histologic variables of apparent clinical importance in terms of management and risk of cancer,13-15 how can endoscopists be expected to make this distinction by a limited observation of the mucosal surface by endoscopy? As a pathologist, it is quite surprising that strong clinical recommendations are made by major governing bodies in gastroenterology to have dysplasia diagnoses in Barrett’s esophagus and IBD confirmed by an expert GI pathologist. However, in the nationally and internationally published adenoma guidelines, there is no mention of the need for an expert GI pathology review of a presumed advanced adenoma despite the fact that expensive colonoscopy intervals and risk of cancer estimates are based on this determination!

Is it really possible that a 9 mm adenomatous polyp has different biological significance compared to a 10 mm one? Are endoscopic measurements reproducible to the 1 mm level? If adenomas less than 10 mm are considered benign and not worthy of histologic evaluation, how often do adenomas of this size harbor clinically important pathology? In a study by Butterfly et al., of 3,291 colonoscopies yielding a total of 1,933 small or diminutive adenomatous polyps, advanced histology (high-grade dysplasia or villous elements) were identified in 10.1 percent of small (5 to 10 mm) adenomas and 1.7 percent of diminutive (≤4 mm) polyps.16 Carcinoma was identified in 0.9 percent of small adenomas. Regardless of the variability in interpretation of villi and high-grade dysplasia, these numbers serve to emphasize the real situation in clinical practice. Some studies, primarily in Japan, have suggested that small carcinomas are more aggressive because they progress rapidly through the adenoma-carcinoma sequence. Japanese investigators have also noted that endoscopically flat, small or superficial colorectal carcinomas have a high frequency of submucosal invasion; these may resemble adenomas, both grossly and microscopically.17 One study of small, de novo, colorectal carcinomas showed that most (85 percent) of these tumors were present in the left colon and rectum.18 There is also a subtype of premalignant serrated polyp, referred to as traditional serrated adenoma, which shows a luminal serrated growth pattern similar to HPs, but represents a true neoplasm with malignant potential.19 Traditional serrated adenomas are more common in the left colon, particularly in the rectum, and may be difficult to distinguish from HPs grossly and microscopically. These lesions have not been taken into account in any long-term clinical studies that evaluated the consequences of the diagnosis and discard polyp strategy.

If the major advantage of this strategy would be to reduce the overall cost of colon cancer screening without negatively impacting quality of care, then the cost analysis calculations need to take into account the consequences of both missed and interval colon cancers and the burden to the medical system imposed by those mistakes. Would you want a 9 mm sessile serrated polyp with true high-grade dysplasia or superficially invasive carcinoma to remain in vivo because it endoscopically resembled an HP? If the answer to that question is no, then I would suggest that all polyps of the colon and rectum be submitted for pathologic evaluation and that clinically important lesions be evaluated by multiple expert pathologists for pathology confirmation prior to instituting surveillance guidelines.

  1. Ignjatovic A, East JE, Suzuki N, et al. Optical diagnosis of small colorectal polyps at routine colonoscopy (Detect InSpect ChAracterise Resect and Discard; DISCARD trial): a prospective cohort study. Lancet Oncol 2009;10:1171-78.
  2. Snover DC, Ahnen DJ, Burt RW, et al. Serrated polyps of the colon and rectum and serrated polyposis. In Carneiro F, Flejou JF, ed. WHO Classification of Tumours of the Digestive System. 4th Ed. Lyon: IARC Press 2010 (In Press).
  3. Huang S, O'Brien M. Hyperplastic polyps, serrated adenomas, and the serrated polyp neoplasia pathway. Am J Gastroenterol 2004;99:2242-55.
  4. Oh K, Redston M, Batts K, et al. Support for hMLH1 and MGMT silencing as a mechanism of tumorigenesis in the hyperplastic adenoma-carcinoma (serrated) carcinogenic pathway. Human Pathol 2005;36:101-11.
  5. Brenner H, Hoffmeister M, Arndt V, et al. Protection from right- and left-sided colorectal neoplasms after colonoscopy: population-based study. J Natl Cancer Inst 2010;102(2):89-95.
  6. Hetzel JT, Huang CS, Coukos JA, et al. Variation in the detection of serrated polyps in an average risk colorectal cancer screening cohort. Am J Gastroenterol 2010 (In Press).
  7. Odze RD, Batts K, Goldstein N, et al. Interobserver variability in the diagnosis of hyperplastic and serrated colonic polyps. Mod Pathol 2007;20 (suppl2):125A.
  8. Farris AB, Misdraji J, Srivastava A, et al. Sessile serrated adenoma: challenging discrimination from other serrated colonic polyps. Am J Surg Pathol 2008;32:30-35
  9. Khalid O, Radaideh S, Cummings O, et al. Reinterpretation of histology of proximal colon polyps called hyperplastic in 2001. World J Gastroenterol 2009 August 14;15(30):3767-70.
  10. Odze RD, Maley CC. Neoplasia without dysplasia: lessons from Barrett Esophagus and other tubal gut neoplasms. Arch Pathol Lab Med 2010;134:896-906.
  11. Lomo L, Blount PL, Sanchez CA, et al. Crypt Dysplasia With Surface Maturation: A Clinical, Pathologic and Molecular Study of a Barrett’s Esophagus Cohort. Am J Surg Pathol 2006;30(4):423-35.
  12. Odze RD. A balancing view: Pathologist/clinician interaction is essential. Am J Gastroenterol 2008;103(6):1331-1333.
  13. Golembeski CP, McKenna BJ, Appelman HD, et al. "Advanced" adenomas: Pathologists don't agree. Mod Pathol 2007;120(Suppl 2):115A.
  14. Rex DK, Alikhan M, Cummings O, et al. Accuracy of pathologic interpretation of colorectal polyps by general pathologists in community practice. Gastrointest Endosc 1999;50:468-74.
  15. Demers RY, Neale AV, Bodev H, et al. Pathologist agreement in the interpretation of colorectal polyps. Am J Gastroenterol 1990;85(4):417-21.
  16. Butterfly LF, Chase MP, Pohl H, et al. Prevalence of clinically important histology in small adenomas. Clin Gastroenterol and Hepatol 2006;4:343-8.
  17. Kuramoto S, Oohara T. How do colorectal cancers develop? Cancer 1989;64:950-5.
  18. Hornick JL, Farraye FA, Odze RD. Clinicopathologic and immunohistochemical study of small apparently "de novo" colorectal adenocarcinomas. Am J Surg Pathol 2007;31(2):207-15.
  19. Kim KM, Lee E, Kim Y, et al. KRAS mutation in traditional serrated adenomas from Korea herald an aggressive phenotype. Am J Surg Pathol 2010;34:667-75.

More on Colorectal Cancer

Principles of GI for the NP and PA

Aug. 10, 2018

Hear from the experts as they provide you with critical updates on treating and managing patients with a variety of GI disorders.

2018 AGA Postgraduate Course

June 2, 2018

Secure your spot for this clinically focused, multi-topic course that offers immediately applicable information. Held in conjunction with DDW®. Save $75 when you register by April 18.

Advocate CMS, close the colonoscopy loophole

March 8, 2018

AGA urges CMS to address the cost sharing problem that arises when screening colonoscopies become diagnostic for Medicare patients.