| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIALS |
From Gastrointestinal Medicine and Nutrition, University of Texas M. D. Anderson Cancer Center, Houston, Texas.
Correspondence: Address correspondence to: Patrick M. Lynch, JD, MD, Gastrointestinal Medicine and Nutrition, Box 436, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030; Fax: 713-794-5082; E-mail: plynch{at}mdanderson.org
In this issue of the Annals of Surgical Oncology, Syrigos et al.1 add to a growing body of literature in support of a role for total colonoscopy (TC) as a screening measure for individuals with a positive family history of colorectal cancer.
The investigators retrospectively reviewed the yield of TC in 249 asymptomatic individuals who underwent TC solely on the basis of having one or two first-degree relatives (FDR’s) with colorectal cancer. Although no invasive cancers were found, the overall yield of polyps were more than 20%, with about one-half of these consisting of adenomas and the remainder largely hyperplastic polyps. Approximately one-fourth of the adenomas were proximal to the splenic flexure and thus, presumably would not have been detected by flexible sigmoidoscopy and/or fecal occult blood testing (FOBT) (small polyps do not bleed much). A substantial proportion of young subjects had polyps (12% of those age 30–39 and 21% of those age 40–49). As in other studies, most hyperplastic polyps were located in the distal colon. In this study, these distal polyps seemed to predict the presence of proximal adenomas.
Should screening colonoscopy for those with positive family history be considered "standard of care"? I would argue that the answer to this is yes but with qualifications. This editorial will discuss some of the key questions that the clinician may want to consider in adopting a position on family history-based colonoscopy screening and applying it to patients in his or her practice.
Is colonoscopy the only appropriate tool for screening high-risk subjects? The current American Cancer Society,2,3 National Comprehensive Cancer Center Network,4 and American Gastroenterology Association5 guidelines basically suggest that colonoscopy is the preferred tool. As shown by Syrigos et al. and others, subjects with a positive family history have a high prevalence of lesions beyond the reach of the flexible sigmoidoscope. FOBT also has disappointing sensitivity for adenomas. Barium enema continues to be an unpopular screening tool, even though it is reasonably sensitive for detection of clinically relevant polyps. However, in one important head-to-head prospective comparison of colonoscopy and barium enema in subjects with previous adenomas, colonoscopy was more sensitive for adenomas greater than 1 cm.6
Two new methodologies, virtual colonoscopy (also known as computed tomography (CT) colonography) and stool-based testing for mutated DNA, are generating considerable interest. Virtual colonoscopy entails the use of spiral CT with very thin cuts (else small lesions would be missed) and new computer techniques for 3-dimensional reconstruction. Currently, standard bowel preps are required, and elimination of all effluent is important. Efforts are underway to develop oral contrast agents that mix with enteral contents so that they may be subtracted from images, eliminating the bowel prep, which many patients find more noxious than the colonoscopy itself. Several single institutional trials have yielded virtual colonoscopy sensitivities of >90% for detection of adenomas of more than 1 cm.7,8 However, a recent multicenter trial (P. Cotton et al., unpublished data, 2002) was unable to match these results, with sensitivity for 1 cm adenomas of <50%. Screening with CT colonography may be further compromised if, as suggested by Syrigos et al. and others, adenomas in family history positive (FH+) subjects are dysplastic out of proportion to their size.
It is known that in the course of the adenoma-carcinoma sequence in the colon, a variety of genetic mutations are acquired in key regulatory genes and their targets. Exfoliated colonic epithelium can be retrieved, its DNA extracted and amplified, and subjected to mutation analysis. Ahlquist et al.9 have reported encouraging data on the use of a panel of markers, including adenomatous polyposis coli, p53, Ras, and BAT-26. Among 61 subjects with colorectal cancer, adenoma, or normals, sensitivity for detection of cancer was 91% and 82% for adenomas >1 cm, with a specificity of 93%.
So which test is appropriate for those at increased risk? FOBT? No. Flexible sigmoidoscopy? No. Barium enema? Only if colonoscopy cannot be carried out for some reason. Virtual Colonoscopy? Maybe someday soon, if better validated. Mutation analysis in stool? Maybe someday soon if better validated. Colonoscopy? Yes.
Is everyone with one or two colorectal cancer-affected FDR’s at about the same risk, or is some stratification possible? If there are differences in risk, are they important enough to investigate? Or is colonoscopy so universally attractive that even crude estimates of increased risk are sufficient to warrant the exam? Many studies (cited by Syrigos et al.) have shown that there is an increased relative risk of colorectal cancer and of adenomas in first-degree relatives of colorectal cancer patients. This risk increases as age at onset of cancer in the relative in question decreases. This is the basis for the common recommendation that screening should begin at age 40 or 10 years younger than the youngest affected individual in the family. However, it is also recognized that there is tremendous heterogeneity within the group of subjects with a positive family history. At one extreme are cases in which only one FDR has colorectal cancer, typically but not always at an early age, and hereditary nonpolyposis colorectal cancer (HNPCC) is demonstrated by mutation analysis. Indeed population studies have shown this to be so. Aaltonen et al.10 performed microsatellite instability testing on 509 unselected Finnish patients with colorectal cancer. Microsatellite instability testing, a marker of HNPCC risk, was found in 12%. Of these, 16% were found to have pathologic mismatch repair gene mutations for HNPCC. Obviously, close relatives of such individuals are at a 50% risk of being mismatch repair gene mutation carriers themselves, with an attendant 80% lifetime risk of colorectal cancer. Any large series of "family history positive, not otherwise specified" subjects will include a small number, perhaps 2% to 3% who actually have HNPCC. Another small group will have attenuated familial adenomatous polyposis, with a sufficiently late onset, often after age 50, and modest enough adenoma burden, as to escape being labeled with familial adenomatous polyposis.
At the other end of the extreme will be a much larger group of patients, typically with later onset disease, the main cause of which will have been environmental (dietary). In this latter group, such dietary risk will be modulated by a more modest genetic modifier of risk. Considerable effort is being devoted to the study of genetic polymorphisms that account for variability in hepatic xenobiotic metabolizing activity. At present, such genetic modifiers of essentially ubiquitous dietary carcinogens probably constitute too weak an influence to be of immediate clinical usefulness in risk stratification.
Assuming colonoscopy is to be carried out, should follow-up of those with adenomas and FH+ be more aggressive than for those at "average risk" who for one reason or another are found to have an adenoma? Likewise, if a baseline exam is negative, should a repeat exam be done at a shorter interval than for those undergoing average risk screening?
Keeping in mind that FH+ subjects represent a truly heterogeneous group, one must consider the range of follow-up strategies. If an average risk subject is found to have an adenoma, he or she is, from that point on, considered at increased risk of developing recurrent adenomas. The National Polyp Study11 concluded that for subjects with small, solitary adenomas without severe dysplasia, repeat colonoscopy can safely be put off for 3 years. If such a follow-up exam is normal, repeat exams at intervals of 3 to 5 years are accepted. Conversely, subjects with HNPCC and who are found to have adenomas are followed very aggressively, at intervals of 1 to 2 years and with no lengthening of the interval as negative exams ensue. This is because adenomas in HNPCC are often flat, subtle, and dysplastic out of proportion to their size, indicating an accelerated adenoma-carcinoma sequence. For those at moderately increased adenoma risk because of their family history, there are no good data regarding the long-term yield of more, as opposed to less, aggressive follow-up.
The study reported here by Syrigos et al. is representative of studies that provide information about one-time screening. It would be of considerable interest to know the long-term fate of subjects from this study following their baseline colonoscopy. What follow-up interval was recommended for those with and without adenomas? What will be the patient adherence to such recommendations? Most importantly, as a guide to other investigators and clinicians, what will be the yield when follow-up colonoscopies are performed?
Received for publication April 22, 2002. Accepted for publication April 22, 2002.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
