Colorectal cancer (CRC) remains the second leading cause of cancer death in North America. It would be wonderful if we could accurately measure a risk score for CRC based on a few simple questions, and use this score to determine most appropriate timing and type of screening. Unfortunately, I do not think the scores are sufficiently discriminant. Investigators from Poland and the U.S. have recently published and validated risk scores aimed at stratification of risk for colorectal cancer by primary care providers.1-3 The elements are similar in all three studies — including age, sex, smoking history, BMI or waist circumference, and family history of CRC. These risk scores are designed to include data that is readily available to a primary care provider. A more complex German scoring system includes nonsteroidal anti-inflammatory drugs (NSAID) use, prior colonoscopy (and polyps) and red meat consumption.4
Results of such stratification could be used to customize screening — provided that the score was sensitive enough. Kaminski et al. generated a risk scoring system and estimated the likelihood of detecting advanced neoplasia to be 1.32 percent for a score of 0 and more than 19 percent for patients with scores of 7 to 8. Imperiale et al. found the risk increased from 1.92 percent in people with a very low risk (score 0), to 24.9 percent in high-risk groups (score greater than 6).2 However, no one ever escapes risk if they live long enough, because age alone adds points every five years after age 50 (1) or 55 (2). Since most patients will have one or more points, they will likely fall into the middle range with rates of advanced neoplasia at 5 percent or more. No male is ever in the very lowest risk group, which limits the usefulness of the current risk scoring systems for men. These scores and their validation remind us that risk of advanced neoplasia and CRC is complicated, and may vary based on many factors. Not included in the U.S. and Polish models is the use of aspirin or NSAIDs, known to reduce risk of recurrent adenomas, incidence of CRC and the mortality of CRC.5 The Polish study applies to Caucasians only. The U.S. studies do not include race or ethnicity as variables in the model. Since black men and women in the U.S. may have a higher risk of advanced lesions compared to age-matched whites, it is not clear if such scoring would apply to blacks.6
I can see several ways in which a scoring system might be useful to primary care providers at the initiation of screening at age 50. First, it would help the primary care provider to identify the higher-risk patients who should definitely be urged to undergo screening. Since the high-risk groups have advanced adenoma rates of more than 15 percent or more, screening with colonoscopy would likely be preferred over other screening methods. A high score would enable counseling, and perhaps help encourage a reluctant patient to get screened. A high score, which includes tobacco use, could provide an opportunity for counseling to stop smoking.
A low score could provide patients a false sense of reassurance, and could impact rates of screening in the future.
I am concerned about the implications of a low score. None of the authors advocate “no screening” for the low-risk group, but suggest that screening by methods other than colonoscopy could be considered. This is quite reasonable. Nevertheless, a low score could provide patients a false sense of reassurance, and could impact rates of screening in the future. For example, if told they are low risk, patients may believe they are low risk for life, and might avoid screening. Such is not the case, since everyone will earn points with age and eventually move into at least the midrange for risk. Once patients reach age 55 to 60 and beyond, they will no longer be in the lowest risk group, so the score may no longer be useful for risk stratification.
What is very clear from these scoring systems and other data is that a 50-to-54-year-old white or Hispanic woman who is a nonsmoker and has no family history of CRC has a very low risk of advanced neoplasia. When compared to white men, age 50 to 54 years, average-risk white and Hispanic women do not have a similar risk of advanced neoplasia until after age 60.6 If we adopted screening strategies based on risk, we would recommend deferring screening in low-risk white and Hispanic women until after age 55. This approach might be simpler than trying to apply a score.
The published data on risk scoring systems provides little evidence that such scoring systems could be applied to surveillance after initial colonoscopy. Once patients have adenomas, they clearly have whatever genetic or lifestyle factors might predispose one to colon neoplasia. Risk factors during surveillance appear to be most closely related to the pathology at the baseline colonoscopy, irrespective of these other factors.7
It would be wonderful to have a precise form of risk stratification for CRC based on clinical criteria, but I am not sure the currently available scoring systems meet this benchmark. The colorectal risk scores do provide an opportunity to engage patients in informed decision making about CRC screening. However, even without a score, we know that male sex, advancing age, smoking and obesity place patients into a higher risk group, for whom screening should be strongly recommended. Young (50-to-54-year-old) white and Hispanic women fall into a very low risk group, for whom the benefits of screening are not so clear. Everyone else falls into some middle range of risk. I am not sure we need a scoring system to tell us this.
Dr. Lieberman serves on the scientific advisory boards of Exact, Ironwood, Given Imaging and MOTUS.
1.Kaminski M,F, Polkowski M, Kraszewska E,Rupinski M, Butruk E, Regular J. A score to estimate the likelihood of detecting advanced colorectal neoplasia at colonoscopy. Gut 2014; 63: 1112-9
2.Imperiale TF, Monahan PO, Stump TE, Glowinski EA, Ransohoff DF. Derivation and validation of a scoring
system to stratify risk fo advanced colorectal neoplasia in
asymptomatic adults. Ann Intern Med 2015; 163: 339-46
3.Shaukat A, Church TR, Shanley R, Kauff ND, O’Brient MJ, Mills GM, Jordan PA, Allen JA, Kim A, Feld AD, Zauber AG, Winawer SF. Development and validation of a clinical score for predicting risk of adenoma at screening colonoscopy. Cancer Epidemiol Biomarkers Prev 2015; 24: 913-20
4. Tao S, Hoffmeister M, Brenner H. Development and validation of a scoring system to identify individuals at high risk for advanced colorectal neoplasms who should
undergo colonoscopy screening. Clin Gastro Hep 2014; 12: 478-85.
5. Chan AT, Arber N, Burn J et al. Aspirin in the chemoprevention of colorectal neoplasia: an overview. Cancer Prev Res. 2012; 5: 164-78
6. Lieberman D, Williams JL, Holub J, Morris C, Logan J, Carney P. Race, ethnicity and gender differences in risk of colorectal neoplasia in average-risk men and women: Implications for initiation age of colorectal cancer screening. Gastroenterology 2014; 147; 351-8
7. 7. Lieberman DA, Rex DK, Winawer SJ, Giardiello FM, Johnson DA, Levin TR. Guidelines for colonoscopy surveillance after screening and polypectomy: A consensus update by the US Multi-Society Task Force on Colorectal Cancer. Gastroenterology 2012; 143: 844-857