Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults: A Guidance Statement From the American College of Physicians (Version 2)FREE
Submit a Comment
Contributors must reveal any conflict of interest. Comments are moderated. Please see our information for authorsregarding comments on an Annals publication.
Abstract
Description:
The purpose of this updated guidance statement is to guide clinicians on screening for colorectal cancer (CRC) in asymptomatic average-risk adults. The intended audience is all clinicians. The population is asymptomatic adults at average risk for CRC.
Methods:
This updated guidance statement was developed using recently published and critically appraised clinical guidelines from national guideline developers since the publication of the American College of Physicians’ 2019 guidance statement, “Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults.” The authors searched for national guidelines from the United States and other countries published in English using PubMed and the Guidelines International Network library from 1 January 2018 to 24 April 2023. The authors also searched for updates of guidelines included in the first version of our guidance statement. The Appraisal of Guidelines for Research and Evaluation II (AGREE II) instrument was used to assess the quality of eligible guidelines. Two guidelines were selected for adoption and adaptation by raters on the basis of the highest average overall AGREE II quality scores. The evidence reviews and modeling studies for these 2 guidelines were also used to synthesize the evidence of diagnostic test accuracy, effectiveness, and harms of CRC screening interventions and to develop our guidance statements.
Guidance Statement 1:
Clinicians should start screening for colorectal cancer in asymptomatic average-risk adults at age 50 years.
Guidance Statement 2:
Clinicians should consider not screening asymptomatic average-risk adults between the ages of 45 to 49 years. Clinicians should discuss the uncertainty around benefits and harms of screening in this population.
Guidance Statement 3:
Clinicians should stop screening for colorectal cancer in asymptomatic average-risk adults older than 75 years or in asymptomatic average-risk adults with a life expectancy of 10 years or less.
Guidance Statement 4a:
Clinicians should select a screening test for colorectal cancer in consultation with their patient based on a discussion of benefits, harms, costs, availability, frequency, and patient values and preferences.
Guidance Statement 4b:
Clinicians should select among a fecal immunochemical or high-sensitivity guaiac fecal occult blood test every 2 years, colonoscopy every 10 years, or flexible sigmoidoscopy every 10 years plus a fecal immunochemical test every 2 years as a screening test for colorectal cancer.
Guidance Statement 4c:
Clinicians should not use stool DNA, computed tomography colonography, capsule endoscopy, urine, or serum screening tests for colorectal cancer.
Colorectal cancer (CRC) is the fourth highest in incidence (153 020) and second in mortality (52 550) among cancer types in the United States (1). Between 2000 and 2019, CRC incidence slightly increased in persons younger than 50 years (6.0 to 8.7 per 100 000), decreased in those aged 50 to 64 years (85 to 74 per 100 000), and more sharply decreased in persons aged 65 years or older (305 to 158 per 100 000); decreases may be attributable to screening (2). Incidence of CRC varies by biological sex and race and ethnicity, with males and non-Hispanic American Indian or Alaska Native persons and non-Hispanic Black persons having the highest rates; however, absolute differences between biological sex and racial and ethnic groups are small (2).
The success of any screening program depends on screening strategy adherence (that is, type and frequency of a test, follow-up testing of abnormal results, and treatment). Benefits accrue from identification and removal of precancerous lesions or localized cancer that may progress and lead to morbidity and mortality; harms include false-positive results, physical and psychological harms, overdiagnosis, overtreatment, and financial and opportunity costs (3). Commonly used screening interventions include stool (fecal immunochemical tests [FIT], guaiac fecal occult blood tests [gFOBT], and stool DNA [sDNA] tests) and direct visualization tests (colonoscopy, flexible sigmoidoscopy [FS], and computed tomography colonography [CTC]).
Several clinical guidelines address CRC screening and vary on age to start and stop screening, screening tests and time intervals, and strength of recommendations. This guidance statement is an update of the American College of Physicians’ (ACP) 2019 guidance statement, “Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults” (4).
Scope, Population, and Intended Audience
The goal of this ACP guidance statement is to guide clinicians on age to start and stop screening and selection of type and frequency of screening tests in asymptomatic average-risk adults. This guidance is based on a critical review of existing guidelines and associated evidence reviews and modeling studies. Average risk for CRC is defined as no prior diagnosis of CRC, adenomatous polyps, or inflammatory bowel disease, and no personal diagnosis or family history of known genetic disorders that predispose a person to a high lifetime risk for CRC (for example, Lynch syndrome) (5, 6).
Methods
The methods article by ACP’s Clinical Guidelines Committee describes the development process for ACP guidance statements, which differs from that for ACP clinical guidelines (7).
Search Results
A search of databases yielded 5 guidelines from the following organizations that met inclusion criteria (Appendix Figure): American Cancer Society (ACS), American College of Gastroenterology, American College of Radiology, U.S. Multi-Society Task Force on Colorectal Cancer, and U.S. Preventive Services Task Force (USPSTF).
Appendix Figure. Study flow diagram.
CRC = colorectal cancer.
Clinical Guidelines Quality Ratings
Guidelines by ACS and USPSTF were rated by most raters as “recommended with modifications” (4.6 of 7 and 5.2 of 7, respectively) and were used to develop our guidance statements. Three of the 5 raters recommended ACS recommendations with modifications, and 2 would not recommend them. All 5 raters recommended USPSTF’s recommendations with modifications (Supplement Table 1).
The ACS used the 2016 USPSTF evidence review and decision modeling to develop its guideline (8). As a result, evidence used in this guidance statement is from USPSTF’s 2021 evidence review (9, 10) and decision modeling (11). When synthesizing the evidence in this manuscript, we note how many new studies were identified by USPSTF’s 2021 evidence review since its 2016 version. The decision modeling for USPSTF (11) was developed by the Cancer Intervention and Surveillance Modeling Network and consisted of 3 independently developed microsimulation models: Simulation Model of CRC (SimCRC), CRC Simulated Population Model for Incidence and Natural History (CRC-SPIN), and Microsimulation Screening Analysis (MISCAN). For benefits and harms outcomes, we abstracted and interpreted means of these 3 models (11). Unless stated otherwise, we refer to these 3 models as the modeling study. Additional methods for this update can be found in the Supplement.
Recommendations from eligible guidelines are displayed in Supplement Tables 2 and 3. Supplement Tables 4 to 6 display the evidence for diagnostic test accuracy, effectiveness, and harms of screening tests for CRC, respectively. The Figure summarizes our updated guidance statement.
Figure. Visual summary of the ACP guidance statement on CRC screening (version 2).
ACP = American College of Physicians; CRC = colorectal cancer; CTC = computed tomography colonography; FIT = fecal immunochemical test; FS = flexible sigmoidoscopy; gFOBT = guaiac fecal occult blood test; NPO = nothing by mouth; sDNA = stool DNA.
Age to Start Screening for CRC
The evidence review did not identify studies that enrolled or reported results for initiating CRC screening in only adults younger than 50 years (9, 10). Evidence from 2 randomized controlled trials (RCTs; no new RCTs) assessed the effectiveness of gFOBT on CRC mortality by age (12, 13). Persons aged 60 years or older had larger reductions in CRC mortality compared with those younger than 60 years (12, 13). Evidence from 3 RCTs (no new RCTs, but additional follow-up data) of FS reported results by age (14–16). The NORCCAP (Norwegian Colorectal Cancer Prevention) (50 to 54 vs. 55 to 64 years) and UKFSST (UK Flexible Sigmoidoscopy Screening) trials (55 to 59 vs. 60 to 64 years) found no differences between age groups for CRC mortality (14, 15). However, the PLCO (Prostate, Lung, Colorectal, and Ovarian) Cancer Screening trial found a greater reduction in CRC mortality in those aged 65 to 74 years versus those aged 55 to 64 years (16). Two RCTs that evaluated FS by age found no differences in all-cause mortality (14, 15).
The modeling study evaluated FIT, sDNA, colonoscopy, CTC, and FS for starting screening at age 45 years versus 50 years (11). Starting screening at age 45 years compared with 50 years yielded more life-years gained (LYG; range: 22 to 27 per 1000 screened or 8 to 10 life-days gained per person) and prevented a small number of CRC cases (range, 2 to 3 per 1000 screened) and deaths (range, 0.9 to 1 per 1000 screened). However, there was also an increase in the number of colonoscopies (range, 161 to 784 per 1000 screened) and colonoscopy complications, such as cardiovascular and gastrointestinal events (for example, serious bleeding, perforation, myocardial infarction, and angina) (range, 0.1 to 2 more per 1000 screened) (11).
Age to Stop Screening for CRC
The evidence review did not identify studies that enrolled or reported results for CRC screening in only adults older than 75 years (9, 10).
With perfect adherence to screening, the modeling study found that stopping screening at ages 80 and 85 years, compared with 75 years, had no to little additional LYG or CRC incidence and mortality prevented but conferred an increase in colonoscopies and a slight increase in colonoscopy complications (11).
Screening Tests for CRC
The updated evidence review by USPSTF found no new screening trials evaluating the effectiveness of stool or direct visualization tests. However, it identified new publications with longer follow-up periods evaluating gFOBT and FS (9, 10).
Stool Tests
gFOBT
Diagnostic Accuracy for CRC.
Effectiveness.
The evidence review identified 5 RCTs (no new RCTs, but additional follow-up data) evaluating the effectiveness of gFOBT compared with no screening (9, 10). Five RCTs, with a range of 11- to 30-year follow-up, found that biennial gFOBT screening reduced CRC mortality at approximately 20 years (risk ratio [RR], 0.91 [95% CI, 0.84 to 0.98]) and 30 years (RR, 0.78 [CI, 0.65 to 0.93]) (12, 13, 17–19). One RCT evaluated annual gFOBT and found similar reductions in CRC mortality at 30 years (13). Four RCTs that evaluated gFOBT found no differences in all-cause mortality (12, 13, 17, 18).
Harms.
There are no known direct serious harms because gFOBT is noninvasive. The evidence review found harms of colonoscopy after an abnormal gFOBT result (Supplement Table 6) (9, 10).
FIT
There are numerous FIT available, and diagnostic accuracy varies by type. The applicability of benefits and harms should be considered with that context.
Diagnostic Accuracy for CRC.
Effectiveness.
Harms.
Fecal immunochemical tests are noninvasive and have no known serious harms. The evidence review found harms of colonoscopy after an abnormal FIT result (Supplement Table 6) (9, 10).
sDNA Tests
Diagnostic Accuracy for CRC.
Effectiveness.
Harms.
Stool DNA tests are noninvasive and have no known serious harms. The evidence review identified no studies evaluating harms of follow-up colonoscopy after an abnormal sDNA test result (9, 10). However, the false-positive rate of sDNA tests to detect CRC is higher than for gFOBT and FIT, which can lead to more colonoscopies, evaluations, and harms.
Direct Visualization Tests
Colonoscopy
Diagnostic Accuracy for CRC.
Studies evaluating the sensitivity of colonoscopy for CRC were underpowered and found a wide range (0.18 to 1.0) (9, 10). Sensitivity was much higher with a narrower range for adenomas 6 mm or greater (range, 0.75 to 0.93) and 10 mm or greater (range, 0.89 to 0.95) (9, 10). Because of the limited data, the modeling study assumed a sensitivity for CRC of 0.95 (11) and used a specificity of 0.86 (21).
Effectiveness.
The evidence review identified 2 NRSs (1 new NRS) evaluating the effectiveness of screening colonoscopy (9, 10). One NRS compared those who had at least 1 colonoscopy with those who had never received one and found that colonoscopy was associated with lower CRC mortality at 24 years of follow-up (adjusted hazard ratio, 0.32 [CI, 0.24 to 0.45]) (22). Both NRSs found colonoscopy reduced CRC incidence, but neither evaluated all-cause mortality (22, 23).
Harms.
The evidence review identified 27 NRSs (6 new NRSs) for screening colonoscopy and reported 3.1 perforations per 10 000 procedures (CI, 2.3 to 4.0) (9, 10). The review also found 22 NRSs (7 new NRSs) and reported 14.6 serious bleeding events per 10 000 procedures (CI, 9.4 to 19.9) (9, 10). One NRS suggested a correlation between serious bleeding events and increasing age at screening (24). The evidence review found that perforation and risk for bleeding was slightly higher in mixed populations (that is, screening and diagnostic) (9, 10).
CTC
Diagnostic Accuracy for CRC.
Effectiveness.
Harms.
The evidence review identified 17 NRSs (1 new NRS) finding no to little risk for serious harms of CTC (9, 10). Seven NRSs (1 new NRS) found 1.3 perforations per 10 000 procedures (CI, 0.0 to 2.9) (9, 10). The largest (n = 40 121) of the 7 NRSs reported 7 perforations, all of which were asymptomatic and in patients undergoing manual insufflation (26). Four NRSs (24, 27–29) reported on CTC and serious bleeding events with only 1 NRS (n = 1384) observing 4 events (24). The review identified 27 NRSs (6 new NRSs) and showed that CTC detected a wide range of extracolonic findings that were deemed either potentially important and requiring follow-up (3.4% to 26.9%) or likely important or incompletely characterized and possibly requiring follow-up (1.3% to 11.4%) (9, 10). Whether extracolonic findings result in benefit or harm is uncertain. Data were inadequate to estimate serious harms from follow-up colonoscopy after an abnormal CTC test result (9, 10). Computed tomography colonography exposes patients to low-dose ionizing radiation, ranging from 0.8 to 5.3 mSv per examination (9, 10).
FS
Diagnostic Accuracy for CRC.
Effectiveness.
The evidence review identified 4 RCTs (no new RCTs, but additional follow-up data) evaluating the effectiveness of FS and colonoscopy referral if high-risk colonic lesions or CRC were detected (9, 10). The RCTs had 11 to 17 years of follow-up; only 1 RCT had 2 rounds of FS screening (54% attendance in the second screen) (16), whereas the other trials had just 1 round (14, 15, 31). Flexible sigmoidoscopy reduced CRC incidence and mortality (incidence rate ratio, 0.78 [CI, 0.74 to 0.83] and 0.74 [CI, 0.68 to 0.80], respectively); however, there was no difference in all-cause mortality (9, 10).
Harms.
The evidence review identified 18 NRSs (4 new NRSs) evaluating serious FS harms. Among 10 NRSs (1 new NRS), the rate of serious bleeding events was 0.5 per 10 000 procedures (CI, 0 to 1.3). In 11 NRSs, perforations occurred at a rate of 0.2 per 10 000 procedures (CI, 0.1 to 0.4) (9, 10). Across 4 NRSs (2 new NRSs), a colonoscopy after an abnormal FS resulted in 20.7 serious bleeding events per 10 000 procedures (CI, 8.2 to 33.2) and 12 perforations per 10 000 procedures (CI, 7.5 to 16.5) (9, 10).
Other Screening Tests for CRC
The USPSTF also reviewed capsule endoscopy, serum, and urine tests (9, 10). A serum test is U.S. Food and Drug Administration–approved for CRC screening in persons who decline recommended screening tests (32). The evidence review found no studies on the effectiveness of capsule endoscopy, serum, or urine screening tests; 1 NRS of harms of screening capsule endoscopy; and 1 to 2 NRSs of each of these screening tests for diagnostic accuracy, mostly focused on adenomas (9, 10). One NRS (n = 689), designed as a diagnostic accuracy study, reported capsule endoscopy harms and found zero serious adverse events and 3 nonserious adverse events (9, 10).
Comparison of Different Screening Strategies
The evidence review identified 20 RCTs and 1 NRS (6 new RCTs) comparing screening tests for detecting CRC; none evaluated sDNA tests (9, 10). These comparative studies generally lacked statistical power to detect differences between groups and were limited to 1 round of CRC screening (9, 10).
The evidence review identified 5 RCTs comparing direct visualization screening tests and found no differences in detected CRC cases (9, 10). The review also identified 11 RCTs comparing stool and direct visualization tests. One-time testing of direct visualization identified more CRC cases than 1-time stool tests (9, 10), but 1 RCT comparing 4 rounds of FIT to 1-time colonoscopy or FIT showed no difference in detecting CRC (31).
Costs of Screening Tests
Supplement Table 7 summarizes the costs of various screening tests and strategies in the United States. Neither ACS nor USPSTF considered cost or resource use in its recommendations (6, 8). Screening more frequently than recommended is unlikely to provide additional meaningful benefit (3). However, it will increase false-positive results, harms, and burden while using already limited health care resources. Additional issues with cancer screening in general include overdiagnosis (a condition or disease that would not cause symptoms or death during a person’s lifetime) and associated overtreatment (unnecessary treatment) (3).
Multiple Chronic Conditions
Comorbidities reduce age-adjusted life expectancy and may influence CRC screening initiation, discontinuation, and frequency. Serious comorbidities include but are not limited to chronic obstructive pulmonary disease, diabetes, heart failure, moderate to severe liver disease, chronic hepatitis, advanced chronic kidney disease or end-stage kidney disease, and dementia (33). Due to the slow growth rate of most adenomas and subsequent CRC should it develop, the time needed to derive a benefit from screening is long (it takes 10 years to reduce 1 CRC death per 1000 persons screened) (34). The evidence review and modeling study suggest that any benefit from reducing mortality is outweighed by harms for patients with a life expectancy of less than 10 years, and possibly longer, due to age or comorbidities (9–11).
Differences by Race and Ethnicity
It is important to note that race and ethnicity are social constructs rather than biological risk factors. Differences in risk for diseases, including CRC, may be mediated by factors such as social determinants of health. Differences between racial and ethnic groups in CRC incidence may be attributable to modifiable risk factors, including variation in screening rates among persons of different races and ethnicities (35).
Absolute differences in CRC incidence and mortality by racial and ethnic groups are small. Those who identify as non-Hispanic American Indian or Alaska Native and non-Hispanic Black have the highest CRC incidence (49 and 44.3 per 100 000, respectively), followed by non-Hispanic White (38.1 per 100 000), Hispanic (34.4 per 100 000), and non-Hispanic Asian or Pacific Islander (30.7 per 100 000) (2). Colorectal cancer mortality occurs in a similar pattern. Those who identify as non-Hispanic American Indian or Alaska Native and non-Hispanic Black have the highest rates (17.2 and 17.6 per 100 000, respectively), followed by non-Hispanic White (13.1 per 100 000), Hispanic (10.7 per 100 000), and non-Hispanic Asian or Pacific Islander (9.1 per 100 000) (2).
Differences by Biological Sex
Males have a higher risk for developing and dying of CRC than females, although absolute differences are small (43.4 vs. 32.8 per 100 000 and 15.7 vs. 11.0 per 100 000, respectively) (2).
Three RCTs assessed the effectiveness of gFOBT on CRC mortality by sex (12, 13, 18). Two RCTs found no difference in CRC mortality by sex (12, 18); however, another trial found greater reductions in males than in females (13). Among males in the same RCT, the largest reductions in CRC mortality were in those aged 60 to 69 years, whereas females aged 70 years or older had the largest reductions (13).
Three RCTs of FS reported results by sex (14–16). In 2 RCTs, males had larger reductions in CRC mortality (15, 16), but no meaningful differences in effects were seen in the other trial (14). One RCT analyzed data by age and sex and found that males aged 50 to 54 years and 55 to 64 years had greater reductions in CRC mortality compared with females in the same age groups (15). Two RCTs found no sex differences in all-cause mortality (14, 15).
Evidence Gaps and Research Needs
Future research should focus on studying the benefits and harms of screening persons younger than 50 years and older than 75 years to further our understanding of the optimal CRC screening intervals and ages to start and stop. Ongoing comparative trials should better inform selection and frequency within (for example, colonoscopy every 10 or 15 years) and between (for example, FIT compared with sDNA) CRC screening tests.
Emerging Evidence
Following ACP guidance statement methods, we did not search for published studies beyond those identified in the reviewed guidelines. However, we believed the NordICC (Northern-European Initiative on Colorectal Cancer) trial, the first published randomized, pragmatic clinical trial on the effectiveness of screening colonoscopy, warranted brief commentary (40). Colorectal cancer and all-cause mortality were not different between the screening colonoscopy and usual care groups (0.28% vs. 0.31%; RR, 0.90 [CI, 0.64 to 1.16]; and 11.03% vs. 11.04%; RR, 0.99 [CI, 0.96 to 1.04], respectively) at 10-year follow-up in the intention-to-screen analysis (40). Only 42% of those invited to be screened received a colonoscopy. In a secondary analysis assuming all participants randomly assigned to be screened were screened, CRC mortality would have been lower than in the usual care group (0.15% screened vs. 0.30% receiving usual care; RR, 0.50 [CI, 0.27 to 0.77]) (40). Findings from NordICC and future published studies, especially comparative studies, should be used to inform CRC screening guidelines, evidence reviews, and microsimulation models.
ACP Guidance Statements
Age to Start Screening for CRC in Asymptomatic Average-Risk Adults
Guidance Statement 1: Clinicians should start screening for colorectal cancer in asymptomatic average-risk adults at age 50 years
There is a net benefit of CRC screening in average-risk adults starting at age 50 years. New evidence confirms our previous conclusion that CRC screening in adults aged 50 to 75 years reduces CRC incidence and mortality but not all-cause mortality (4, 9, 10). Results stratified by age and studies with a higher mean age showed that those aged 65 to 75 years had the greatest benefit. This may be anticipated as the median age at diagnosis of CRC is 67 years (41). Although a net screening benefit still existed in those aged 50 to 64 years, it was lower at younger age and deemed small in adults aged 50 to 54 years. A detailed rationale can be found in our previous guidance statement (4).
Guidance Statement 2: Clinicians should consider not screening asymptomatic average-risk adults between the ages of 45 to 49 years. Clinicians should discuss the uncertainty around benefits and harms of screening in this population
First, no studies of effectiveness and harms only enrolled participants younger than 50 years or directly stratified results by this age group (9, 10). Sensitivity and specificity data of CRC screening tests in adults younger than 50 years are mainly on adenomas, not CRC (9, 10). In the absence of evidence, it is unknown if diagnostic accuracy of CRC screening tests would be similar to that in older populations. However, we know the predictive value of tests would be lower because the incidence of adenomas and CRC is lower in younger adults.
Second, there is a potential for exacerbating health care disparities, an important factor when considering expanding CRC screening to 45 to 49 years old, particularly given the uncertain net benefit. Opportunity costs and resources need to be weighed as decisions are made about unproven screening programs when there is a shortage of internal medicine physicians in the United States. These shortages lead to prolonged scheduling times for routine medical care services, including colonoscopies. The limited time and resources should be used on prioritizing CRC screening access for adults aged 50 to 75 years and other interventions with proven efficacy or cost savings.
Third, the decision modeling for USPSTF consisted of 3 independently created microsimulation models. Nevertheless, we have concerns about assumptions and parameters used in the modeling that likely provide estimates of screening effectiveness and harms that are optimistic compared with clinical settings. All models assumed perfect adherence which, regardless of age, would likely overestimate benefits of CRC screening and is not consistent with surveillance and empirical data (42–44). The sensitivity and specificity estimates used in the models were assumed or not always consistent with the evidence review (9–11). All 3 models did not simulate the pathway of serrated polyps to CRC (11). All adenomas had the potential to progress to CRC in SimCRC and CRC-SPIN. However, MISCAN allowed for some adenomas not to progress (11), which is important particularly for estimating benefit at younger ages. As a result, MISCAN was likely the most similar to a “real-world” setting for younger ages and produced 41% to 57% lower relative benefit in LYG and 0% to 75% lower benefits in CRC incidence and mortality compared with SimCRC and CRC-SPIN.
Fourth, the net benefit of screening is much less favorable in average-risk adults between ages 45 and 49 years than in those aged 50 to 75 years. At the population level, MISCAN found 13 to 17 LYG per 1000 screened starting at age 45 years compared with 50 years (11). This translates to 5 to 6 additional life-days gained per person. MISCAN also indicated a benefit of preventing CRC cases (1 to 2 per 1000 screened or a 0.1% to 0.2% reduction) and reducing CRC mortality (0.4 to 1 per 1000 screened or a 0.04% to 0.1% reduction) in those who started screening at age 45 years compared with 50 years (11). Although there has been a small increase in CRC incidence among persons aged 45 to 49 years (45), it is lower than in those aged 50 to 64 years and 65 to 74 years (35.1 vs. 71.9 vs. 128.9 per 100 000, respectively) (2). Harms that occur with CRC screening include cardiovascular and gastrointestinal events (for example, serious bleeding, perforation, myocardial infarction, and angina) (range, 0.1 to 2 more per 1000 screened), unnecessary follow-ups, and costs for findings deemed clinically unimportant.
Even if we assumed the modeling study had no limitations and accepted the results at face value, we would conclude that the small estimated benefits and harms roughly balance each other out, resulting in an inadequate net benefit to warrant CRC screening in average-risk adults aged 45 to 49 years. Clinicians should discuss the data on the diagnostic accuracy, incidence, uncertainty around benefits, and harms in average-risk adults aged 45 to 49 years.
Age to Stop Screening for CRC in Average-Risk Adults
Guidance Statement 3: Clinicians should stop screening for colorectal cancer in asymptomatic average-risk adults older than 75 years or in asymptomatic average-risk adults with a life expectancy of 10 years or less
There is no new evidence from RCTs about the benefits and harms of screening for adults older than 75 years or those with limited life expectancy (9, 10). In addition, despite favorable assumptions and parameters in the modeling study, stopping screening at ages 80 and 85 years, compared with 75 years, had no to little additional benefits with an increase in screening tests and a slight increase in serious harms (11). Although estimating life expectancy can be challenging and inaccurate, persons older than 75 years who are in good health and lack history of CRC screening may still derive a net benefit from 1-time screening. A detailed rationale is provided in our previous guidance statement (4). Additional consideration of these issues will need to be made between clinicians and patients when discussing stopping screening in older adults or in those who are in poorer health.
The USPSTF (grade C) guideline recommends selectively screening average-risk adults aged 76 to 85 years based on patients’ overall health, prior screening history, and preferences (6). The ACS guideline (qualified recommendations) recommends individualized screening decisions for adults aged 76 through 85 years based on patient preferences, life expectancy, health status, and prior screening history and to discourage adults aged 85 years from continuing CRC screening (8).
Selecting a Screening Test and Frequency for CRC
Guidance Statement 4a. Clinicians should select a screening test for colorectal cancer in consultation with their patient based on a discussion of benefits, harms, costs, availability, frequency, and patient values and preferences
Guidance Statement 4b. Clinicians should select among a fecal immunochemical or high-sensitivity guaiac fecal occult blood test every 2 years, colonoscopy every 10 years, or flexible sigmoidoscopy every 10 years plus a fecal immunochemical test every 2 years as a screening test for colorectal cancer
Guidance Statement 4c. Clinicians should not use stool DNA, computed tomography colonography, capsule endoscopy, urine, or serum screening tests for colorectal cancer
Among effective screening strategies in Guidance Statement 4b, data are lacking to conclude superiority of one strategy over another (9, 10). Clinicians should also consider availability, cost, and patients’ values and preferences.
There were no new RCTs evaluating the effectiveness of stool or direct visualization tests (9, 10). Clinicians should use a FIT or high-sensitivity gFOBT biennially instead of annually because we observed no differences in effectiveness between these screening intervals. Fewer tests would reduce patient burden and harms. For colonoscopy, modeling data suggest that screening for CRC every 10 years results in the largest net benefit; however, data also showed that screening every 15 years preserves most of the benefit in CRC mortality and LYG while reducing colonoscopy harms, burden, and costs (11). The frequency of colonoscopy and other screening tests should be reevaluated as more benefits and harms data beyond microsimulation models become available. A primary reason for selecting screening tests other than sDNA, serum, urine, CTC, and capsule endoscopy is that these screening tests have no studies of effectiveness. Stool DNA tests had a high sensitivity but lower specificity for CRC, which would lead to unnecessary colonoscopies and other evaluations. The number of unnecessary colonoscopies would also increase with more frequent use (USPSTF currently recommends screening every 1 to 3 years with sDNA). Our guidance statement considered cost, but USPSTF and ACS did not. Over a 10-year time frame, sDNA tests would have a cost equal to that of colonoscopy, whereas other accurate stool tests have lower cost (Supplement Table 7). Computed tomography colonography also leads to a high frequency of extracolonic findings of uncertain benefit or harm. A positive CTC requires follow-up colonoscopy, which reduces its utility as a direct visualization test. A detailed rationale can be found in our previous guidance statement (4).
The USPSTF recommends screening with FIT or a high-sensitivity gFOBT every year, an sDNA test every 1 to 3 years, CTC every 5 years, FS every 5 years, FS every 10 years with an annual FIT, or colonoscopy every 10 years (6). The ACS recommends the same screening strategies but recommends screening with an sDNA test every 3 years and does not specifically recommend using FS plus FIT (8).
Supplemental Material
References
1.
American Cancer Society. Cancer Statistics Center. Accessed at https://cancerstatisticscenter.cancer.org/?_ga=2.172833251.1314374044.1666044285-1140487461.1666044285#!/ on 2 February 2023.
2.
National Cancer Institute Surveillance, Epidemiology, and End Results Program. Colon and rectum recent trends in U.S. age-adjusted mortality rates, 2000-2020. Accessed at https://seer.cancer.gov/statistics-network/explorer/application.html on 11 October 2022.
3.
Wilt TJ, Harris RP, Qaseem A; High Value Care Task Force of the American College of Physicians. Screening for cancer: advice for high-value care from the American College of Physicians. Ann Intern Med. 2015;162:718-725. [PMID: 25984847] doi: 10.7326/M14-2326
4.
Qaseem A, Crandall CJ, Mustafa RA, et al; Clinical Guidelines Committee of the American College of Physicians. Screening for colorectal cancer in asymptomatic average-risk adults: a guidance statement from the American College of Physicians. Ann Intern Med. 2019;171:643-654. [PMID: 31683290] doi: 10.7326/M19-0642
5.
Screening for colon and rectal cancer in average-risk adults. CA Cancer J Clin. 2018;68:282-283.
6.
Davidson KW, Barry MJ, Mangione CM, et al; US Preventive Services Task Force. Screening for colorectal cancer: US Preventive Services Task Force recommendation statement. JAMA. 2021;325:1965-1977. [PMID: 34003218] doi: 10.1001/jama.2021.6238
7.
Qaseem A, Kansagara D, Lin JS, et al; Clinical Guidelines Committee of the American College of Physicians. The development of clinical guidelines and guidance statements by the clinical guidelines committee of the American College of Physicians: update of methods. Ann Intern Med. 2019;170:863-870. [PMID: 31181568] doi: 10.7326/M18-3290
8.
Wolf AMD, Fontham ETH, Church TR, et al. Colorectal cancer screening for average-risk adults: 2018 guideline update from the American Cancer Society. CA Cancer J Clin. 2018;68:250-281. [PMID: 29846947] doi: 10.3322/caac.21457
9.
Lin JS, Perdue LA, Henrikson NB, et al. Screening for Colorectal Cancer: An Evidence Update for the U.S. Preventive Services Task Force. Agency for Healthcare Research and Quality; 2021.
10.
Lin JS, Perdue LA, Henrikson NB, et al. Screening for colorectal cancer: updated evidence report and systematic review for the US Preventive Services Task Force. JAMA. 2021;325:1978-1998. [PMID: 34003220] doi: 10.1001/jama.2021.4417
11.
Knudsen AB, Rutter CM, Peterse EFP, et al. Colorectal Cancer Screening: An Updated Decision Analysis for the U.S. Preventive Services Task Force. Agency for Healthcare Research and Quality; 2021.
12.
Scholefield JH, Moss SM, Mangham CM, et al. Nottingham trial of faecal occult blood testing for colorectal cancer: a 20-year follow-up. Gut. 2012;61:1036-1040. [PMID: 22052062] doi: 10.1136/gutjnl-2011-300774
13.
Shaukat A, Mongin SJ, Geisser MS, et al. Long-term mortality after screening for colorectal cancer. N Engl J Med. 2013;369:1106-1114. [PMID: 24047060] doi: 10.1056/NEJMoa1300720
14.
Atkin W, Wooldrage K, Parkin DM, et al. Long term effects of once-only flexible sigmoidoscopy screening after 17 years of follow-up: the UK Flexible Sigmoidoscopy Screening randomised controlled trial. Lancet. 2017;389:1299-1311. [PMID: 28236467] doi: 10.1016/S0140-6736(17)30396-3
15.
Holme Ø, Løberg M, Kalager M, et al; NORCCAP Study Group. Long-term effectiveness of sigmoidoscopy screening on colorectal cancer incidence and mortality in women and men. A randomized trial. Ann Intern Med. 2018;168:775-782. [PMID: 29710125] doi: 10.7326/M17-1441
16.
Miller EA, Pinsky PF, Schoen RE, et al. Effect of flexible sigmoidoscopy screening on colorectal cancer incidence and mortality: long-term follow-up of the randomised US PLCO cancer screening trial. Lancet Gastroenterol Hepatol. 2019;4:101-110. [PMID: 30502933] doi: 10.1016/S2468-1253(18)30358-3
17.
Lindholm E, Brevinge H, Haglind E. Survival benefit in a randomized clinical trial of faecal occult blood screening for colorectal cancer. Br J Surg. 2008;95:1029-1036. [PMID: 18563785] doi: 10.1002/bjs.6136
18.
Kronborg O, Jørgensen OD, Fenger C, et al. Randomized study of biennial screening with a faecal occult blood test: results after nine screening rounds. Scand J Gastroenterol. 2004;39:846-851. [PMID: 15513382] doi: 10.1080/00365520410003182
19.
Faivre J, Dancourt V, Lejeune C, et al. Reduction in colorectal cancer mortality by fecal occult blood screening in a French controlled study. Gastroenterology. 2004;126:1674-1680.
20.
Chiu HM, Chen SL, Yen AM, et al. Effectiveness of fecal immunochemical testing in reducing colorectal cancer mortality from the One Million Taiwanese Screening Program. Cancer. 2015;121:3221-3229. [PMID: 25995082] doi: 10.1002/cncr.29462
21.
Schroy PC 3rd, Coe A, Chen CA, et al. Prevalence of advanced colorectal neoplasia in White and Black patients undergoing screening colonoscopy in a safety-net hospital. Ann Intern Med. 2013;159:13-20. [PMID: 23817700] doi: 10.7326/0003-4819-159-1-201307020-00004
22.
Nishihara R, Wu K, Lochhead P, et al. Long-term colorectal-cancer incidence and mortality after lower endoscopy. N Engl J Med. 2013;369:1095-1105. [PMID: 24047059] doi: 10.1056/NEJMoa1301969
23.
García-Albéniz X, Hsu J, Bretthauer M, et al. Effectiveness of screening colonoscopy to prevent colorectal cancer among Medicare beneficiaries aged 70 to 79 years. A prospective observational study. Ann Intern Med. 2017;166:18-26. [PMID: 27669524] doi: 10.7326/M16-0758
24.
Zafar HM, Harhay MO, Yang J, et al. Adverse events following computed tomographic colonography compared to optical colonoscopy in the elderly. Prev Med Rep. 2014;1:3-8. [PMID: 25530940] doi: 10.1016/j.pmedr.2014.08.001
25.
Johnson CD, Chen MH, Toledano AY, et al. Accuracy of CT colonography for detection of large adenomas and cancers. N Engl J Med. 2008;359:1207-1217. [PMID: 18799557] doi: 10.1056/NEJMoa0800996
26.
Iafrate F, Iussich G, Correale L, et al. Adverse events of computed tomography colonography: an Italian National Survey. Dig Liver Dis. 2013;45:645-650. [PMID: 23643567] doi: 10.1016/j.dld.2013.02.020
27.
Multicentre Australian Colorectal-neoplasia Screening (MACS) Group. A comparison of colorectal neoplasia screening tests: a multicentre community-based study of the impact of consumer choice. Med J Aust. 2006;184:546-550.
28.
Kim DH, Pickhardt PJ, Hanson ME, et al. CT colonography: performance and program outcome measures in an older screening population. Radiology. 2010;254:493-500. [PMID: 20093521] doi: 10.1148/radiol.09091478
29.
Sali L, Mascalchi M, Falchini M, et al; SAVE Study Investigators. Reduced and full-preparation CT colonography, fecal immunochemical test, and colonoscopy for population screening of colorectal cancer: a randomized trial. J Natl Cancer Inst. 2016;108.
30.
Weissfeld JL, Schoen RE, Pinsky PF, et al; PLCO Project Team. Flexible sigmoidoscopy in the PLCO cancer screening trial: results from the baseline screening examination of a randomized trial. J Natl Cancer Inst. 2005;97:989-997. [PMID: 15998952] doi: 10.1093/jnci/dji175
31.
Segnan N, Senore C, Andreoni B, et al; SCORE3 Working Group-Italy. Comparing attendance and detection rate of colonoscopy with sigmoidoscopy and FIT for colorectal cancer screening. Gastroenterology. 2007;132:2304-2312. [PMID: 17570205] doi: 10.1053/j.gastro.2007.03.030
32.
U.S. Food and Drug Administration. Epi proColon. Accessed at https://www.accessdata.fda.gov/cdrh_docs/pdf13/p130001c.pdf on 2 February 2023.
33.
Cho H, Klabunde CN, Yabroff KR, et al. Comorbidity-adjusted life expectancy: a new tool to inform recommendations for optimal screening strategies. Ann Intern Med. 2013;159:667-676. [PMID: 24247672] doi: 10.7326/0003-4819-159-10-201311190-00005
34.
Lee SJ, Boscardin WJ, Stijacic-Cenzer I, et al. Time lag to benefit after screening for breast and colorectal cancer: meta-analysis of survival data from the United States, Sweden, United Kingdom, and Denmark. BMJ. 2013;346:e8441. [PMID: 23299842] doi: 10.1136/bmj.e8441
35.
Carethers JM. Racial and ethnic disparities in colorectal cancer incidence and mortality. Adv Cancer Res. 2021;151:197-229.
36.
Chukmaitov A, Bradley CJ, Dahman B, et al. Association of polypectomy techniques, endoscopist volume, and facility type with colonoscopy complications. Gastrointest Endosc. 2013;77:436-446. [PMID: 23290773] doi: 10.1016/j.gie.2012.11.012
37.
Wang P, Xu T, Ngamruengphong S, et al. Rates of infection after colonoscopy and osophagogastroduodenoscopy in ambulatory surgery centres in the USA. Gut. 2018;67:1626-1636. [PMID: 29777042] doi: 10.1136/gutjnl-2017-315308
38.
Arora G, Mannalithara A, Singh G, et al. Risk of perforation from a colonoscopy in adults: a large population-based study. Gastrointest Endosc. 2009;69:654-664. [PMID: 19251006] doi: 10.1016/j.gie.2008.09.008
39.
Ko CW, Riffle S, Michaels L, et al. Serious complications within 30 days of screening and surveillance colonoscopy are uncommon. Clin Gastroenterol Hepatol. 2010;8:166-173. [PMID: 19850154] doi: 10.1016/j.cgh.2009.10.007
40.
Bretthauer M, Løberg M, Wieszczy P, et al; NordICC Study Group. Effect of colonoscopy screening on risks of colorectal cancer and related death. N Engl J Med. 2022;387:1547-1556. [PMID: 36214590] doi: 10.1056/NEJMoa2208375
41.
National Cancer Institute. Age and cancer risk. 2021. Accessed at https://www.cancer.gov/about-cancer/causes-prevention/risk/age on 20 October 2022.
42.
Cyhaniuk A, Coombes ME. Longitudinal adherence to colorectal cancer screening guidelines. Am J Manag Care. 2016;22:105-111.
43.
van der Vlugt M, Grobbee EJ, Bossuyt PM, et al. Adherence to colorectal cancer screening: four rounds of faecal immunochemical test-based screening. Br J Cancer. 2017;116:44-49. [PMID: 27923037] doi: 10.1038/bjc.2016.399
44.
Centers for Disease Control and Prevention. Use of colorectal cancer screening tests. Accessed at https://www.cdc.gov/cancer/colorectal/statistics/use-screening-tests-BRFSS.htm on 22 August 2022.
45.
Cho MY, Siegel DA, Demb J, et al. Increasing colorectal cancer incidence before and after age 50: implications for screening initiation and promotion of "on-time" screening. Dig Dis Sci. 2022;67:4086-4091. [PMID: 34486085] doi: 10.1007/s10620-021-07213-w
Comments
Sign In to Submit A CommentInformation & Authors
Information
Published In
Annals of Internal Medicine
Volume 176 • Number 8 • August 2023
Pages: 1092 - 1100
History
Published online: 1 August 2023
Published in issue: August 2023
Keywords
Copyright
© 2023 American College of Physicians.
Authors
Metrics & Citations
Metrics
Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
Download article citation data for:
Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults: A Guidance Statement From the American College of Physicians (Version 2). Ann Intern Med.2023;176:1092-1100. [Epub 1 August 2023]. doi:10.7326/M23-0779
Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults: A Guidance Statement From the American College of Physicians (Version 2). Ann Intern Med.2023;176:1092-1100. [Epub 1 August 2023]. doi:10.7326/M23-0779
View More
Login Options:
Purchase
You will be redirected to acponline.org to sign-in to Annals to complete your purchase.
Access to EPUBs and PDFs for FREE Annals content requires users to be registered and logged in. A subscription is not required. You can create a free account below or from the following link. You will be redirected to acponline.org to create an account that will provide access to Annals. If you are accessing the Free Annals content via your institution's access, registration is not required.
Create your Free Account
You will be redirected to acponline.org to create an account that will provide access to Annals.
Once again, the lowly PCP gets conflicting recommendations from the experts
These latest recommendations from the ACP conflict with those of the USPSTF and the ACS. These experts obviously have no idea about the demands and stresses on practicing primary care internists, who struggle to address multiple clinical issues, both acute and preventative, in a typical 15-20 minute PCP visit and who roll their eyes whenever an expert advises "shared decision making" in discussing a matter like CRC screening. Being risk adverse and always concerned about malpractice lawsuits, most of us will opt for the most conservation recommendation, i.e., starting CRC screening at age 45. Why can't these experts and professional societies for once make our lives a little easier and issue harmonious recommendations?
US Multi-Society Task Force Response to ACP CRC Screening Guidance
We are disappointed that the American College of Physicians (ACP) guides clinicians to consider not screening asymptomatic average-risk adults between the ages of 45 to 49 years for colorectal cancer (CRC). This contrasts with recommendations from multiple respected organizations, including the US-MSTF(1), the USPSTF, the ACG and the ACS. We are concerned that this statement may undermine efforts to increase CRC screening in the face of significant increases in the incidence of CRC in those under age 50, and emerging data showing the benefit of screening in this population. Accordingly, we ask your readers to consider the following. While epidemiologic trends in this age group are described as a “small increase in CRC incidence,” this does not capture the actual magnitude of the public health burden of CRC in 45-49-year-olds. A recent study of CRC incidence rates from 2000-2015 demonstrated a 46% increase in CRC incidence between ages 49 and 50,(2) suggesting the presence of a large undetected pre-clinical CRC burden. In 2023, 12% of colon cancers and 16% of rectal cancers are estimated to occur at ages <50 years, 43% of which were diagnosed between ages 45-49 years.(3) CRC will be the leading cause of cancer-related death among 20 to 49-year-olds by 2030. Recent data demonstrates that initiating screening at age 45 results in a reduction in cumulative CRC incidence by age 60, compared to screening at age 50.(4) While we share the ACP’s concern for potential healthcare disparities, wherever colonoscopy access may be limited, alternative screening tests (e.g., FIT) can promote access. Withholding screening in 45-49-year-olds is a grave missed opportunity to decrease burden of disease in those under age 50 and confer protective benefit as birth cohorts age. While discounting the benefits of CRC screening, the ACP is simultaneously omitting the fact that the harms of screening colonoscopy are significantly lower for younger compared to older individuals.(5) The ACP ultimately recommends “discussing the uncertainty around benefits and harms of screening,” however, given the contradictory guidance the ACP provides compared to other organizations, we fear this recommendation will place undue burden on physicians who are already overtaxed and most patients will reasonably ask, “what do you recommend, doctor?”. We hope the guidance by practicing physicians (within and outside of ACP) will be clear: individuals age 45 to 49 years should get screened for colorectal cancer and should use the screening test they are most likely to complete.
References
1. Patel SG, May FP, Anderson JC, Burke CA, Dominitz JA, Gross SA, et al. Updates on Age to Start and Stop Colorectal Cancer Screening: Recommendations From the U.S. Multi-Society Task Force on Colorectal Cancer. Gastroenterology. 2022;162(1):285-99.
2. Abualkhair WH, Zhou M, Ahnen D, Yu Q, Wu XC, Karlitz JJ. Trends in Incidence of Early-Onset Colorectal Cancer in the United States Among Those Approaching Screening Age. JAMA Netw Open. 2020;3(1):e1920407.
3. Siegel RL, Wagle NS, Cercek A, Smith RA, Jemal A. Colorectal cancer statistics, 2023. CA Cancer J Clin. 2023;73(3):233-54.
4. Ma W, Wang M, Wang K, Cao Y, Hertzmark E, Ogino S, et al. Age at Initiation of Lower Gastrointestinal Endoscopy and Colorectal Cancer Risk Among US Women. JAMA Oncol. 2022;8(7):986-93.
5. Ladabaum U, Mannalithara A, Desai M, Sehgal M, Singh G. Age-Specific Rates and Time-Courses of Gastrointestinal and Nongastrointestinal Complications Associated With Screening/Surveillance Colonoscopy. Am J Gastroenterol. 2021;116(12):2430-45.
Disclosures:
SGP: Olympus America FPM: Freenome Inc, Bayer Pharmaceuticals, Exact Sciences, Geneoscopy JCA: none CAB: Ferring Pharmaceuticals Inc., Novo Nordisk Inc., Shionogi Inc., Freenome, Inc., SLA pharmaceuticals, Janssen Pharmaceuticals, Cancer Prevention Pharmaceuticals, Emtora Biosciences, Ferring Inc. JAD: Premera Blue Cross (spouse) SAG: Olympus America, Microtech, Cook, Medtronic BCJ: Motus, GI (advisory board) AS: Freenome Inc., Iterative Scopes Inc. DJR: Freenome Inc.,
Additional considerations pertaining to the United States Preventive Services Taskforce modeling
The recent guideline from the American College of Physicians (ACP) on colorectal cancer (CRC) screening [1] understandably highlights limitations in the United States Preventive Services Taskforce (USPSTF) modeling study [2]. However, there is another limitation worth serious consideration – The USPSTF modeling study assumes “the observed increase in colorectal cancer incidence among 20- to 44-year-olds in recent years is a cohort effect, and that the increase in risk will be carried forward as individuals age," and the authors incorporate this assumption quantitatively [2, p.v]. They acknowledge the uncertainty in this choice, but it is important to emphasize that uncertainty and to understand the impact this assumption has. Specifically, they estimate baseline CRC risk to be 7.7% to 8.5% over the course of an unscreened 40-year-old’s life (with life expectancy of 40.2 years). However, one sees a substantially lower estimate – namely 4.7% – based on current data from the Surveillance, Epidemiology, and End Results Program (SEER) along with an adaptation of methods described by Welch and Passow (also demonstrable algebraically), population-level data on screening uptake, and the impact of screening on incidence, even if one assumes everyone screened does so with colonoscopy, and estimating an approximate 50% relative reduction in CRC incidence based on the USPSTF systematic review [3]. An absolute difference of 3% may seem to have debatable meaningfulness. However, it is certainly noteworthy considering we are already talking about modest differences across the board in terms of screening efficacy, and it is certainly noteworthy considering the magnitude of differences forwarded by the USPSTF modeling study for people 40 years of age over a lifetime of screening: For example, they estimate 28 vs. 26 CRC deaths averted per 1,000 screened with colonoscopy vs. FIT starting at age 45. This suggests an absolute risk difference (ARD) of –0.2% for colonoscopy vs. FIT. Comparably, they estimate 1 fewer death per 1,000 who start screening at age 45 vs. 50 (ARD –0.1%). In addition to limitations already noted [1], the aforementioned assumptions about incidence generate serious pause about the veracity of these differences. Modeling studies may sometimes be the best evidence available, but one might do well to heed the USPSTF modeling authors’ comments: “models only approximate reality”, “are no substitute for empirical evidence”, and their intent “was not to estimate effectiveness of regimens in real-world settings” [2, p.42]. Unfortunately, their modeling also leaves one with considerable reservation about their results.
References
1. Qaseem A, Harrod CS, Crandall CJ, et al.; Clinical Guidelines Committee of the American College of Physicians. Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults: A Guidance Statement From the American College of Physicians (Version 2). Ann Intern Med. 2023 Aug 1. doi: 10.7326/M23-0779. Epub ahead of print.
2. Knudsen AB, Rutter CM, Peterse EFP, et al. Colorectal Cancer Screening: An Updated Decision Analysis for the U.S. Preventive Services Task Force [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2021 May. (Technical Report, No. 202s.) Available from: https://www.ncbi.nlm.nih.gov/books/NBK570833/.
3. Mayer M. Making use of SEER estimates in the context of cancer screening. figshare; 2023. https://doi.org/10.6084/m9.figshare.23961231.
Annual versus Biennial FOBT/FIT
We read the recent ACP guidance on colorectal cancer (CRC) screening with interest. It is unclear why the guidance recommended biennial fecal occult blood/fecal immunochemical testing (FOBT/FIT) given that we have strong evidence from a randomized clinical trial done in the US that high-sensitivity FOBT performed annually reduces CRC mortality by 33% at 18 years of follow-up, and the benefit is sustained through 30 years of follow up, while the biennial arm had reductions of 18% in CRC mortality.(1,2) The two other trials conducted in Europe used only biennial fecal occult blood testing and reported a reduction in CRC mortality of 18% and 13% (3, 4). There is also real world evidence of reduction of 25% and 52% in CRC incidence and mortality respectively with annual FIT-based CRC screening program in the US. (5) Given that the guidance is meant for US healthcare systems and US population and the current standard is annual FOBT/FIT, the guidance not only creates confusion, but ignores direct, high-quality evidence derived from target populations. Furthermore, FOBT/FIT once every calendar year is a current National Committee for Quality Assurance’s HEDIS measures used to define high value care and determine reimbursement. This quality measure allows up to a 23 month gap between tests. If the ACP guideline were implemented, gaps of up to 35 months would be possible, which are definitely outside of that supported by evidence.
References:
1. Mandel JS, Church TR, Ederer F, Bond JH. Colorectal cancer mortality: effectiveness of biennial screening for fecal occult blood. J Natl Cancer Inst. 1999;91(5):434-7
2. Shaukat A, Mongin SJ, Geisser MS, Lederle FA, Bond JH, Mandel JS, et al. Long-term mortality after screening for colorectal cancer. N Engl J Med. 2013;369:1106-14.
3. Kronborg O, Fenger C, Olsen J, Jørgensen OD, Søndergaard O. Randomised study of screening for colorectal cancer with faecal-occult-blood test. Lancet. 1996;348(9040):1467-71.
4. Scholefield JH, Moss SM, Mangham CM, Whynes DK, Hardcastle JD. Nottingham trial of faecal occult blood testing for colorectal cancer: a 20-year follow-up. Gut. 2012;6:1036-40.
5.Levin TR, Corley DA, Jensen CD, Schottinger JE, Quinn VP, Zauber AG et al. Effects of Organized Colorectal Cancer Screening on Cancer Incidence and Mortality in a Large Community-Based Population. Gastroenterology. 2018;155:1383-1391.
Clinical Guidelines Misused as Prior Authorization Guidelines for Denial
Screening for Colorectal Cancer in Asymptomatic Average Risk Patients reviewed five published guidelines.1 Of note not one of the seven experts who were rating for the ACP accepted other organizations’ guidelines without modifications. One of the rationales offered to support the recommendation to not screen those under age 50 is the potential for exacerbating health care disparities. The concern is genuine, but the solution proposed is problematic. It exacerbates the health care disparity of early onset CRC having greater life span morbidity and mortality from late diagnosis. Other subjective and objective limitations of the guidelines are made clear in the accompanying editorial.2 The ACP guidelines are published with an accompanying notice that they may not apply to all patients or individual clinical situations, and that they should not be used as a replacement for a clinician’s judgment. These ACP guidelines will be subject to interpretation and modification by knowledgeable clinicians and specialty consultants. While publishing new and updated clinical guidelines can enhance clinician knowledge and patient care, they are often misappropriated by the insurance industry and used as criteria to deny services appropriate for individual patients. One of the current challenges for clinicians is the labor intensive and uncompensated expense of the insurance industry demand that they complete a proprietary prior authorization process for quality assurance and cost containment. The prior authorization industry utilizes artificial intelligence (AI) computer algorithms, clerical insurance claims adjusters, and prior authorization reviewers to deny services that do not strictly comply with guidelines. The ACP guidelines statement should clarify that these are population based general recommendations and are not designed for and are inappropriate to use for the denial of clinician submitted prior authorization review for individual unique patients. In addition, with the exponential growth of medical knowledge and advances, clinical guidelines often become outdated well before their published expiration date. The misapplication of clinical guidelines may result in delays of the adoption of new medical knowledge, and result in great harm to patient care and further advances in the field. The primary dictum in the Hippocratic oath of medicine: Primum non nocere (First do no harm), should also apply to clinical guidelines.
References
1 Qaseem A, Harrod C, Crandall C, et al. Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults: A Guidance Statement from the American College of Physicians (Version 2). Ann Intern Med. 2023; 176:1092-1100 doi:10.7326/M23-0779 2Bretthauer M, Yang YX. New American College of Physicians Guidance on Colorectal cancer Screening: Less is More. Ann Intern Med. 2023; 176:1127-1128. Doi:10.7326/M23-1695
Disclosures:
No conflict of interest to disclose
CT colonography is a validated screening option
We were disappointed to read the American College of Physicians (ACP) colorectal cancer screening guidance statement, particularly in regards to the exclusion of CT colonography (CTC) (1). As a multidisciplinary group of academic physicians at the University of Wisconsin, we know the power of choice to improve screening adherence, having incorporated CTC as a part of a robust colorectal cancer screening program. The decision not to recommend CT colonography is in stark contrast to the USPSTF and ACS guidelines. Despite utilizing the same USPSTF systematic reviews, the ACP panel surprisingly arrived at the opposite decision in key areas. As with the short-sighted statement not to screen younger patients (age 45-49), the decision not to recommend CTC is presumably related to the lack of large randomized clinical trials demonstrating effectiveness. Although this appears to suggest that the ACP statements require a more stringent level of evidence, it represents the opposite. Given the tremendous advances in our understanding of colorectal cancer (including underlying genetics, pathobiology, and longitudinal development from benign precursors), ignoring numerous published studies that create a cohesive story on effective screening interventions simply because of a perceived absent trial design hides behind a veneer of an evidence-based decision. In reality, previous randomized trials already demonstrate the preventive nature of polyp detection, and do not need to be unnecessarily repeated. Obviously, USPSTF and ACS have disagreed with the ACP guidance statements based on the peer-reviewed published literature, despite the apparent ‘lack of RCTs demonstrating effectiveness’. CTC is a valuable screening option with favorable risk profile. In addition to simply detecting early cancers like FIT/FOBT and stool DNA, it can detect the precursor adenomas and sessile serrated lesions whereas stool-based testing falls short in this regard to prevent cancers entirely. At the University of Wisconsin, we have published the impact of CTC inclusion, which has resulted in an increase of screening rates by all modalities (2). Given the state of CRC screening adherence in the US today, the ACP guidance recommendations are disappointing as they advocate for a list of screening options which fall woefully short of the target screening goal of 80% set by the National Colorectal Cancer Roundtable (3). In agreement with the USPSTF and ACS, we believe that a menu of options between stool tests, CTC, and colonoscopy, starting at 45 years, is the best way forward to help impact the incidence of this truly preventable cancer.
References:
1. Qaseem A, Harrod CS, Crandall CJ, Wilt TJ, Clinical Guidelines Committee of the American College of P. Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults: A Guidance Statement From the American College of Physicians (Version 2). Ann Intern Med. 2023.
2. Smith MA, Weiss JM, Potvien A, et al. Insurance Coverage for CT Colonography Screening: Impact on Overall Colorectal Cancer Screening Rates. Radiology. 2017;284(3):717-24.
3. National Colorectal Cancer Roundtable. Achieving 80% colorectal cancer screening rates in every community. https://nccrt.org/80-in-every-community;Accessed 11/17/2022.
Disclosures:
DHK: None NKL: Research support from Exact Sciences JMW: Research support from Exact Sciences PJP: Advisor to Bracco, GE HealthCare, and Nanox-AI
New ACP Guidelines may reduce screening benefit by one-fifth
The new American College of Physicians (ACP) Guidelines for Colorectal Cancer Screening[1] differ in important ways from guidelines by the U.S. Preventive Services Task Force and American Cancer Society. Although we appreciate the ambition of the ACP to optimize the balance between the harms and benefits of screening, we disagree with two of their recommendations: 1) biennial rather than annual fecal-immunochemical testing (FIT), and 2) not screening average-risk adults between the ages of 45 to 49 years. Those changes are consequential – if followed, they are projected to curtail approximately one fifth of the benefits afforded by screening. In recommending biennial FIT, the ACP cites no difference in effectiveness coupled with greater patient burden and harms, despite studies consistently showing that annual FIT is more effective than biennial FIT.[2] Moreover, annual FIT is a cost-saving strategy, reducing both cancer mortality and costs. With respect to initiating screening at age 45 rather than 50, the ACP concludes, similar to other guidelines, that earlier screening would increase life-years but increase both the number of colonoscopies and harms due to complications. The ACP guidelines postulate that the benefit of earlier screening would not outweigh the harms. The method used by the ACP to reach these conclusions is unspecified. The National Cancer Institute-funded Cancer Intervention and Surveillance Modeling Network (CISNET) uses modeling to compare the benefits and harms of screening under clearly defined assumptions, with multiple models to address uncertainty. We do not expect models to perfectly predict the future, but projections quantify the balance of benefits and harms of screening using the best currently available evidence. Across the models, shifting from annual FIT from 45 to 75 to biennial FIT from 50 to 75 would reduce colonoscopies by 35%-37%, but also increase cancer deaths by 16%-18% and reduce life-years gained by 18%-21%.[3,4] Finally, we disagree with the ACP regarding guidelines that anticipate non-adherence. While non-adherence reduces the population-level impact of screening, proposing guidelines based on non-adherence would result in recommendations for more – not less – intensive screening.[5] Compared to other guidelines, the ACP guidelines place greater weight on harms and costs and less on the benefit derived from annual vs biennial FIT screening, and modeling studies that illustrate the effectiveness of earlier screening, without being transparent about the method used to compare benefits and harms. The discrepancy between APC and other guidelines will cause confusion, and if followed may increase unnecessary CRC deaths.
References
1. Qaseem, A., Harrod, C. S., Crandall, C. J. & Wilt, T. J. Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults: A Guidance Statement From the American College of Physicians (Version 2). Ann. Intern. Med. (2023) doi:10.7326/M23-0779.
2. Pokharel, R., Lin, Y.-S., McFerran, E. & O’Mahony, J. F. A Systematic Review of Cost-Effectiveness Analyses of Colorectal Cancer Screening in Europe: Have Studies Included Optimal Screening Intensities? Appl. Health Econ. Health Policy. 21, 701–717 (2023).
3. Knudsen, A. B. et al. Colorectal Cancer Screening: An Updated Modeling Study for the US Preventive Services Task Force. JAMA 325, 1998 (2021).
4. CISNET Publication Support and Modeling Resources, Colorectal Cancer Screening. JAMA 2021 : Colorectal Cancer Screening: An Updated Modeling Study for the US Preventive Services Task Force https://resources.cisnet.cancer.gov/projects/#crcr/uspstf2021/explorer.
5. Pedersen, K., Kristiansen, I. S., Sy, S., Kim, J. J. & Burger, E. A. Designing Guidelines for Those Who Do Not Follow Them: The Impact of Adherence Assumptions on Optimal Screening Guidelines. Value Health 26, 1217–1224 (2023).
Epidemiologic Evidence Supporting Colorectal Cancer Screening at 45
We read with interest the new American College of Physicians (ACP) suggestion that average-risk colorectal cancer (CRC) screening begin at age 50 instead of 45, in distinction from USPSTF recommendations(1). Numerous lines of Surveillance, Epidemiology, and End Results (SEER) based epidemiologic evidence support starting screening at 45. Key evidence includes relatively high absolute CRC incidence rates (IRs) of those in their 40’s (2). In addition, rising IRs and increasing burdens of advanced stage CRC support earlier screening(3,4). As research team members and authors of several early-onset CRC epidemiology studies, we wanted to provide our input and review a select group of our supporting analyses. In terms of absolute IRs, SEER registries may underestimate CRC case burdens in those under 50 compared with older than 50 because average-risk screening has historically not been performed to detect preclinical CRC.2 Shedding light on this, an analysis of SEER 18 CRC IRs in one-year age increments demonstrated a steep 46.1% increase (34.9/100,000 to 51.0/100,000) in IRs from age 49 to 50 suggesting that those in their mid to late 40’s have a considerable preclinical CRC burden which is not uncovered until screening is initiated at age 50.2 92.9% of lesions identified at 50 are invasive, potentially requiring interventions including surgery, chemotherapy and radiation, arguing against length-time bias. CRC IRs are also rising. Our group’s analysis in Annals of Internal Medicine, cited as one of the supporting lines of evidence in the USPSTF guidelines for screening at 45, demonstrated a 13% increase in colonic adenocarcinoma and 16% increase in rectal adenocarcinomas in recent years in those ages 40-49 (3,5). Finally, there has been an increasing burden of advanced stage CRC in younger patients.4 Since 2000, there has been a 41% increase in advanced stage CRC in those ages 40-49. The proportion (percent stage contribution) of advanced stage disease in those 40-49 is now 27%.4 These data, along with studies from multiple other research groups, paint a concerning picture of CRC burden of those in their 40s and support screening at age 45. These data also highlight the importance of identifying those at risk for CRC prior to 45 by identifying concerning symptoms and cancer family histories. The increasing burden of early-onset CRC is part of a general trend of increasing rates of other cancer types, thus creating an impetus for new approaches for prevention and earlier diagnosis in younger populations.
References:
1. Qaseem A, Harrod CS, Crandall CJ, Wilt TJ. Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults: A Guidance Statement From the American College of Physicians (Version 2). Annals of Internal Medicine. 2023; 176(8): 1092-1100.
2. Abualkhair WH, Zhou M, Ahnen D, et al. Trends in Incidence of Early-Onset Colorectal Cancer in the United States Among Those Approaching Screening Age. JAMA Network Open. 2020; 3(1): e1920407.
3. Montminy EM, Zhou M, Maniscalco L, et al. Contributions of Adenocarcinoma and Carcinoids to Early-onset Colorectal Cancer. Annals of Internal Medicine. 2021; 174 (2): 157-66.
4. Montminy EM, Zhou M, Maniscalco L, et al. Shifts in the Proportion of Distant Stage Early-Onset Colorectal Adenocarcinoma Demonstrate Rising Distant Cancer Incidence in the United States. Cancer Epidemiology, Biomarkers, and Prevention. 2022; 31(2): 334-341.
5. US Preventive Services Task Force. Screening for Colorectal Cancer: US Preventive Services Task Force Recommendation Statement. JAMA. 2021; 325(19): 1965–77.
Disclosures:
Jordan J Karlitz: Disclosure-Senior Medical Director, GRAIL Eric M Montminy: No disclosures
Bridging Guideline Gaps: The role of Large Language Models for Evidence Summarization
Recently, the American College of Physicians (ACP) published guidance statement for colorectal cancer screening (CRC) for asymptomatic average-risk adults. The guidance recommends screening for CRC in asymptomatic average-risk adults at age 50 years and advises against certain methods such as computed tomography colonography or stool DNA testing (1). This guidance has differences when compared to guidelines from the American College of Gastroenterology which has recommendations to start CRC screening at age 45 for average risk individuals (2). Both set of guidelines have merits and demerits and getting a deep understanding of the level of evidence and data supporting differences in recommendations can be rather cumbersome. Clinical guidelines can be difficult to follow due to long narratives, conflicting recommendations as in this case, and alert fatigue if electronic health record-based alerts are incorporated. Despite these limitations and time pressures in primary care, a discussion on CRC screening is warranted as CRC screening improves outcomes and is deemed cost effective; and patients demonstrate a keen interest(3). Large language models such as chat generative pre-trained transformer (ChatGPT), are being explored for utilization in healthcare with several features such as text generation, answering clinical questions and summarization of evidence but a variable performance. These available language models have general training but do have robust neural networks that make them trainable with specific literature. We utilized ChatGPT version 4.0 (ChatGPT-4) to help summarize evidence from these 2 guidelines for CRC screening with varied recommendations to inform clinicians and allow discussions with patients. In order to extract information from guidelines using the `askyourpdf` plugin within ChatGPT-4, we provided a PDF uniform resource locator (URL) and obtained `doc_id` from the `askyourpdf` platform. We provided the specific doc_id to ChatGPT-4 which subsequently downloaded and stored the content, allowing it to be queried. We queried the trained version of ChatGPT-4 to compare recommendations for CRC screening from the two guidelines, and the language model was able to summarize the differences, the data behind the differences and provided a framework for clinicians to use to discuss recommendations, provide patient education and enable shared decision making. While there can be differences in clinical guidelines, large language models can be used (one time with no knowledge of coding needed) by clinicians. The trained models can be available for individual use for shared clinical decision making and especially useful in time constrained environments. We recommend individual clinicians with busy clinical practices explore training language models to aid them with day-to-day practice.
References:
1. Qaseem A, Harrod CS, Crandall CJ, Wilt TJ, Clinical Guidelines Committee of the American College of P, Balk EM, et al. Screening for Colorectal Cancer in Asymptomatic Average-Risk Adults: A Guidance Statement From the American College of Physicians (Version 2). Ann Intern Med. 2023;176(8):1092-100.
2. Shaukat A, Kahi CJ, Burke CA, Rabeneck L, Sauer BG, Rex DK. ACG Clinical Guidelines: Colorectal Cancer Screening 2021. Am J Gastroenterol. 2021;116(3):458-79.
3. Baeker Bispo J, Bandi P, Jemal A, Islami F. Receipt of Clinician Recommendation for Colorectal Cancer Screening Among Underscreened U.S. Adults. Ann Intern Med. 2023;176(9):1985-7.
Disclosures:
Conflicts of interest: SK: Research grants from Rebiotix / Ferring, Seres, Finch, Vedanta and consulting fees from Takeda, Immuron, Niche and ProbioTech Inc, outside of the submitted work.
Author Response to Comments from Patel, Shaukat, Mayer, and Weiss
Dr. Patel and colleagues’ primary concern about the American College of Physicians’ (ACP) guidance statement was the rise of colorectal cancer (CRC) in adults aged 45-49 years. However, CRC incidence in this group is small and has only increased since 2000 from 29 cases per 100,000 individuals to 35 cases per 100,000 individuals. Moreover, the net benefit for CRC screening is not realized in average risk adults until age 50 years. Earlier screening will likely exacerbate health care disparities, particularly colonoscopy access. A more equitable approach would be to use these resources in priority populations where evidence more convincingly supports net benefit. Broadening screening strategies does not necessarily translate to higher value for patients.
We disagree with Dr. Shaukat and colleagues that an annual guaiac fecal occult blood test (gFOBT) or fecal immunochemical test was a superior screening strategy than biennial. Evidence from thirty-year follow-up data between annual and biennial gFOBT screening showed no meaningful difference in CRC mortality (1.28% vs. 1.52%; risk difference: -0.24% [95% CI -0.50% to 0.03%]) (2). Although each strategy reduced CRC mortality compared to a control (2). Less intensive screening strategies reduces screening burden, false-positive tests, and follow-up colonoscopies.
We agree with Dr. Mayer who highlighted additional limitations of modeling parameters of lifetime CRC risk in an unscreened population. Specifically, the modeling assumed a cohort effect in 20- to 44-year-olds estimating a baseline risk of CRC to be 7.7 to 8.5% in those unscreened. Surveillance data suggests the risk is lower, around 5% (1).
Dr. Weiss commented about potential misuse of ACP’s guidance in coverage decisions. ACP’s guidance may not apply to all patients or individual clinical situations and should not be used as a replacement for a clinician’s judgement or coverage decisions based on that judgement. We appreciate Drs. Weiss and Patel and colleagues sharing our concern about the potential for increasing healthcare disparities with expanding CRC screening to younger, asymptomatic patients. However, we disagree that use of stool tests in younger populations is justified, as evidence does not show a net benefit in this age group.
Some commenters noted concerns about potential confusion of conflicting guidelines. The majority of guideline developers’ recommendation for screening adults aged 45-49 years (3) is “conditional/weak”; thus, they “suggest” screening. This means the desirable effects probably outweigh risk and burden, but there is appreciable uncertainty. In clinical practice, a conditional recommendation does not apply to all patients, and implementation may differ depending on circumstance or patients’ values and preferences. Conditional recommendations do not mean “45 is the new 50” for implementation. ACP states clinicians should consider not screening because benefits probably do not outweigh risks and burden for many patients. The key message for adults aged 45-49 years across the majority of guidelines is similar. It is unfortunate that not all guideline developers are aligned, but inconsistencies highlight evidence limitations and the importance of incorporating patients’ values and preferences and individual and population benefits, harms, and costs.
REFERENCES
Disclosures:
Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M23-0779