Society: AGA

Introduction: The ultimate goal of colorectal cancer (CRC) screening is to prevent CRC by detecting advanced adenomas that potentially progress to CRC. Previous studies demonstrated that the Asia-Pacific Colorectal Screening (APCS) score could select more advanced neoplasia (AN) subjects and reduce the number needed to screen (NNS). Using the recommended cutoff FIT for the average-risk individual at 100 ng/ml could have missed AN in the high-risk individuals. Hypothetically, using the lower cutoff of FIT in those with high risk may improve the sensitivity for AN detection. Therefore, we aimed to evaluate the performance of the cutoff FIT of 25 ng/ml in the high-risk population and the cutoff FIT of 100 ng/ml in the non-high-risk population.
Methods: We conducted an organized CRC screening program in four districts in Thailand. Asymptomatic subjects aged 50-75 were invited and interviewed by the village health volunteer. Using the APCS score (i.e., age, male, smoking, and a first-degree relative with CRC), subjects were stratified into the high-risk or non-high-risk group. The positive FIT was called in the high-risk group with quantitative FIT using the cutoff ≥25 ng/ml (FIT25) and ≥100 ng/ml (FIT100) in the non-high-risk group. Positive FIT subjects were invited for colonoscopy. We assessed the positive predictive value (PPV) and NNS to detect AN—AN comprised of advanced adenoma and CRC. We compared the two cutoffs (FIT25/100) in high and non-high-risk individuals and the single cutoff in all individuals (FIT100). The analysis population was intention-to-treat.
Results: Of 12,000 invited subjects, 10262 subjects (85.5%) participated in the FIT program. The mean age was 60 years, and 33.7% were male. Of those, 8578 (83.6%) screenees were classified as non-high-risk, and 1684 (16.4%) were high-risk. The return rate of FIT was 99.6% (n=10221). The acceptance rate for colonoscopy was 68.3% (n=915). The positivity rate of FIT25/100 was 13.1% (n=1340) in the FIT25/100 group and 9.4% (n=961) in the FIT100 group (p<0.001). In the FIT25/100 and FIT100 groups, AN and CRC were detected in 158 and 34 screenees and 133 and 33 screenees, respectively. The PPV for AN (11.8% vs. 13.8%, p=0.17) and CRC (2.5% vs. 3.4%, p=0.21) were not significantly different between the FIT25/100 and FIT100 groups. The NNS to detect one AN in FIT25/100 was 64.9 (95%CI, 63.7 - 66.2) and significantly lower than that of in FIT100 (77.2 (95%CI, 75.7 - 78.7); p<0.001) (Table 1). This strategy reduced 15.8% in NNS to detect AN compared to FIT100 for all individuals (p<0.001).
Conclusion: Using a lower FIT cutoff in the high-risk individuals showed better performance with lower NNS while maintaining high PPV for AN and CRC detection compared to using a high cutoff for all individuals. This strategy could reduce the workload in CRC screening in limited-resource countries.

INTRODUCTION
Colorectal cancer (CRC) is the 2nd leading cause of cancer deaths, and screening is crucial to improve outcomes by early detection. Fecal immunochemical test (FIT) is the most common CRC screening method worldwide, and recommended annually for those at average risk. FIT-based screening programs vary in quality and efficacy, and rely on providers correctly ordering FIT, reliable patient participation, and timely diagnostic follow-up after abnormal results. We aimed to assess the frequency of and reasons behind inappropriate FIT utilization at the Phoenix Veterans Affairs Healthcare System.

METHODS
A retrospective review of FITs ordered from January to February 2019 was performed. For each FIT, the patient chart was reviewed to identify patient demographics, date of prior screening, and presence of high-risk features i.e., personal history of advanced polyps or CRC, IBD, or family history of CRC. FIT was deemed inappropriate if it was performed for any indication other than CRC screening, in a high-risk patient, or within the recommended interval from prior screening. Among positive FITs, performance of follow-up colonoscopy and reasons for lack of diagnostic follow-up was assessed.

RESULTS
A total of 546 patients underwent FIT during the observed period. Of them, 52 tested positive (9.5%), 490 tested negative (89.7%), and 4 tests were canceled (0.7%). Of all FITs, 312 of 546 were obtained for CRC screening (57%). The remainder (234 tests, 42.8%) were completed for inappropriate indications i.e., evaluation of anemia (9.5%), gastrointestinal bleeding (2.9%), other symptoms (2.9%), or unknown indication (27.5%). In those who underwent FIT for CRC screening, 88 of 312 (28.2%) had high-risk features including history of CRC (1%) or adenoma (25%), 1st degree relative with CRC (5.1%), or IBD (0.6%). In addition, 69 patients (22.1%) had recent FIT within 1 year or adequate colonoscopy within 5 years. FIT for CRC screening was appropriate in only 192 (35.2%) of all tests, and inappropriate in 354 (64.8%). Lastly, only 53.8% of patients who had a positive FIT underwent colonoscopy, due to patient refusal (13.5%), lack of referral for colonoscopy (30.8%), or GI advising against colonoscopy (5.8%).

CONCLUSIONS
In our review, inappropriate FIT use was substantial, with 64.8% of the 546 patient tests analyzed deemed inappropriate. The most common reason of inappropriate FIT was its use outside of CRC screening (42.8% of tests). When performed for CRC screening, 38.5% tests were inappropriately performed in patients with at least 1 high-risk feature (28.2%) or within the recommended surveillance interval from prior testing (22.1%). This review highlights the need for optimizing adherence to CRC screening guidelines by promoting accurate identification of eligible patients for screening with FIT and avoiding the use of FIT when a colonoscopy is indicated.

<b>Table 1. Factors leading to inappropriate FIT use determination.</b> <i>Patients who underwent FIT were deemed to have had an inappropriate test if they underwent FIT for any other indication than CRC screening. Among those who underwent FIT for CRC screening, FIT was deemed inappropriate if the patient had any of the listed historical risk factors (either confirmed per review of the patient’s chart, or unknown and therefore unable to confirm that patient did not have the specific risk factor).</i>

Table 1. Factors leading to inappropriate FIT use determination. Patients who underwent FIT were deemed to have had an inappropriate test if they underwent FIT for any other indication than CRC screening. Among those who underwent FIT for CRC screening, FIT was deemed inappropriate if the patient had any of the listed historical risk factors (either confirmed per review of the patient’s chart, or unknown and therefore unable to confirm that patient did not have the specific risk factor).

<b>Table 2. Follow-up diagnostics after abnormal (positive) FIT.</b>

Table 2. Follow-up diagnostics after abnormal (positive) FIT.

Introduction: Fecal immunochemical test (FIT) effectiveness for colorectal cancer (CRC) prevention and early detection is contingent on timely colonoscopy after abnormal results. Receipt of follow-up colonoscopy is low in safety-net settings, including Federally Qualified Health Centers (FQHCs), which provide primary care services to 30 million Americans annually. We aimed to determine patient factors associated with receipt of, and time to, follow-up colonoscopy in one of the largest FQHCs in California.

Methods: We identified patients ages 50-75 with an abnormal FIT result between 1/1/2016 and 8/13/2019 at Northeast Valley Health Corporation (NEVHC) in Los Angeles County. Data were obtained from electronic health records. Outcomes were 1) receipt of follow-up colonoscopy at 6 mos. and 2) number of days between FIT result and follow-up colonoscopy. We used multivariable logistic regression to assess correlates of colonoscopy at 6 months after abnormal FIT and Cox regression to identify patient characteristics associated with time to colonoscopy.

Results: Our study included 1,475 patients (abnormal FIT rate 12.4%) and the overall follow-up colonoscopy completion rate was 19.3% at 6 months. The mean (sd) time to colonoscopy completion was 8.5 (7.5) months, or 259.6 (228.6) days. Median age at FIT completion was 58.6 years, 62.4% were female, 80.5% were Latino/a, 74.0% reported Spanish language preference, and 49.5% were Medicaid insured (Table). Patients of other race/ethnicity (aOR 0.12, 95%CI 0.02-0.97) and with Spanish preference (aOR 0.55, 95%CI 0.35-0.85) were less likely to have completed colonoscopy within 6 months compared to White patients and English speakers (Table). Patients with health insurance or with an abnormal FIT in later study years were more likely to have completed colonoscopy within 6 months (Table). Black (aHR 0.41, 95%CI 0.18-0.95) and other race/ethnicity (aHR 0.36, 95%CI 0.13-0.99) patients were more likely to have longer time to colonoscopy compared to White patients. Patients with Medicare had shorter time to colonoscopy compared to uninsured patients (HR 1.45, 95%CI 1.05-2.01), and patients with an abnormal FIT in later study years also had shorter time to colonoscopy (Figure).

Discussion: FQHCs face considerable challenges to obtaining colonoscopies for patients with abnormal FIT results. In this large FQHC, patients with abnormal FIT were less likely to undergo colonoscopy within 6 months if their primary language was Spanish or if they were uninsured and were more likely to have a long delay to colonoscopy if they were Black or uninsured. We also observed improvement in FIT follow-up over time, likely reflecting the impact of system-level interventions at NEVHC, including tracking and navigation for patients with abnormal FIT results. Additional policies, resources, and interventions are urgently needed.

Background: Major guidelines have recently recommended starting screening at an earlier age in response to the rising incidence of young-onset colorectal cancer (CRC). It is, however, unclear whether screening from a younger age can confer additional benefit to the already established effectiveness of screening for subjects aged 50 or older in reducing CRC mortality and incidence.
Method: We retrospectively designed and compared the “early screening cohort” who started FIT screening at the age of 40 to 49 (study group) and the “ordinary screening cohort” who started only after 50 (control group) through the natural setting of the community-based fecal immunochemical test (FIT) screening for residents aged 40 to 69 in two communities (Keelung and Tainan) prior to the launch of the Taiwan CRC Screening Program eligible for those aged 50 to 69 years. Both cohorts were followed up till 2019 to identify incident CRC cases through the linkage to Taiwan Cancer Registry. The cumulative incidence of CRC was calculated and expressed as CRC per 100,000 person-years in each group and compared. Multivariable analysis with Cox proportional hazards regression model was used to calculate the adjusted hazard ratio for evaluating incremental effectiveness with adjustment for gender, birth year, and family history of colorectal cancer in their first-degree relatives.
Results: Of those 492,224 residents aged 40 to 49 in these two cities from 2001 to 2009, 261,823 subjects participated in the Taiwan CRC Screening Program when they were 50 or older. Of them, 39,334 subjects participated in their first screening when they were 40 to 49, and the remaining 222,489 started FIT screening only after 50. With a follow-up time of 9.8 and 10.5 years, the cumulative incidence of CRC was 45.3 and 67.8 per 100,000 person-years in the study and control groups, respectively (p<.0001). Those who participated in FIT screening from the age of 40 to 49 had a significantly lower risk of incident CRC with an adjusted relative risk (aRR) of 0.71 (95%CI=0.61-0.83). If the study group was stratified into those who started screening at 40-44 and 45-49, then the aRR was 0.71 (95%CI=0.59-0.85) in the latter but there was no significant difference between these two subgroups (aRR=0.90, 95%CI=0.61-1.33).
Conclusion: FIT screening starting at age 40 to 49 as opposed to those who began screening at 50 led to an incremental 29 % reduction of incident CRC, with equivalent effectiveness beginning screening from 45, supporting the most up-to-date screening age recommendation.

Demographics and incidence of CRC in the early screening and ordinary screening cohorts

Demographics and incidence of CRC in the early screening and ordinary screening cohorts

Cumulative incidence of CRC in early screening and ordinary screening cohorts

Cumulative incidence of CRC in early screening and ordinary screening cohorts

Background:
Fecal immunochemical test (FIT) and FIT-DNA test are stool-based tests that are recommended by all gastroenterology societies and United States Preventative Services Taskforce (USPSTF) for average-risk colorectal cancer (CRC) screening. Previous studies reported a higher adenoma detection rate (ADR) with FIT-DNA than FIT. However, similar data are lacking for performance of stool-based tests on sessile serrated polyp/lesions (SSPs) detection. Therefore, we performed a meta-analysis aimed at evaluating the sessile serrated polyp detection rate (SSPDR) of FIT and FIT-DNA test in individuals undergoing average-risk CRC screening.
Methods
A comprehensive literature search of multiple databases (until September 2022) was performed to identify studies reporting SSPDR in patients with positive FIT or FIT-DNA for average-risk CRC screening. The outcome was overall detection rates of SSP and advanced serrated polyps (ASP; SSP ≥ 10 mm). SSPDR and ASPDR were compared between FIT and FIT-DNA cohorts. Subgroup analyses were performed based on FIT cutoff, continent, and study type. All analyses were performed using R statistical software (Metafor package).
Results
A total of 482,405 patients were included from 23 studies (Table 1). The mean age was 62.3 ± 4.4 years and there were 52.4% females from 21 studies. The pooled SSPDR for all positive stool-based tests in was 5.2% (95% confidence interval [CI]: 4.0 - 20.3; I²=99.5%, 18 studies). The pooled SSPDR in FIT-DNA positive individuals was 11.3% (95% CI: 6.6 – 18.6; I²= 95.4%) which was significantly higher compared to FIT positive individuals (4.1%, 95% CI: 2.9 – 5.6; I²= 99.6%; p=0.001; Figure 1a). The overall pooled ASPDR was 1.4% (95% CI: 0.81 – 2.3; I²= 96.7%). ASP detection rate was higher in FIT-DNA positive individuals [3.8 % (95% CI: 1.7 – 8.6; I²= 92.8%)] as compared to FIT positive individuals (0.71%, 95% CI: 0.36 – 1.4; I²= 75.4%; p< 0.01; Figure 1b). SSPDR in FIT-DNA positive individuals was also significantly higher than FIT positive individuals in North American subgroup (11.3% vs. 7.2%, p<0.001) (Figure 2a) and FIT ≥ 10 ug/g group (11.3% vs. 5.8%, p<0.001) (Figure 2b). Pooled SSPDR was significantly higher in FIT≥ 10 ug/g group as compared to FIT ≥ 20 ug/g group (5.8% vs. 2.1%, p<0.006).

Conclusion:
Our study reports SSP and ASP detection rates of 5.2% and 1.4% for positive stool-based screening tests in the average-risk population. FIT-DNA outperformed FIT tests in both SSP and ASP detection. Individuals undergoing colonoscopy after positive FIT-DNA had significantly higher rates of SSL as compared to the FIT ≥ 10 ug/g group and the North American subgroup. Further studies are needed to study cost-effectiveness and validate our findings.

Figure 1 Forest plot showing pooled SSP (A) and ASP (B) detection rate of FIT and FIT-DNA positive groups

Figure 1 Forest plot showing pooled SSP (A) and ASP (B) detection rate of FIT and FIT-DNA positive groups

Figure 2 Forest plot showing pooled SSP detection rate of FIT-DNA positive individuals as compared to FIT in North American group (A) and FIT ≥ 10 ug/g group (B)

Figure 2 Forest plot showing pooled SSP detection rate of FIT-DNA positive individuals as compared to FIT in North American group (A) and FIT ≥ 10 ug/g group (B)

Michelle K. Kim

Appearances