USPSTF Colorectal Cancer Screening Guidelines: An Extended Look at Multi-Year Interval Testing

ABSTRACT

Objectives: The US Preventive Services Task Force (USPSTF) released draft recommendations regarding colorectal cancer (CRC) screening in October 2015. Despite evidence that annual fecal blood testing test use is uncommon in screen eligible adults, with only 10.4% reporting the use of such a test in 2012, and features poor adherence over time, the USPSTF recommended only 3 noninvasive screening strategy options, all including annual fecal occult blood testing: 1) annual fecal immunochemical test (FIT) alone; 2) annual FIT in combination with flexible sigmoidoscopy every 10 years; and 3) annual high-sensitivity fecal occult blood test (hsFOBT). Mt-sDNA is the only FDA-approved CRC screening test, is covered by Medicare every 3 years, and is included as an every-3-year (3y) option in the American Cancer Society guidelines. We demonstrate that USPSTF modeling includes an embedded sensitivity analysis of less frequent than annual test adherence, which provides support for the inclusion of mt-sDNA 3y as a recommended test.

Study Design: A descriptive analysis of USPSTF modeling of the clinical impact of various stool based CRC screening strategies.

Methods: We analyzed the data generated by the USPSTF CRC screening models describing the impact of noninvasive CRC screening strategies on CRC incidence, CRC related mortality, life years gained (LYG), colonoscopy volume and associated complication, test efficiency (a measure of benefits (LYG) and harms (colonoscopies generated), and identified strategies that provide 90% or more of the LYG by screening with colonoscopy every 10 years. We compared mt-sDNA at 3y intervals and FIT and hsFOBT at 2-year (2y) and 3y intervals and did not consider annual testing.

Results: We found that only mt-sDNA 3y, FIT 2y, and FIT 3y were within 98% of the efficiency frontier. However, only mt-sDNA 3y generates more than 90% of the life-years gained with screening colonoscopy. These results meet the USPSTF criteria for a recommendation for mt-sDNA 3y for routine screening.

Conclusions: Given poor adherence to annual testing, any screening opportunity with a test, such as mt-sDNA, that has high sensitivity for CRC and for the most significant precancerous lesions would be an important screening option for patients for maximizing screening effectiveness in reducing CRC incidence and mortality.

Am J Manag Care. 2016;22(2):e77-e81

Take-Away Points

As of October 2015, US Preventive Services Task Force (USPSTF) draft recommendations for noninvasive colorectal cancer (CRC) screening depend on annual screening with fecal occult blood tests despite uncommon annual test use and poor adherence over time.

• CRC is the second leading US cause of cancer death with, annually, 133,000 new cases, 50,000 deaths (2015 estimate), and more than $14 billion in related costs.
• The multi-target stool DNA (mt-sDNA) test has significantly higher sensitivity for cancerous and precancerous lesions than fecal blood testing. This could be of great utility in irregularly screened patients and in allowing longer test intervals and lower health system screening-process burdens.
• USPSTF could recommend mt-sDNA screening every 3 years based on USPSTF modeling, providing support for appropriate medical policy.

In its October 5, 2015, draft recommendation (draft statement) regarding colorectal cancer (CRC) screening, the US Preventive Services Task Force (USPSTF) assigned an “A” grade to CRC screening starting at age 50 and continuing until age 75.¹ In addition to the option of screening colonoscopy every 10 years (10y), the USPSTF recommended 3 screening options that include annual fecal occult blood testing: 1) annual fecal immunochemical test (FIT) alone, 2) annual FIT in combination with flexible sigmoidoscopy 10y, and 3) annual high-sensitivity fecal occult blood test (hsFOBT). These recommendations were made despite evidence that annual CRC screening test use across the United States is the least frequent strategy used by screen-eligible adults aged 50 to 75 years (10.4%),² and subject to poor year-over-year adherence in those patients that use it in routine clinical practice. Higher rates can be obtained in highly resourced screening environments focused on annual screening, although reported initial uptake remains less than 50%.^3,4 Further, FIT⁵ and hsFOBT⁶ demonstrate significantly inferior single application sensitivity (point sensitivity) to multi-target stool DNA testing (mt-sDNA).⁴ Mt-sDNA—specifically, the Cologuard test—can be performed at 3-year (3y) intervals.^7,8

Test sensitivity, patient preferences,^9,10 and access are key considerations for ensuring screening effectiveness and adherence. In view of the poor adherence to regular repeat testing, it is critical that the first opportunity we have in screening a patient should be with a test that has a high sensitivity for both CRC and significant precancerous lesions. This would maximize screening effectiveness in reducing CRC incidence and mortality.

Given the lack of adherence to annual testing and patient preference for longer intervals, a comparison of mt-sDNA 3y with biennial or triennial FIT/hsFOBT would be more clinically relevant than comparison with annual testing by the same methods. However, with minimal clinical rationale and no expressed scientific rationale, and without accounting for patient preference factors, the Cancer Intervention and Surveillance Modeling Network (CISNET), which provided the model data to inform USPSTF recommendations, grouped “stool tests” together and ruled that only a single strategy could be “recommended” from that group.¹¹ By considering only annual strategies, the USPSTF recommended both annual FIT and hsFOBT for routine screening, but not mt-sDNA. Mt-sDNA was included in the draft statement as an “alternative test” that “may be useful in select clinical circumstances.” Mt-sDNA is currently recommended for use at 3-year intervals, not annually.⁶

Mt-sDNA uses a single random stool sample to identify 11 biomarkers associated with CRC and precancerous lesions: 9 for altered DNA, 1 for the reference gene beta-actin, and 1 for hemoglobin. Biomarker results are algorithmically combined, provide a composite score, and generate a single “negative” or “positive” patient result. A prospective, cross-sectional, 90-site, 10,000-patient screening study demonstrated that mt-sDNA was significantly superior to FIT for detecting any class of colorectal neoplasia, including 92% versus 74% and 69% versus 46% sensitivity for CRC and high-grade dysplasia, respectively. Specificity (no lesions found during colonoscopy) was 90% for mt-sDNA versus 96% for FIT.⁵

Poor screening adherence can have grave consequences for patients. Although preventable and treatable, CRC is the second leading cause of cancer death in the United States, with an estimated 133,000 new cases and 50,000 deaths estimated for 2015.¹² Recent studies^2,13 indicate that screening rates plateaued in 2010, with an estimated 23 million Americans eligible for, but not participating in, CRC screening.²

The aim of this analysis was to demonstrate that the modeling performed by CISNET supports a recommendation of mt-sDNA 3y for routine screening based on the USPSTF’s criteria, in light of data demonstrating that patients are unlikely to adhere to annual testing.

METHODS

We reviewed the CISNET consortium modeling data in the technical report accompanying the draft statement¹¹ and identified an embedded sensitivity analysis of nonannual adherence by considering mt-sDNA 3y, FIT every 2 years (2y), FIT 3y, hsFOBT 2y, and hsFOBT 3y all as a group, and not considering data from annual testing. CISNET used 3 well-characterized CRC disease state models—Simulation Model of Colorectal Cancer (SimCRC), Microsimulation Screening Analysis (MISCAN) for Colorectal Cancer, and Colorectal Cancer Simulated Population Model for Incidence and Natural history (CRC-SPIN)—to provide screening outcomes under various stool-screening scenarios. These scenarios involved different modalities (hsFOBT, FIT, and mt-sDNA), intervals (2y, 3y, and every 5 years), and screening age groups (50-75, 50-80, 50-85, 55-75, 55-80, and 55-85 years). All scenarios assumed 100% adherence to the screening modality and interval.

Model screening outcomes included CRC incidence reduction, CRC mortality reduction, life-years gained (LYG), total colonoscopies performed, and screening complications, including adverse events related to follow-up colonoscopy. Screening scenarios were plotted against LYG and colonoscopies performed to create an “efficiency frontier,” a line connecting the most efficient screening scenarios based on the trade-off between LYG and colonoscopies. By excluding annual stool tests, we recreated CISNET’s analysis of the “efficiency frontier” and compared these screening scenarios with each other and with 10y colonoscopy screening. Per CISNET, tests were recommended if they were within 98% of the LYG at the equivalent point on the efficiency frontier and generated at least 90% of the LYG generated by 10y screening colonoscopy.

RESULTS

Figure

Table

CISNET modeling showed that mt-sDNA 3y, FIT 2y, and FIT 3y were within 98% of the efficiency frontier, whereas hsFOBT 2y and hsFOBT 3y were not (¹¹). In comparison with FIT 2y and 3y and hsFOBT 2y and 3y, only mt-sDNA 3y generated greater than 90% of the LYG by screening colonoscopy 10y in at least 1 of the 3 models (SimCRC only) (¹¹).

DISCUSSION

By considering only annual strategies, the USPSTF recommended both annual FIT and hsFOBT for routine screening, but not mt-sDNA. Mt-sDNA was included in the draft statement as an “alternative test” that “may be useful in select clinical circumstances” and is currently recommended for use at 3-year intervals, not annually. Screening test adherence is an important determinant of clinical effectiveness, but annual strategies are difficult to execute and are burdensome on patients, clinicians, and health systems. Biennial or triennial adherence more accurately reflect clinical experience with stool testing and provide a more realistic base for projecting outcomes using modeling. This critical issue was not reflected in the USPSTF’s draft statement; however, the USPSTF modeling data itself provide a basis for the evaluation and recommendation of noninvasive screening strategies performed at intervals of more than 1 year.

Combining modeling data, empirically derived data, and clinical experience allows for reasonable clinical decisions on test strategies. We found that the CISNET modeling data support the use of mt-sDNA 3y for routine CRC screening using USPSTF criteria: mt-sDNA 3y lies within 98% of the efficiency frontier and provides greater than 90% of the LYG by screening colonoscopy (SimCRC). Further, the models show that hsFOBT 2y and hsFOBT 3y are not within 98% of the efficiency frontier and do not provide greater than 90% of the LYG by screening colonoscopy. Therefore, logically, the superior performance of mt-sDNA 3y over hsFOBT 2y allows the clinical utility inferred for hsFOBT 2y (from previous randomized controlled trials that showed CRC-related mortality reduction with biennial FOBT screening) ⁶ to be inferred as well for mt-sDNA 3y. Finally, clinical experience supports the use of high-sensitivity nonannual testing, and patient preference for high-sensitivity long-interval testing is known.^2,9

Patient-level cost concerns are mitigated by coverage. The USPSTF-recommended screening tests are free of co-pays and deductibles under the Affordable Care Act, which will help increase access to more effective screening options by eliminating patient out-of-pocket expense. Medicare already covers mt-sDNA every 3 years for traditional Medicare beneficiaries without co-payments or deductible (Medicare cost $509). At a list price of $649, which includes a nationwide 24 hour, 7 day a week patient navigation/compliance system, mt-sDNA is less expensive than colonoscopy and more expensive than FIT/FOBT; at an average cost of $600, it has a cost effectiveness ratio of $11,313 per quality-adjusted life-year compared with not screening, which is well within an acceptable range.⁸

The first opportunity we have to screen a patient for CRC may be the last opportunity. Screening tests should have minimal opportunity for complications and high sensitivity for CRC and significant precancerous lesions to maximize screening effectiveness in reducing CRC incidence and mortality.

CONCLUSIONS

Three factors—CISNET modeling, relied on by the USPSTF in its draft statement; inferential clinical utility data supporting CRC mortality reduction; and clinical experience with lack of adherence to annual testing—all support a recommendation by the USPSTF for mt-sDNA 3y as a routine colorectal cancer screening test.

Author Affiliations: Exact Sciences Laboratories, LLC (BMB, MP) and Exact Sciences Scientific Advisory Board (BL), Madison, WI.

Source of Funding: Exact Sciences Corporation (Madison, WI).

Author Disclosures: Dr Berger and Mr Parton are employees and stockholders of Exact Sciences Corporation, which is the developer and manufacturer of Cologuard, the only mt-sDNA available on the market; USPSTF guideline recommendations could have an impact on the commercial success of Cologuard and on Exact Sciences’ financial performance. Dr Levin is a member of Exact Sciences Scientific Advisory Board and has received consulting fees, though not for this publication.

Authorship Information: Concept and design (BB, MP); acquisition of data (MP); analysis and interpretation of data (BB, MP); drafting of the manuscript (BB, BL); critical revision of the manuscript for important intellectual content (BB, MP, BL); statistical analysis (MP); provision of patients or study materials (MP); administrative, technical, or logistic support (MP); and supervision (BB).

Address correspondence to: Barry M. Berger, MD, FCAP, Chief Medical Officer, Exact Sciences Corporation, 441 Charmany Dr, Madison, WI 553719. E-mail: bberger@exactsciences.com.

REFERENCES

1. Draft recommendation statement: colorectal cancer—screening [table: recommended screening strategies for colorectal cancer]. US Preventive Services Task Force website. http://www.uspreventiveservicestaskforce.org/Page/Document/draft-recommendation-statement38/colorectal-cancer-screening2#table-recommended-screening-strategies-for-colorectal-cancer. Published October 2015. Accessed January 28, 2016.

2. CDC. Vital signs: colorectal cancer screening test use—United States, 2012. MMWR Morb Mortal Wkly Rep. 2013;62(44):881-888.

3. Gellad ZF, Stechuchak KM, Fisher DA, et al. Longitudinal adherence to fecal occult blood testing impacts colorectal cancer screening quality. Am J Gastroenterol. 2011;106(6):1125-1134. doi:10.1038/ajg.2011.11.

4. Jensen CD, Corley DA, Quinn VP et al. Fecal immunochemical test program performance over 4 rounds of annual screening: a retrospective cohort study [published online January 26, 2016]. Ann Intern Med. doi:10.7326/M15-0983.

5. Imperiale TF, Ransohoff DF, Itzkowitz SH, et al. Multitarget stool DNA testing for colorectal-cancer screening. N Engl J Med. 2014;370(14):1287-1297. doi:10.1056/NEJMoa1311194.

6. Lin JS, Piper MA, Perdue LA et al. Screening for colorectal cancer: an updated systematic review for the U.S. Preventive Services Task Force [AHRQ publication No. 14-05203-EF-1]. US Preventive Services Task Force website. http://tinyurl.com/jyzo8ye. Published October 2015. Accessed January 2016.

7. Colorectal cancer prevention and early detection. American Cancer Society website. http://www.cancer.org/cancer/colonandrectumcancer/moreinformation/colonandrectumcancerearlydetection/colorectal-cancer-early-detection-screening-tests-used. Updated January 27, 2016. Accessed January 28, 2016.

8. Berger BM, Schroy PC, Dinh TA. Screening for colorectal cancer using a multitarget stool DNA test: modeling the effect of the inter-test interval on clinical effectiveness. Clin Colorectal Cancer. 2016 [in press]. http://dx.doi.org/10.1016/j.clcc.2015.12.003.

9. Abola MV, Fennimore TF, Chen MM, et al. Stool DNA-based versus colonoscopy-based colorectal cancer screening: patient perception and preferences. Fam Med and Comm Health. 2015;3(3):2-8. doi:http://dx.doi.org/10.15212/FMCH.2015.0125.

10. Inadomi JM, Vijan S, Janz NK, et al. Adherence to colorectal cancer screening: a randomized clinical trial of competing strategies. Arch Intern Med. 2012;172(7):575-582. doi:10.1001/archinternmed.2012.332.

11. Zauber A, Knudsen A, Rutter CM, et al; Writing Committee of the Cancer Intervention and Surveillance Modeling Network (CISNET) Colorectal Cancer Working Group. Technical report: evaluating the benefits and harms of colorectal cancer screening strategies: a collaborative modeling approach [AHRQ pub. No. 14-05203-EF-2]. US Preventive Services Task Force website. http://tinyurl.com/hw9b9zl. Published October 2015. Accessed January 28, 2016.

12. Cancer facts and figures 2015 [table. estimated number* of new cancer cases and deaths by sex, US, 2015]. American Cancer Society website. http://www.cancer.org/acs/groups/content/@editorial/documents/document/acspc-044552.pdf. Published 2015. Accessed January 28, 2016.

13. Sabatino S, White M, Thompson T, Klabunde C; Centers for Disease Control and Prevention. Cancer screening test use—United States, 2013. MMWR Morb Mortal Wkly Rep. 2015;64(17):464-468.

Download PDF: USPSTF Colorectal Cancer Screening Guidelines: An Extended Look at Multi-Year Interval Testing