Fecal Immunochemical Tests for Colorectal Cancer Screening: A Systematic Review of Accuracy and Compliance

Details

Files
Project Line:
Health Technology Review
Project Number:
M0010-000

Context and Policy Issues

Colorectal cancer (CRC) is the second leading cause of cancer-related deaths among men and women in Canada. Compared to no screening, screening for CRC is associated with a reduction in incidence and mortality. Early detection results in an improved prognosis for individuals with CRC. At least two types of stool-based screening tests are used in Canada to detect occult blood from bleeding cancers and adenomas. These include the guaiac fecal occult blood test (gFOBT) and the fecal immunochemical test (FIT). The purpose of this report is to review the evidence regarding the diagnostic accuracy and patient compliance in screening with FIT.

Research Questions

  1. What is the accuracy (sensitivity and specificity) of fecal immunochemical tests compared to guaiac tests for colorectal cancer screening?
  2. What is the evidence that compliance is higher with the fecal immunochemical test compared to the guaiac test for colorectal cancer screening?

Methods

The following bibliographic databases were searched through the Ovid interface: MEDLINE, MEDLINE In-Process & Other Non-Indexed Citations, and EMBASE. Parallel searches were run in PubMed and The Cochrane Library. Methodological filters were applied to limit the retrieval to health technology assessments, systematic reviews, meta-analyses, randomized controlled trials, controlled clinical trials, observational studies, and Canadian guidelines. The search was restricted to English language clinical articles published between 2004 and April 2009. Regular alerts were established on EMBASE and MEDLINE, and information that was retrieved via alerts was current to June 15, 2009.

Grey literature (literature that is not commercially published) was identified by searching the websites of health technology assessment and related agencies, professional associations, and other specialized databases. Google and other Internet search engines were used to search for additional web-based materials and information. These searches were supplemented by hand searching the bibliographies and abstracts of key papers. Two independent reviewers screened articles for selection.

Summary of Findings

The literature search identified one health technology assessment (HTA), two systematic reviews, two randomized controlled trials (RCTs) that were not included in the two systematic reviews, and six observational studies that reported on the diagnostic accuracy of FIT compared with gFOBT. In addition, one RCT and one observational study were identified that evaluated patient compliance.

A 2007 HTA by the New Zealand Health Technology Assessment program examined the clinical and cost-effectiveness of screening tests for colorectal cancer. Six observational studies and one RCT evaluating FIT were included, and six types of FIT were evaluated. In most studies, FITs did not perform as well as gFOBTs in diagnostic accuracy. In a single study, one FIT, HemeSelect, performed equally as well as or better than the two gFOBTs (Hemoccult II and Hemoccult SENSA). The authors of the HTA concluded that the evidence regarding the clinical effectiveness of FIT compared with gFOBT is weak, mainly because of the paucity of high quality studies.

The Norwegian Knowledge Centre for Health Services conducted a systematic review of methods for CRC screening. The authors identified two systematic reviews and one guideline regarding the diagnostic accuracy of FIT. The sensitivities of FIT ranged from 82% to 94% across the three reports, and the specificity was reported to be 87% across the three reports. The authors did not make any conclusions about the accuracy of FIT compared with gFOBT. Another systematic review published in 2008 by Whitlock et al. included nine cohort studies evaluating FITs, including HemeSelect and FlexSure. The authors concluded that FITs that had better sensitivities and similar specificities as Hemoccult II may represent reasonable CRC screening test alternatives. Results from the systematic review by Whitlock et al. were used to help inform a 2008 US Preventive Services Task Force (USPSTF) recommendation on colorectal cancer screening.

Two RCTs comparing the diagnostic accuracy of FIT to gFOBT are included in the current report. A 2009 RCT found that the FIT (OC-Sensor Micro) detected more advanced neoplasia at a hemoglobin cut-off level of 50 nanograms (ng)/millilitre (mL) and 200 ng/mL than did the gFOBT (Hemoccult II). The detection rates for CRC were similar at all cut-off levels. A 2008 RCT compared the performance of FIT with that of gFOBT and reported that, while the specificity of gFOBT to detect CRC or advanced adenomas was statistically significantly higher than FIT, the detection rates of FIT were statistically significantly higher than those of gFOBT. A third RCT compared patient participation rates between FIT and gFOBT, and found a higher rate of compliance with FIT (35.8%) compared with gFOBT (30.4%).

Four observational studies that were included reached similar conclusions regarding the comparison of FIT and gFOBT tests. Overall, they found that FIT was more sensitive for the detection of cancers and significant adenomas than gFOBT, and had a higher specificity than gFOBT for detecting advanced adenomas. Another study evaluated the use of a two-tier screening approach that involved the use of the relatively inexpensive gFOBT test as a first-round screening tool, and only followed up with FIT in those individuals with a positive gFOBT. The authors concluded that this approach was effective in identifying those with cancer or clinically significant adenomas.

A comparison of the bedside FIT (Prevent ID CC) to gFOBT (Hemoccult) and to the human hemoglobin enzyme-linked immunosorbent assay (ELISA [Immundiagnostik AG]) was made in a 2006 study. The authors concluded that the beside FIT was an accurate test with similar performance characteristics to the ELISA, and that the sensitivity of the bedside FIT was statistically significantly greater than that of the gFOBT.

A 2005 study in a community-based rural setting found that FIT (InForm) had a higher patient participation rate than gFOBT (Hemoccult II).

Conclusions and Implications for Decision-Making or Policy-Making

The results of this review suggest that FIT may be effective as a method of screening for CRC and advanced adenomas, and that FIT may be more effective when compared to other screening tests, such as gFOBT. In particular, the HemeSelect, FlexSure OBT, and OC-Sensor Micro FITs have demonstrated improved diagnostic performance characteristics compared with the gFOBT. All included studies that compared participation rates of FIT with other screening tests demonstrated that FIT had higher completion rates than the other tests, including gFOBT. The type of FIT and associated costs, the appropriate hemoglobin cut-off to use, and the capacity for follow-up by colonoscopy or flexible sigmoidoscopy may contribute to deciding if FIT is an appropriate CRC screening tool.