Fecal occult blood test for colorectal cancer screening: an evidence-based analysis.

Abstract

UNLABELLED:

The colorectal cancer (CRC) screening project was undertaken by the Medical Advisory Secretariat (MAS) in collaboration with the Cancer Care Ontario (CCO).In November 2007, the Ontario Health Technology Advisory Committee (OHTAC) MAS to conduct an evidence-based analysis of the available data with respect to colorectal cancer diagnosis and prevention. The general purpose of the project was to investigate the effectiveness, cost effectiveness, and safety of the various methods and techniques used for colorectal cancer screening in average risk people, 50 years of age and older.The options currently offered for colorectal cancer screening were reviewed and five technologies were selected for review:Computed tomographic (CT) colonographyMagnetic resonance (MR) colonographyWireless capsule endoscopy (PillCam Colon)Fecal occult blood test (FOBT)Flexible sigmoidoscopyIn this review, colonoscopy was considered as the "gold standard" technique by which the effectiveness of all other modalities could be evaluated. An economic analysis was also conducted to determine cost-effectiveness of different screening modalities.Evidence-based analyses have been prepared for each of these technologies, as well as summary document that includes an economic analysis, all of which are presented at the MAS Web site: http://www.health.gov.on.ca/english/providers/program/mas/tech/techmn.html

OBJECTIVE:

The objective of this evidence review is to examine the effectiveness and cost-effectiveness of fecal occult blood testing (FOBT), including guaiac FOBT (gFOBT) and immunochemical FOBT (iFOBT), for use in colorectal cancer (CRC) screening in asymptomatic, average-risk adults. Specifically: Is the use of gFOBT or iFOBT associated with a reduction in CRC and overall mortality?What are the sensitivity and specificity of gFOBT and iFOBT for the detection of 1) CRC and 2) large polyps (≥ 1 cm)?

CLINICAL NEED:

CRC is the most common cause of non-tobacco related cancer death in Canada. It has been estimated that in 2007, 7,800 people were diagnosed with CRC in Ontario and 3,250 died from the disease, making the province's incidence and mortality rate of CRC amongst the highest in the world.

DESCRIPTION OF TECHNOLOGY/THERAPY:

THERE ARE TWO GENERAL TYPES OF FOBT THAT ARE CATEGORIZED ACCORDING TO THE ANALYTE DETECTED: guaiac FOBT (gFOBT) and immunochemical FOBT (iFOBT). Blood in the stool is a nonspecific finding but may originate from CRC or larger (>1 cm) polyps (small adenomatous polyps do not tend to bleed). Bleeding from cancers and larger polyps may be intermittent and not always detectable in a single sample. The FOBT thus requires regular testing that consists of collecting specimens from consecutive bowel movements. A positive gFOBT or iFOBT involves a diagnostic workup with colonoscopy to examine the entire colon in order to rule out the presence of cancer or advanced neoplasia. METHODS OF EVIDENCE-BASED ANALYSIS: A literature search was conducted from January 2003 to June 2008 that included OVID MEDLINE, MEDLINE In-Process and Other Non-Indexed Citations, EMBASE, the Cumulative Index to Nursing & Allied Health Literature (CINAHL), The Cochrane Library, and the International Agency for Health Technology Assessment/Centre for Review and Dissemination.

INCLUSION CRITERIA:

Patients at average risk for CRCAll patients must be at least 50 years of ageBiennial FOBT as a screening modality and use of colonoscopy as the reference standardSystematic reviews and randomized controlled trials (RCTs)

OUTCOMES:

CRC mortality, overall mortality, sensitivity, specificity, adverse effects

EXCLUSION CRITERIA:

Studies involving fewer than 100 patientsStudies that do not report sufficient data for analysis COMPARISONS OF INTEREST: Evidence exists for these comparisons of interest: gFOBT compared with the reference "gold standard" colonoscopy (or double-contrast barium enema where colonoscopy is incomplete or contraindicated)iFOBT compared with the reference gold standard colonoscopy (or DCBE where colonoscopy is incomplete or contraindicated)gFOBT compared with iFOBTThe quality of the diagnostic studies was examined according to the 'GRADE Working Group criteria' for grading quality of evidence and strength of recommendations for diagnostic tests and strategies.

SUMMARY OF FINDINGS:

SINGLE-TEST STUDIES: There is limited direct/indirect evidence that iFOBT has sensitivity/specificity superior to that of unrehydrated gFOBT for CRC detection: sensitivity for gFOBT:13% and 25%pooled iFOBT sensitivity:81%There is evidence that iFOBT and gFOBT have lower sensitivities for adenoma detection than for CRC detection: sensitivity for rehydrated gFOBT22%pooled iFOBT sensitivity28% REPEATED-TEST STUDIES: No trials have examined CRC mortality outcomes after repeated testing of iFOBT. Two RCTs from the United Kingdom and Denmark showed significant reduction in CRC mortality using unrehydrated gFOBT biennially Relative risk reductions of 13% (UK trial) and 16% (Danish trial); absolute difference of 0.1% (UK trial) and 0.2% (Danish trial).No significant reduction in overall mortalityInterval cancers (CRC that develop in the intervals between routine screening) UNITED KINGDOM TRIAL: 236 CRCs detected by positive test, 236 interval CRCs after negative testDANISH TRIAL: 120 CRCs detected by positive test, 146 interval CRCs after negative testUnrehydrated gFOBT has low sensitivity for CRC detection (45% in the UK trial and 54% in the Danish trial). true positive rate50% (United Kingdom and Danish RCTs)false positive rate5%-10%true negative rate90%-95% (from observational studies as RCTs did not report specificity)false negative rate50%ES Table 1:Guaiac FOBT - GRADE Quality of Evidence for InterventionsOutcomeDesignQualityConsistencyDirectnessOverall QualityCRCMortalityRCTDanishN = 137,485United KingdomN = 152,850No seriouslimitationsYes (RR reduction in 2 trials13% and 16%; absolutedifference 0.1% and 0.2%respectively).Age rangeDanish and UnitedKingdom study 45-75yearsHighCRC indicates colorectal cancer; FOBT, fecal occult blood test; GRADE, Grading of Recommendations Assessment, Development and Evaluation; RCT, randomized controlled trial.*Unlikely to be an important uncertainty.ES Table 2:GRADE Quality of Evidence for Diagnostic Tests: Implications of Testing Focusing on AccuracyNew Test and Reference TestPutative BenefitDiagnostic AccuracyPatient Outcomes and Expected Impact on ManagementSensitivitySpecificityTrue PositiveTrue NegativeFalse PositiveFalse NegativePresumed Influence on Outcomes Important to PatientsiFOBT and ColonoscopySimple, non-invasiveLessLessBenefit from diagnosis and treatment after confirmatory colonoscopySmall risk of bowel perforation during colonoscopyBenefit of reassuranceAnxiety/worry leading up to confirmatory colonoscopySmall risk of bowel perforation during confirmatory colonoscopyDetriment from delayed diagnosisDirectness of Evidence (Test Results) for Outcomes Important to PatientsSome uncertainty (until after confirmatory colonoscopy)No UncertaintyUncertaintyUncertaintyFOBT indicates fecal occult blood test; GRADE, Grading of Recommendations Assessment, Development and Evaluation.Es Table 3:Immunochemical FOBT - GRADE Quality of Evidence for Diagnostic StudiesNo. ofStudiesDesignLimitationsIndirectnessInconsistencyImprecise dataQuality6Diagnostic cohort (single test)(reference standard for positive and negative iFOBT results was colonoscopy)No serious limitationsTP Some uncertaintyTN No uncertaintyFP UncertaintyFN UncertaintyTP rate = 69%TN rate = 94%FP rate = 6%FN rate = 30%(from direct comparison study)Diagnostic cohort iFOBT sensitivities: 50% to 90%High I(2)in pooled sensitivity and specificityWide range in confidence intervals in direct comparison studyLow*FN indicates false negative; FOBT, fecal occult blood test; FP, false positive; Development and Evaluation; TN, true negative; TP, true positive.*Uncertainty until after confirmatory colonoscopy†Stress/worry for patient until confirmatory colonoscopy‡Detrimental effects due to delayed diagnosis.§For these 3 reasons, downgrade quality from High to Moderate.║For these 3 reasons, downgrade quality from Moderate to Low. CONSIDERATIONS FOR THE ONTARIO HEALTH SYSTEM: Executive Summary Table 4 shows the potential system pressures and benefit/risk analysis for the use of FOBT and colonoscopy to screen for CRC in average-risk adults, ages 50 and over in Ontario. Es Table 4:Summary of Potential System Pressures for FOBT ScreeningCriterionColonoscopyFOBTPrimarily prevent or detect cancer?Prevent and detectDetectFrequency of screeningEvery 10 yearsMust repeat at regular intervalsEvery 2 yearsMust repeat at regular intervalsLevel of evidenceObservational studiesRCTsBenefitsUsed as gold standard in studiesINTERVENTION GRADE QUALITY: High (gFOBT)DIAGNOSTIC GRADE QUALITY: Low (iFOBT)No RCTs examining the effectiveness of repeated iFOBT on CRC mortality reduction were identifiedLimited direct/indirect evidence that iFOBT has superior sensitivity/specificity to unrehydrated gFOBT for detection of CRCRisks0.1% risk of serious bleeding and perforation requiring surgery0.3% risk of serious complications (stroke/bleeding requiring hospitalization/ myocardial infarction)High interval cancer rateThe small benefit in CRC mortality reduction (absolute difference 0.1% to 0.2%) also coincides with a 0.3% risk of serious complications.Preparation requirementsNo food 1 day prior to examOffice/hospital visitComplete bowel preparationSedationEliminate citrus fruit and juices and vitamin C from diet for 3 days prior to/during stool collection.Person applies 2 samples per bowel movement (each occurring on 3 different days) onto test areas of FOBT cards.Resources required for screening asymptomatic, average-risk adults ≥ 50 yearsIncreased demand for colonoscopies and colonoscopists or nurses who perform colonoscopies. (ABSTRACT TRUNCATED)

• Authors: Medical Advisory Secretariat
• Journal: Ontario health technology assessment series (Ont Health Technol Assess Ser) Vol. 9 Issue 10 Pg. 1-40 ( 2009) ISSN: 1915-7398 [Print] Canada
• PMID: 23074514 (Publication Type: Journal Article)