Technology Evaluation Center

Immunochemical versus Guaiac Fecal Occult Blood Tests

Assessment Program Volume 19, No. 5 July 2004

Executive Summary Performing an annual fecal occult blood test (FOBT) is one of several recommended options for colorectal cancer screening in the average risk population beginning at age 50. Annual or biennial screening with guaiac-based FOBTs (gFOBT) has been shown in large, randomized trials to have a significant and beneficial effect on colorectal cancer incidence and mortality. However, while the specificity of these tests is generally high, sensitivity is poor. Complicated dietary restrictions prior to testing and sampling instructions may limit patient compliance. Newer, immunochemical FOBTs (iFOBT) are reported to have improved performance characteristics compared to guaiac tests, no dietary restrictions, and in the case of one immunochemical test, InSure™, simpler instructions for sampling. This Assessment will evaluate whether there is sufficient evidence to evaluate the performance of iFOBTs in general, or of specific iFOBTs, and to compare performance to standard gFOBTs. In addition, this Assessment will examine the evidence on patient compliance with various FOBT formats to determine if compliance is more likely with any or with a specific iFOBT versus gFOBTs. Based on the available evidence, the Blue Cross and Blue Shield Medical Advisory Panel made the following judgments about whether immunochemical fecal occult blood tests meet the Blue Cross and Blue Shield Association Technology Evaluation Center (TEC) criteria. 1. The technology must have final approval from the appropriate governmental regulatory bodies. InSure™ (Enterix, Inc.), Instant-View ® (Alpha Scientific Designs, Inc.), immoCARE® (Care Products, Inc.), FlexSure® OBT (Beckman Coulter), HemeSelect® (Beckman Coulter), and MonoHaem® (Chemicon International, Inc.) are iFOBTs that have received U.S. Food and Drug Administration (FDA) approval. However, FlexSure® OBT and HemeSelect® assays are no longer available commercially in the U.S. Guaiac-based FOBTs and most iFOBTs are categorized as waived under Clinical Laboratory Improvement Amendments of 1988 (CLIA) regulation and may be developed in any laboratory with a CLIA license for waived tests, such as a physician’s office lab. Tests may be waived from regulatory oversight if they meet certain requirements established by the statute. Minimal scientific and technical knowledge, training, and experience are required to perform waived tests.

®

®

BlueCross BlueShield Association An Association of Independent Blue Cross and Blue Shield Plans

2. The scientific evidence must permit conclusions concerning the effect of the technology on health outcomes. Annual or biennial screening with gFOBTs has been shown in large, randomized trials to have a significant and beneficial effect on colorectal cancer incidence and mortality. However, while the NOTICE OF PURPOSE: TEC Assessments are scientific opinions, provided solely for informational purposes. TEC Assessments should not be construed to suggest that the Blue Cross Blue Shield Association, Kaiser Permanente Medical Care Program or the TEC Program recommends, advocates, requires, encourages, or discourages any particular treatment, procedure, or service; any particular course of treatment, procedure, or service; or the payment or non-payment of the technology or technologies evaluated. ©2004 Blue Cross and Blue Shield Association. Reproduction without prior authorization is prohibited.

1

Technology Evaluation Center

specificity of these tests is generally high, sensitivity is poor. Complicated dietary restrictions prior to testing and sampling instructions may limit patient compliance. Newer, immunochemical tests are reported to have improved performance characteristics compared to guaiac tests, no dietary restrictions, and in the case of 1 immunochemical test, InSure™, simpler instructions for sampling. By comparing iFOBT performance characteristics and patient compliance to those for standard gFOBTs, effects on health outcomes can be inferred. Is there sufficient evidence to evaluate the performance of iFOBTs in general? Are there relative differences in performance among iFOBTs? Is iFOBT performance as good as or better than that of standard gFOBTs? The minimum study inclusion criteria for evaluating iFOBT performance were as follows: ■ ■ ■ ■ ■ ■ ■ ■

Published as a full-length article in the English language Enrolled a well-described patient group for CRC screening Excluded patients with known causes of bleeding Successfully completed FOBT by at least 85% of enrollees Performed endoscopy on all enrollees Compared 2 or more FOBTs including at least 1 iFOBT Used no rehydration for Hemoccult II® tests Reported FOBT results for cancer, adenomas larger than 1 cm, or both combined

Seven studies met these criteria, enrolling a total of 3,000 individuals for gFOBT testing and 3,834 for iFOBT testing. None of the studies enrolled an average-risk colorectal cancer screening population. To compare performance of screening tests, smaller studies of higher-risk screening populations in which all receive endoscopy were considered acceptable to increase the yield of neoplasia while avoiding verification bias. The majority of comparative data on iFOBTs are derived from studies of FlexSure® OBT (n=2,946) and HemeSelect ® (n=1,853), neither of which are currently available in the U.S. Considering only iFOBTs that are FDA approved and currently available in the U.S., only 1 included study evaluated InSure™ (n=443) and 1 evaluated MonoHaem® (n=81); none evaluated Instant-View ® or immoCARE®. However, it is not clear that iFOBTs can be evaluated as an assay class. iFOBTs vary in their detection limit, determined by adding known quantities of fresh, human blood. For example, MonoHaem® has the highest detection limit for hemoglobin and, judging from 1 small study (n=81), poor clinical sensitivity compared to Hemoccult II®. However, the InSure™ assay reportedly has a detection limit that is 6 times lower than that of FlexSure® but in 1 study (n=443) InSure™ and FlexSure® performed equally. Thus, artificially determined detection limits may not predict comparative clinical performance. Therefore, the evidence on clinical performance of immunochemical FOBTs currently available in the U.S. compared to guaiac FOBTs is limited to 1 study comparing InSure™ to FlexSure® and 1 study comparing MonoHaem® to Hemoccult II® and is insufficient for drawing conclusions. Does the evidence indicate that compliance is more likely with an iFOBT compared to gFOBT? Is there a specific iFOBT that increases compliance? For evaluation of the effect of iFOBTs on patient compliance, all available studies were included. Only 1 study addressed the effect of sampling method on patient compliance with FOBT testing. In a population randomly chosen from electoral rolls in Australia, not in association with a health care setting, sampling from 2 versus 3 stools and sampling toilet water around a stool with a brush (InSure™) versus capturing a dry specimen for sampling with a spatula (Hemoccult SENSA®, FlexSure® OBT) was associated with significantly increased compliance. Whether or not these factors affect compliance when FOBT testing is recommended in a U.S. health care setting, e.g.,

2

©2004 Blue Cross and Blue Shield Association. Reproduction without prior authorization is prohibited.

Immunochemical versus Guaiac Fecal Occult Blood Tests

by a patient’s general physician, is not known. The evidence is insufficient to evaluate the effect of sampling method on patient compliance with FOBT testing. Several studies have tested the effect of dietary restrictions versus none for compliance with performing the same FOBT, with conflicting results. Because study settings, populations, and specific dietary restrictions differ across studies, it is not possible to draw conclusions regarding the effect of dietary restrictions on patient compliance with FOBT testing. 3. The technology must improve the net health outcome; and 4. The technology must be as beneficial as any established alternatives. There is insufficient evidence to permit conclusions regarding the use of immunochemical fecal occult blood testing for colorectal cancer screening and health outcomes. 5. The improvement must be attainable outside the investigational settings. Whether or not the use of immunochemical fecal occult blood testing for colorectal cancer screening improves health outcomes has not been demonstrated in the investigational setting. Based on the above, the use of immunochemical fecal occult blood testing for colorectal cancer screening does not meet the TEC criteria.

Contents Assessment Objective

4

Review of Evidence

11

Background

4

18

Methods

9

Summary of Application of the Technology Evaluation Criteria References

21

Appendix

24

Formulation of the Assessment

10

Published in cooperation with Kaiser Foundation Health Plan and Southern California Permanente Medical Group. TEC Staff Contributors

Author—Margaret A. Piper, Ph.D., M.P.H.; TEC Executive Director—Naomi Aronson, Ph.D.; Managing Scientific Editor—Kathleen M. Ziegler, Pharm.D.; Research/Editorial Staff—Claudia J. Bonnell, B.S.N., M.L.S.; Maxine A. Gere, M.S. Blue Cross and Blue Shield Association Medical Advisory Panel Allan M. Korn, M.D., F.A.C.P.—Chairman, Senior Vice President, Clinical Affairs/Medical Director, Blue Cross and Blue Shield Association; David M. Eddy, M.D., Ph.D.—Scientific Advisor, Senior Advisor for Health Policy and Management, Kaiser Permanente, Southern California. ■ Panel Members Peter C. Albertsen, M.D., Professor, Chief of Urology, and Residency Program Director, University of Connecticut Health Center; Edgar Black, M.D., Vice President, Chief Medical Officer, BlueCross BlueShield of the Rochester Area; Helen Darling, M.A., President, Washington Business Group on Health; Josef E. Fischer, M.D., F.A.C.S., Mallinckrodt Professor of Surgery, Harvard Medical School and Chair, Department of Surgery, Beth Israel Deaconess Medical Center—American College of Surgeons Appointee; Alan M. Garber, M.D., Ph.D., Professor of Medicine, Economics, and Health Research and Policy, Stanford University; Steven N. Goodman, M.D., M.H.S., Ph.D., Associate Professor, Johns Hopkins School of Medicine, Department of Oncology, Division of Biostatistics (joint appointments in Epidemiology, Biostatistics, and Pediatrics)—American Academy of Pediatrics Appointee; Michael A.W. Hattwick, M.D., Woodburn Internal Medicine Associates, Ltd. American College of Physicians Appointee; I. Craig Henderson, M.D., Adjunct Professor of Medicine, University of California, San Francisco; Mark A. Hlatky, M.D., Professor of Health Research and Policy and of Medicine (Cardiovascular Medicine), Stanford University School of Medicine; Bernard Lo, M.D., Professor of Medicine and Director, Program in Medical Ethics, University of California, San Francisco; Barbara J. McNeil, M.D., Ph.D., Ridley Watts Professor and Head of Health Care Policy, Harvard Medical School, Professor of Radiology, Brigham and Women’s Hospital; Brent O’Connell, M.D., M.H.S.A., Vice President and Medical Director, Pennsylvania Blue Shield/Highmark, Inc.; Stephen G. Pauker, M.D., M.A.C.P., F.A.C.C., Sara Murray Jordan Professor of Medicine, Tufts University School of Medicine; and Vice-Chairman for Clinical Affairs and Associate Physician-in-Chief, Department of Medicine, New England Medical Center; William R. Phillips, M.D., M.P.H., Clinical Professor of Family Medicine, University of Washington—American Academy of Family Physicians’ Appointee; Earl P. Steinberg, M.D., M.P.P., President, Resolution Health, Inc.; Paul J. Wallace, M.D., Executive Director, Care Management Institute, Kaiser Permanente; A. Eugene Washington, M.D., M.Sc., Executive Vice Chancellor, University of California, San Francisco; Jed Weissberg, M.D., Associate Executive Director for Quality and Performance Improvement, The Permanente Federation. CONFIDENTIAL: This document contains proprietary information that is intended solely for Blue Cross and Blue Shield Plans and other subscribers to the TEC Program. The contents of this document are not to be provided in any manner to any other parties without the express written consent of the Blue Cross and Blue Shield Association. ©2004 Blue Cross and Blue Shield Association. Reproduction without prior authorization is prohibited.

3

Technology Evaluation Center

Assessment Objective Performing an annual fecal occult blood test (FOBT) is one of several recommended options for colorectal cancer screening in the average risk population beginning at age 50. Annual or biennial screening with guaiac-based FOBTs (gFOBT) has been shown in large, randomized trials to have a significant and beneficial effect on colorectal cancer incidence and mortality. However, while the specificity of these tests is generally high, sensitivity is poor. Complicated dietary restrictions prior to testing and sampling instructions may limit patient compliance. Newer, immunochemical FOBTs (iFOBT) are reported to have improved performance characteristics compared to guaiac tests, no dietary restrictions, and in the case of 1 immunochemical test, InSure™, simpler instructions for sampling. This Assessment 1 will evaluate whether there is sufficient evidence to evaluate the performance of iFOBTs in general, or of specific iFOBTs, and to compare performance to standard gFOBTs. In addition, this Assessment will examine the evidence on patient compliance with various FOBT formats to determine if compliance is more likely with any or with a specific iFOBT versus gFOBTs.

Background Colon Cancer Epidemiology Estimates for both colorectal cancer incidence and mortality in the U.S. in 2003 ranked third among all cancers in men and women (American Cancer Society 2003). The primary risk factor is age; more than 90% of cases are diagnosed in adults over age 50. It is estimated that at age 50, a person has about a 5% remaining lifetime risk of being diagnosed with colorectal cancer. About 20% of cases occur in persons

with specific risk factors (e.g., inflammatory bowel disease), and about 6% arise from persons with uncommon genetic syndromes such as familial adenomatous polyposis (U.S. Preventive Services Task Force 2002). The incidence also is increased in individuals with a personal or family history of colorectal cancer or polyps. Colorectal cancer in the early stages is largely asymptomatic and frequently cured by surgery alone. Survival rates are much better when diagnosed and treated at an early stage (Table 1). Thus, screening for early colorectal cancer is recommended beginning at age 50 for those with no risk factors other than age. Colorectal Cancer Screening Guidelines for Adults at Average Risk The U.S. Preventive Services Task Force (2002 and Pignone et al. 2002), the American Cancer Society (Smith et al. 2003), and the American Gastroenterological Association (Winawer et al. 2003) have all published evidence-based guidelines recommending that clinicians screen men and women 50 years of age or older for colorectal cancer by one of the following methods: 1. annual fecal occult blood test; 2. flexible sigmoidoscopy every 5 years; 3. annual FOBT plus flexible sigmoidoscopy every 5 years; 4 double contrast barium enema every 5 years; or 5. colonoscopy every 10 years. However, none of the guidelines found sufficient evidence to determine which strategy is best in terms of the balance of benefits and potential harms or cost-effectiveness. All guidelines recommend earlier and more intensive surveillance for those at higher than average risk based on personal history of colorectal cancer or adenomatous polyps;

Table 1. Percent Survival after Colorectal Cancer (CRC) Diagnosis (American Cancer Society, Cancer Facts & Figures 2003)

Overall

1-year

5-year

10-year

83

62

55

Localized (37% of CRC)

90

Regional metastases

65

1 In a 2003 Bulletin, the Technology Evaluation Center (TEC) summarized a cost-effectiveness analysis prepared by the Agency for Healthcare Research and Quality (AHRQ; van Ballegooijen et al. 2003).

4

©2004 Blue Cross and Blue Shield Association. Reproduction without prior authorization is prohibited.

Immunochemical versus Guaiac Fecal Occult Blood Tests

history of a predisposing illness (e.g., inflammatory bowel disease); family history of colorectal cancer or adenomatous polyps; or a predisposing genetic mutation. Colorectal Neoplasia and Bleeding Colorectal cancers and some precancerous adenomas often bleed periodically. Consequently, small amounts of blood in the stool (fecal occult blood) in the absence of other explanatory conditions is a marker for neoplasia. One study indicated that blood loss in patients with known adenomas exceeded 2 mL/day (approximate upper limit of normal) only when adenomas were larger than 2 cm in diameter (Macrae and St. John 1982). In this study, the FOBT false-negative rate for adenomas smaller that 1 cm was 93% and for 1–2 cm adenomas was 77%. Adenomas larger than 2 cm had a false-negative rate of 25%, similar to that for cancers (31%). Grossman et al. (1989) performed colonoscopy on 402 asymptomatic patients with histories of colorectal adenomas. The rate of cancer diagnosis for 133 patients with adenomas smaller than 1 cm and no first-degree relatives with colorectal cancer was similar to age-adjusted autopsy rates of cancer, suggesting that these patients may not be at increased risk for the development of cancer. However, patients with a history of adenoma 1 cm or larger were significantly more likely to have cancer compared to autopsy data. Atkin et al. (1992) similarly documented a very low risk of cancer in patients with only a single, small, mildly to moderately dysplastic tubular adenoma. Muto et al. (1975) conducted histologic studies of adenomatous polyps and found that of those less than 1 cm in diameter only 1% had histologic evidence of malignancy. About 10% of those between 1 and 2 cm harbored malignant tissue, as did nearly 50% of those greater than 2 cm. In a study of endoscopically removed adenomas, Shinya et al. (1979) found that only 5% of adenomas had histologic evidence of malignancy whereas 15 to 27% of larger polyps did. Thus, small adenomas (i.e., 1 screen)

60

67

80

Cumulative risk of

2.6

5.1

(Not applicable)

Relative CRC mortality reduction, % (95% CI)

13 (3–22)

15 3 (1–27)

36 (0–68)

16 (7–23)

21 (0–56)

Relative CRC mortality reduction among screenees, % (95% CI)

27

30–34

(Not reported)

23 (11–43)

64

Absolute risk reduction for death from CRC per 1,000

1.2 (0.3–2.1)

1.7 (0.1–3.3)

1.2 (0.64–3.0)

0.85 (0.4–1.35)

2.7 (0.4–9.2)

Number needed to screen, over follow-up time

840 (476–3,605)

595 (306–10,781)

833 (329–∞)

1173 (741–2,807)

370 (108–∞)

67

71.5 9.1

a positive test, %

1 2 3

Includes listed studies plus controlled trials of rehydrated Hemoccult® not listed here All patients were offered sigmoidoscopy Including complications from treatment

Technology Evaluation Center

8 Table 3. Effect of Non-rehydrated gFOBT on Mortality from Colorectal Cancer in Controlled Clinical Trials

Immunochemical versus Guaiac Fecal Occult Blood Tests

et al. 2000) a questionnaire was mailed to U.S. gastroenterologists chosen at random from a national database; responses were obtained from 1,828 (24%). The FOBT tests of choice were Hemoccult II® (72%) or Hemoccult SENSA® (22%) and 78% of respondents reported providing patients with advice about dietary restrictions before either performing or ordering an FOBT. However, 86% reported performing FOBTs on a single stool specimen obtained from digital rectal exam. Similarly, Nadel et al. (2002), evaluating results of the National Health Interview Survey reported that approximately half of FOBT testing is performed with single samples taken during a physical exam rather than with the home kit. A single in-office FOBT is likely to be less sensitive than the FDA-approved 3-card homeperformed FOBT because only one sample is taken (Pignone et al. 2002) and evidence substantiating its use is lacking. Based on these results, it seems possible that physicians may not rigidly adhere to guidelines regarding the patients’ performance of FOBTs. Moreover, physicians frequently use the FOBT for reasons other than colorectal cancer screening, such as for hematemesis, melena, heartburn, or dyspepsia (Sharma et al. 2000), for which the test has not been validated. Despite the apparent lack of consistent use in practice, the use of FOBT can be accurately evaluated only when it is used for validated indications (colorectal cancer screening) and with the manufacturer’s approved performance instructions. Screening Recommendations and FOBT Types In April 2002 the American Cancer Society Colorectal Cancer Advisory Group concluded that “the evidence showing improved specificity with immunochemical tests, and the lack of requirements to adhere to dietary restrictions prior to the test, was sufficiently persuasive to update the guideline ... to include the following statement: in comparison with guaiac-based tests for the detection of occult blood, immunochemical tests are more patient-friendly, and are likely to be equal or better in sensitivity and specificity” (Smith et al. 2003). The screening guidelines of the American Gastroenterological Association note only that “newer guaiac-based and immunochemical tests are available that have improved sensitivity and appear to maintain acceptable

specificity (Winawer et al. 2003). The USPSTF evidence review for its guidelines refers only to guaiac-based tests used in studies of FOBT screening and mortality outcomes (Pignone et al. 2002). A joint committee representing the World Health Organization and the World Organization for Digestive Endoscopy published a report on choice of FOBT for colorectal cancer screening (Young et al. 2002). The Hemoccult SENSA® is the recommended gFOBT due to low cost, greater sensitivity than Hemoccult II®, but better specificity than rehydrated Hemoccult II®. Reliable compliance but poor colonoscopy resources are more compatible with the higher specificity of Hemoccult II®. However, if compliance is uncertain or unreliable but colonoscopy resources are sufficient, iFOBT may be considered.

Methods Search Methods An initial search strategy of “(fecal OR faecal) AND immuno* AND (colon OR cancer)” yielded 139 citations from 1985 through June 2004. Titles and abstracts were reviewed; full copies of potentially relevant papers published in the English language were retrieved for review. The electronic search was also supplemented with hand searches of prior reviews in this area, recent articles, and the Cochrane Collaboration Library; expert contacts also identified relevant studies. Study Selection For Key Question 1, included studies were published in peer-reviewed journals and met the minimum or preferred criteria outlined in Table 4. In particular, only studies that compared at least 2 FOBTs, one of them being an iFOBT, were included so that performance characteristics could be compared without confounding by different patient populations. Screening test performance characteristics are most accurate when obtained in studies that enroll the target population of randomly selected individuals at average risk of colorectal cancer. Because gFOBTs require dietary restrictions, populations representative of both geographical and cultural diversity are ideal. However, in these studies, yield of neoplasia is low and large numbers of enrollees are

©2004 Blue Cross and Blue Shield Association. Reproduction without prior authorization is prohibited.

9

Technology Evaluation Center

Table 4. Minimum and Preferred Criteria for Study Inclusion Minimum

Preferred

Enrolled a well-described patient group for CRC screening; actual CRC/adenoma status unknown

Population or community-based participant selection Multi-ethnic, geographically diverse screening population

Patients with known causes of bleeding excluded FOBT successfully completed by at least 85% of enrollees Endoscopy for all

Colonoscopy for all

Compares 2 or more FOBTs including

Includes Hemoccult II® as 1 of the assays compared

at least 1 iFOBT No rehydration for Hemoccult II ® tests Reports FOBT results for cancer, and adenomas >1 cm, or both combined, separately from adenomas 1 cm). In all but 391 patients (see Robinson et al. 1995), the FOBT tests compared in each study were run on each patient and the results are matched by patient. Thus, statistical comparisons of proportions from independent samples are inappropriate for determining significant differences between performance characteristics such as sensitivity and specificity. Rather, McNemar’s test, which takes paired data into account, is most often used in this situation. However, none of the included studies presented raw data in a format that allowed McNemar’s test to be conducted. Two studies compared sensitivities and specificities by 14

McNemar’s testing and reported the results. Young et al. (2003) found no significant difference in any parameter between InSure™ and FlexSure® OBT iFOBTs. Greenberg et al. (2000) reported that sensitivity results for neoplasia by Hemoccult SENSA®, FlexSure®, and HemeSelect® were not significantly different from each other, but all were significantly greater than Hemoccult II®; for specificity, FlexSure® OBT was significantly lower than Hemoccult II®. For the rest of the studies, assays can only be compared by best results within each study. Table 7 shows the best result for sensitivity and for specificity among assays compared within each study. Best values for positive predictive value were obtained in nearly all cases by the same assay as had the best specificity (see Appendix, Tables A and B), both being driven by the false-positive rate. This simple presentation suggests the following conclusions regarding the comparative performance of gFOBTs and iFOBTs: ■

■

iFOBTs have better sensitivity than Hemoccult II® but not necessarily better sensitivity than Hemoccult SENSA®, and; iFOBTs have better specificity than Hemoccult SENSA®, but specificity is not clearly as good as or better than Hemoccult II®.

However, this overall comparison assumes that iFOBTs as an assay class perform similarly. As shown in Table 2, this may not be the case. iFOBTs vary in their detection limit, determined by adding known quantities of fresh, human blood. For example, MonoHaem® has the highest detection limit for hemoglobin and, judging from 1 small study (Gopalswamy et al. 1994), poor clinical sensitivity compared to Hemoccult II®. However, the InSure™ assay reportedly has a detection limit that is 6 times lower than that of FlexSure® but in 1 study (Young et al. 2003) InSure™ and FlexSure® performed equally. Thus, artificially determined detection limits may not predict comparative clinical performance. The majority of published comparative data on iFOBTs are derived from studies of FlexSure® OBT (n=2,946) and HemeSelect® (n=1,853), neither of which are currently available in the U.S. (Table 2). Only 1 study (Young et al. 2003; n=443) links the currently available InSure™ assay to other iFOT data via

©2004 Blue Cross and Blue Shield Association. Reproduction without prior authorization is prohibited.

gFOBTs

iFOBTs

Meet Current Guidelines

Available

Not available in U.S.

in U.S. Citation

n

HO

HOS

IS

FS

HS

MH

Sensitivity for Neoplasia Rozen et al. 1997

403

x

x

x

✓

Greenberg et al. 2000*

554

x

x

✓

x

Gopalswamy et al. 1994

81

x

x

Robinson et al. 1995

417 + 391†

x

Rozen et al. 2000

1,410

✓

x

Wong et al. 2003a

135

✓

x

Young et al. 2003*

443

✓

x

✓

x

✓

Specificity for Neoplasia Rozen et al. 1997

403

x

x

x

✓

Greenberg et al. 2000*

554

✓

x

x

x

Gopalswamy et al. 1994

81

✓

x

Robinson et al. 1995

417 + 391†

✓

Rozen et al. 2000

1,410

x

✓

Wong et al. 2003a

136

x

✓

Young et al. 2003*

443

✓ x

15

†

x

x

✓

=

✓

Best value among assays compared in each study Other assays compared in study

Abbreviations: HO Hemoccult II® HOS Hemoccult SENSA® FS FlexSure® OBT *

x

IS HS MH

InSure™ HemeSelect® MonoHaem®

Only Young et al. 2003 and Greenberg 2000 performed appropriate significance testing (McNemars test for matched pairs). For Greenberg et al. 2000, sensitivity results for neoplasia by Hemoccult SENSA®, FlexSure®, and HemeSelect® were not significantly different from each other, but all were significantly greater than Hemoccult II®; for specificity, FlexSure® OBT was significantly lower than Hemoccult II®. For Young et al., FlexSure® OBT and InSure™ did not differ significantly for either parameter. Data not available to perform significance testing for other studies. Hemoccult® performed on 417 patients; HemeSelect® performed on 417 + 391 patients.

Immunochemical versus Guaiac Fecal Occult Blood Tests

©2004 Blue Cross and Blue Shield Association. Reproduction without prior authorization is prohibited.

Table 7. Highest Value for Sensitivity and Specificity for Neoplasia (Cancer or Adenoma >1 cm) among Assays Compared in Each Included Study

Technology Evaluation Center

comparison to FlexSure®, with which it shares statistically similar performance characteristics. Only one included study evaluated MonoHaem® (n=81) and none evaluated Instant-View® or immoCARE®, all of which are FDA-approved and currently available in the U.S. Summary The evidence on clinical performance of immunochemical FOBTs currently available in the U.S. compared to guaiac FOBTs is limited to 1 study comparing InSure™ to FlexSure® (Young et al. 2003; n=443) and 1 study comparing MonoHaem® to Hemoccult II® (Gopalswamy et al. 1994; n=81) and is insufficient for drawing conclusions. Does the evidence indicate that compliance is more likely with an iFOBT compared to gFOBT? Is there a specific iFOBT that increases compliance? Study Inclusion and Overview Three individual studies and one meta-analysis of 5 other randomized trials were evaluated for this Assessment and are summarized in Table 8. One study (Cole et al. 2003) evaluated FOBT compliance with Hemoccult SENSA® (dietary restrictions) compared to FlexSure® OBT (no dietary restrictions, same sampling method) and to InSure™ (no dietary restrictions, different sampling method). The other 2 studies (Cole et al. 2001; King et al. 1992) and the meta-analysis (Pignone et al. 2001) all evaluated FOBT compliance with versus without dietary restrictions for the same assay. Studies reported by Cole et al. (2001 and 2003) and by King et al. (1992) were all conducted in Australia and none in a clinical setting. Cole et al. recruited patients from electoral rolls (2001) or from an epidemiology database of individuals willing to be contacted about health-related issues (2003) and enrolled approximately 600 participants per study arm. In a smaller study of approximately 400 total participants, King et al. (1992) approached residents in a defined geographic area but did recruit names from randomly chosen general practices in the area and included a letter from each patient’s general practitioner encouraging participation in the initial invitation to screening. Each study incorporated some form of random selection of participants and resulted in essentially matched study arms. Dietary restrictions were similar across these 3 studies and advocated avoidance

16

of red meat, certain raw fruits and vegetables, aspirin or nonsteroidal anti-inflammatory drugs, and vitamin C during the testing process. The randomized trials evaluated by Pignone et al. (2001) were conducted in the U.S. (2 studies, n=3,783 and 786), Canada (1 study, n=5003), or the U.K. (2 studies, n=634 and 153). Four trials were conducted in health care settings. In a fifth, members of the American Association of Retired Persons received an invitation to perform FOBT screening by mail or at a group meeting. The specific dietary restrictions differed among studies in complexity and in specific items; only avoidance of red meat during testing was common to all studies. Hemoccult II® assays were used in all studies. Evidence Evaluation Cole et al. (2003) reported a baseline FOBT compliance rate of 23.4% for Hemoccult SENSA® with dietary restrictions and sample collection by spatula from 3 dry stool specimens. InSure™ requires only 2 stool samples taken with a brush from toilet water around the stool; compliance was significantly better than with FlexSure® (30.5%) or Hemoccult® SENSA at 39.6%. Cole et al. (2003) was the only study addressing the effect of sampling method. Although encouraging, the significance of these results (derived from participants randomly chosen from electoral rolls in Australia) for clinical settings in the U.S., where FOBT testing is most likely to be administered, is not clear. The absolute compliance rates were much lower than those seen in the other studies, which enrolled either epidemiology database members who had already indicated an interest in health care issues, or participants recruited in or via health care settings. Compliance in a screening setting not associated with a patient’s usual health care may be more influenced by factors such as sampling method. Whether or not compliance with recommendations for FOBT colorectal cancer screening administered in a clinical setting, e.g., by a patient’s primary care physician, would be similarly influenced by sampling method is unknown. Cole et al. (2003) also reported that compliance was significantly improved to 30.5% with FlexSure®, which does not have dietary restrictions but the same sampling method as Hemoccult SENSA®. Cole et al. (2001) found

©2004 Blue Cross and Blue Shield Association. Reproduction without prior authorization is prohibited.

Citation Cole et al. 2003

Cole et al. 2001

King et al. 1992

Pignone et al. 2001

* † ‡ **

Population (active colonic disease or bleeding excluded) Individuals aged 50–69 randomly selected from the electoral roll for certain urban postcodes in Australia In U.S., initiated in M.D. office—may influence compliance—this study setting artificial for this country

n Eligible First 606 Second 606 Third 606

Demographic Comparison Matched* except more were economically advantaged in FlexSure ® cohort (p=0.047)

Individuals aged 50–69 randomly selected from the S. Australian Dept of Human Services Behavioral Epidemiology Database

602

(random assignment)

Matched* except more were economically disadvantaged in No Diet Restrictions cohort (p=0.04)

General practitioner patient lists in the Southern Sydney Area Health Service, Australia

199

Matched*

Meta-analysis of effect

5018

of dietary restrictions on FOBT participation in 5 randomized trials

5341

601

FOBT Type

Diet/drug Restrictions

# Stools to Sample Sampling Method

% Participation

Hemoccult SENSA®

Yes

3

Spatula, avoid toilet water

23.4

FlexSure® OBT®

No

3

Spatula, avoid toilet water

30.5†

InSure™

No

2

Brush, sample stool immersed in toilet water

39.6†,‡

FlexSure ® OBT

Yes

3

Spatula, avoid toilet water

53.3

FlexSure ® OBT

No

3

Spatula, avoid toilet water

65.9†

Hemoccult SENSA®

Yes

3

Spatula, avoid toilet water

51.4

No

3

Spatula, avoid toilet water

59.8

Yes

3

Spatula, avoid toilet water

51.2

190 (All Hemoccult ®)

No

51.9**

By gender, age, and, for Cole et al. (2001) and Cole et al. (2003), by socio-economic index. p