Volume 43, Issue 7 p. 755-764

Systematic Review with Meta-analysis

Free Access

Systematic review with meta-analysis: faecal occult blood tests show lower colorectal cancer detection rates in the proximal colon in colonoscopy-verified diagnostic studies

H. W. Hirai,

School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

Stanley Ho Big Data Decision Analytics Research Centre, The Chinese University of Hong Kong, Shatin, Hong Kong

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

K. K. F. Tsoi,

Corresponding Author

School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

Stanley Ho Big Data Decision Analytics Research Centre, The Chinese University of Hong Kong, Shatin, Hong Kong

Correspondence to:

Dr K. K. F. Tsoi, 4/F, School of Public Health and Primary Care, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong.

E-mail: kelvintsoi@cuhk.edu.hk

Search for more papers by this author

J. Y. C. Chan,

School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

S. H. Wong,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

J. Y. L. Ching,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

M. C. S. Wong,

School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

J. C. Y. Wu,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

F. K. L. Chan,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

J. J. Y. Sung,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

S. C. Ng,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

H. W. Hirai,

School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

Stanley Ho Big Data Decision Analytics Research Centre, The Chinese University of Hong Kong, Shatin, Hong Kong

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

K. K. F. Tsoi,

Corresponding Author

School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

Stanley Ho Big Data Decision Analytics Research Centre, The Chinese University of Hong Kong, Shatin, Hong Kong

Correspondence to:

Dr K. K. F. Tsoi, 4/F, School of Public Health and Primary Care, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong.

E-mail: kelvintsoi@cuhk.edu.hk

Search for more papers by this author

J. Y. C. Chan,

School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

S. H. Wong,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

J. Y. L. Ching,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

M. C. S. Wong,

School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

J. C. Y. Wu,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

F. K. L. Chan,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

J. J. Y. Sung,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

S. C. Ng,

Institute of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Department of Medicine and Therapeutics, LKS Institute of Health Science, State Key Laboratory of Digestive Disease, The Chinese University of Hong Kong, Shatin, Hong Kong

Search for more papers by this author

First published: 09 February 2016

https://doi.org/10.1111/apt.13556

Citations: 42

As part of AP&T's peer-review process, a technical check of this meta-analysis was performed by Dr Yuan. This article was accepted for publication after full peer-review.

Summary

Background

The performance of faecal occult blood tests (FOBTs) to screen proximally located colorectal cancer (CRC) has produced inconsistent results.

Aim

To assess in a meta-analysis, the diagnostic accuracy of FOBTs for relative detection of CRC according to anatomical location of CRC.

Methods

Diagnostic studies including both symptomatic and asymptomatic cohorts assessing performance of FOBTs for CRC were searched from MEDINE and EMBASE. Primary outcome was accuracy of FOBTs according to the anatomical location of CRC. Bivariate random-effects model was used. Subgroup analyses were performed to evaluate test performance of guaiac-based FOBT (gFOBT) and immunochemical-based FOBT (iFOBT).

Results

Thirteen studies, with 17 cohorts, reporting performance of FOBT were included; a total of 26 342 patients (mean age 58.9 years; 58.1% male) underwent both colonoscopy and FOBT. Pooled sensitivity, specificity, positive likelihood ratio and negative likelihood ratio of FOBTs for CRC detection in the proximal colon were 71.2% (95% CI 61.3–79.4%), 93.6% (95% CI 90.7–95.7%), 11.1 (95% CI 7.8–15.8) and 0.3 (95% CI 0.2–0.4) respectively. Corresponding findings for CRC detection in distal colon were 80.1% (95% CI 70.9–87.0%), 93.6% (95% CI 90.7–95.7%), 12.6 (95% CI 8.8–18.1) and 0.2 (95% CI 0.1–0.3). The area-under-curve for FOBT detection for proximal and distal CRC were 90% vs. 94% (P = 0.0143). Both gFOBT and iFOBT showed significantly lower sensitivity but comparable specificity for the detection of proximally located CRC compared with distal CRC.

Conclusion

Faecal occult blood tests, both guaiac- and immunochemical-based, show better diagnostic performance for the relative detection of colorectal cancer in the distal colon than in the proximal bowel.

Introduction

Colorectal cancer (CRC) is the third most common cancer and the second most common cause of cancer deaths worldwide.1 It is one of the main cancers that is preventable with screening. Although colonoscopy is regarded as the most effective screening tool, it can be invasive and labour-intensive. Faecal occult blood test (FOBT) is non-invasive and easily performed at home or in clinics; it is more suitable for population-based screening especially in resource-limited countries.2 Annual screening with FOBT has been demonstrated to reduce CRC incidence by 20% and mortality by 33%.3, 4

Guaiac-based FOBT (gFOBT) and immunochemical-based FOBT (iFOBT) are two commonly used FOBTs. As gFOBT is non-specific to human haemoglobin, dietary restriction is needed before stool sample collection5; whereas, iFOBT has minimal restrictions.6 The diagnostic performance of iFOBT has been shown better than that of gFOBT7-10; however, some reported FOBT detection rates might not be verified by colonoscopic result.11-14 These studies have been included in meta-analyses to evaluate the diagnostic performance of FOBT15, 16; verification bias might be induced and potentially leaded to an overestimation of sensitivity of FOBT by about 34%.17

Studies assessing whether the performance of FOBT is more accurate in the distal colon compared with the proximal colon have produced conflicting results,18-21 partly due to small sample size. This systematic review and meta-analysis pooled together colonoscopy-verified cohorts to assess diagnostic accuracy of FOBTs on the relative detection of CRC in the proximal and distal colon respectively.

Methods

This systematic review was performed based on standard guidelines for conducting and reporting systematic review of diagnostic studies, including Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) 22 and guidelines from Cochrane Diagnostic Test Accuracy Working Group.23, 24

Search strategy

A list of FOBTs with the commercial brands was identified from a general search engines in Google. Literature search was extended to the academic database of MEDLINE and EMBASE from the earliest available dates stated in the individual database until 30 April 2015. Each screening test was separately searched using keywords including colorectal cancer, faecal occult blood test and colonoscopy. Diagnostic studies comparing accuracy of FOBTs were manually identified from title or abstract preview of all search records. The selection was limited to peer-reviewed articles published in English language. Literature search was also extended to Google Scholar. Manual search was also performed to the bibliographies of review articles and any research studies included in the review.

Inclusion and exclusion criteria

Cross-sectional studies were included if they met the following inclusion criteria: (i) involved patients using FOBT to detect CRC; (ii) used colonoscopy as the gold standard to confirm the diagnosis of cancer and healthy colon and (iii) reported FOBT result, or data that could derive sensitivity and specificity, of CRC and healthy colon. Studies from both symptomatic and asymptomatic patients were included without restriction to screening setting. Studies were excluded if they were not written in English.

Data extraction

Two investigators (HWH and JYC) independently assessed the relevance of search results and collected data into a data extraction form. Year of publication, study location, number of patients, mean age, percentage of males, number of patients with normal colonoscopy, number of patients diagnosed with CRC, cut-off values for qualitative FOBTs were extracted. Diagnostic measurement such as sensitivity, specificity, positive likelihood ratio (positive LR), negative likelihood (negative LR), true-positive, false-positive, true-negative and false-negative values were recorded as well. Sensitivity referred to number of patients with CRC divided by number of patients with positive FOBT result; specificity referred to number of patients with normal colonoscopy divided by number of patient with negative FOBT result.

The primary outcome of this study was the accuracy of FOBTs in the detection of CRC according to anatomical location (proximal vs. distal colon). Subgroup analysis was performed to compare diagnostic performance of gFOBT and iFOBT in the proximal vs. the distal colon. All included patients have undergone colonoscopy with confirmed histology. For studies reporting more than one FOBTs specimen, results from the first specimen were included. For quantitative FOBT with varied cut-off values, results were adopted based on that particular study's cut-off recommendation. When discrepancies were found, a third investigator (KKT) made the definitive decision for study eligibility and data extraction. In studies which have recruited cancer patients, stool collection might take place after colonoscopy; or following endoscopic biopsy of polyps or cancer.

Risk of bias and reporting quality

Potential risks of bias in each screening test were evaluated by QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies 2 instrument),25 which assessed patient selection, execution of index test and the reference standard, and flow of patients. All studies with high risk of bias were calculated and presented as a percentage. An 8-point scale was designed for the evaluation of reporting quality, including the description of: (i) study population, (ii) participant recruitment, (iii) sampling of participant selection, (iv) data collection plan, (v) reference standard and its rationale, (vi) technical specifications, (vii) rationale for units and cut-offs, and (viii) methods for calculating of diagnostic accuracy with confidence intervals. This quality score was presented as a median and range.

Data synthesis and statistical analysis

Statistical analyses were performed using statistical software package Stata (Version 11). Bivariate random-effects model was applied to pool the sensitivity and specificity from each included studies in order to assess FOBT accuracy for the detection of CRC.26 Forest plots were used to present the pooled sensitivity and specificity. In order to trade-off between sensitivity and specificity, the diagnostic odds ratio (DOR) and area-under-curve were used as a single indicator of test performance.27 A hierarchical summary receiver-operating characteristic (HSROC) curve was generated to present the summary estimates of sensitivities and specificities along with their corresponding 95% CI and prediction region.28 The area under the HSROC curve (AUC) was calculated and an area between 0.9 and 1.0 indicated that the diagnostic accuracy was good.29 When the Hessian matrix of bivariate random-effects approach was unstable or asymmetric, a random effect model following the approach of DerSimonian and Laird was applied to estimate the pooled sensitivity and specificity, and a Moses–Littenberg summary receiver operating characteristic (SROC) curve would be generated to present the summary estimates of sensitivity and specificity with AUC.30, 31 Statistical heterogeneity among the trials was assessed by I², which described the percentage of total variation across studies due to the heterogeneity rather than chance alone. P < 0.1 was considered as statistical significant.

Results

Literature search and study selection

A total of 5756 abstracts were identified and 153 potential studies were further investigated from the bibliographies. All titles and abstracts were screened; 230 articles were relevant to the use of FOBTs for the detection of CRC. Studies were excluded for the following reasons: studies lacked details on sensitivity and specificity (n = 98); studies related to other stool tests such as stool DNA test and other biomarkers (n = 42); studies report FOBT results but no colonoscopic findings (n = 8); studies did not report FOBT results of patients with healthy colon or studies did not use normal colonoscopy patients as reference group when reporting specificity (n = 35); duplicated studies (n = 3); did not report FOBT results according to anatomical location of CRC (n = 30), and a study evaluated FOBT results on overall gastrointestinal cancer (Figure 1). A total of 13 colonoscopy-verified diagnostic studies, with 17 cohorts, have been included to assess the diagnostic accuracy of FOBT in the proximal and distal colon (Table 1). Included studies were published between 1992 and 2014 from eight countries [Australia (n = 2), Denmark (n = 1), Germany (n = 1), Hong Kong (n = 1), Japan (n = 2), South Korea (n = 2), Taiwan (n = 2) and UK (n = 2)].

**Figure 1**
Open in figure viewer PowerPoint

Summary of literature search. CRC, colorectal cancer; FOBT, faecal occult blood test; gFOBT, guaiac-based faecal occult blood test; iFOBT, immunochemical-based faecal occult blood test.

Table 1. Characteristics of included studies

Ref	Study_ID	Country	Study period	Male (%)	Mean age	No. of subject				Screening	FOBT brandb^b FOBT brands: 1 = FlexSure OBT; 2 = HemeSelect; 3 = Hemoccult; 4 = Hemoccult II; 5 = Hemoccult Sensa; 6 = InSure; 7 = Nescauto Hemo Plus; 8 = OC-Hemocatch; 9 = OC-Light; 10 = OC-Sensor; 11 = PreventID-CC.		No. of cohorts
Ref	Study_ID	Country	Study period	Male (%)	Mean age	Total	CRC	Normal	Othersa^a Others: all colonic findings other than cancer and normal such as IBD, hyperplastic or adenoma.	Screening	gFOBT	iFOBT	No. of cohorts
Studies for primary outcome analysis
32	Bjerregaard 2009	Denmark	2002–2003	42.2%	63c^c Age is represented as median.	256	8	120	128	No	5		1
33	Chiu 2013	Taiwan	2005–2010	59.2%	59.8	18297	28	14252	4017	Yes		9	1
34	Hope 1996	Australia	1991–1992	40.0%	52.9	160	3	136	21	No	4		1
35	Kaul 2013	United Kingdom	NA	45.5%	66c^c Age is represented as median.	126	17	36	73	No		9	1
36	Kim 2014	South Korea	NA	NA	58.2	326	175	151	0	No		10	1
37	Koga 2013	Japan	2009–2012	55.8%	65c^c Age is represented as median.	224	117	107	0	No		8	1
38	Lee 2013	Taiwan	2011–2013	60.5%	53	3172	39	2965	168	Yes	5	10	2
39	Oono 2010	Japan	2007–2008	52.0%	65	1073	91	758	224	No		7	1
40	Park 2010	South Korea	2007–2008	51.4%	59.3	770	13	479	278	Yes	4	10	2
41	Shastri 2008	Germany	2003–2007	65.0%	54.6	640	55	516	69	No		11	1
42	Sung 2003	Hong Kong	NA	44.0%	56.5	505	4	353	148	Yes	4		1
43	Thomas 1992	United Kingdom	NA	60.2%	69c^c Age is represented as median.	350	50	252	48	No	3	2	2
44	Young 2003	Australia	1999–2001	46.5%	NA	443	36	179	228	No		1, 6	2

CRC, colorectal cancer; FOBT, faecal occult blood test; gFOBT, guaiac-based faecal occult blood test; iFOBT, immunochemical-based faecal occult blood test; Ref, reference number.
^a Others: all colonic findings other than cancer and normal such as IBD, hyperplastic or adenoma.
^b FOBT brands: 1 = FlexSure OBT; 2 = HemeSelect; 3 = Hemoccult; 4 = Hemoccult II; 5 = Hemoccult Sensa; 6 = InSure; 7 = Nescauto Hemo Plus; 8 = OC-Hemocatch; 9 = OC-Light; 10 = OC-Sensor; 11 = PreventID-CC.
^c Age is represented as median.

Study characteristics

This meta-analysis pooled 13 studies,32-38, 18, 8, 39-41, 19 17 cohorts (6 gFOBT; 11 iFOBT), reporting the diagnostic performance of FOBT according to the anatomical location of CRC. A total of 26 342 patients were included with mean age 58.9 years and 58.1% were male. A total of 636 patients (2.4%) were diagnosed with CRC (Table 1). Twelve of 13 studies (92.3%) had good reporting quality with a score 7 and 8. The risk of bias of the included studies was assessed by QUADAS-2; no study was assigned with high risk of bias. The potential source of the risk of bias was unknown ‘Flow and timing’ between performance of FOBT and colonoscopy.

Diagnostic performance of proximally located colorectal cancer

For FOBT diagnostic performance in the proximal colon, the pooled sensitivity, specificity, positive LR, and negative LR were 71.2% (95% CI = 61.3–79.4%), 93.6% (95% CI = 90.7–95.7%), 11.1 (95% CI = 7.8–15.8) and 0.3 (95% CI = 0.2–0.4) respectively (Table 2). For the six gFOBT cohorts, the pooled sensitivity, specificity, positive LR and negative LR of were 62.6% (95% CI = 34.9–83.9%), 90.0% (95% CI = 84.2–93.8%), 6.3 (95% CI = 3.4–11.5) and 0.4 (95% CI = 0.2–0.8) (Figure 2a). Corresponding findings for CRC detection in the 11 iFOBT cohorts were 71.1% (95% CI = 60.9–79.6%), 95.2% (95% CI = 92.1–97.1%), 14.7 (95% CI = 9.5–22.7) and 0.3 (95% CI = 0.2–0.4) (Figure 3a).

Table 2. Summary of diagnostic performance of faecal occult blood test according to anatomical location of colorectal cancer

Types of FOBT	No. of cohorts	No. of subjecta^a Subjects with CRC or normal colon but without FOBT result were excluded.	CRC		Pooled sensitivity	Pooled specificity	Pooled positive LR	Pooled negative LR	AUC		DOR
Types of FOBT	No. of cohorts		CRC		(95% CI)	(95% CI)	(95% CI)	(95% CI)	%	P-value	DOR
Proximal CRC
gFOBT	6	3770	47	1.2%	62.6% (34.9–83.9%)	90.0% (84.2–93.8%)	6.3 (3.4–11.5)	0.4 (0.2–0.8)	88%	Refb^b P-value of proximal versus distal of corresponding tests.	15.1
iFOBT	11	20 148	200	1.0%	71.1% (60.9–79.6%)	95.2% (92.1–97.1%)	14.7 (9.5–22.7)	0.3 (0.2–0.4)	91%	Refb^b P-value of proximal versus distal of corresponding tests.	48.5
Overall	17	23 918	247	1.0%	71.2% (61.3–79.4%)	93.6% (90.7–95.7%)	11.1 (7.8–15.8)	0.3 (0.2–0.4)	90%	Refb^b P-value of proximal versus distal of corresponding tests.	36.2
Distal CRC
gFOBT	6	3793	70	1.8%	75.4% (30.4–95.6%)	90.0% (84.3–93.8%)	7.5 (3.7–15.3)	0.3 (0.1–0.3)	92%	0.027b^b P-value of proximal versus distal of corresponding tests.	27.6
iFOBT	11	20 397	449	2.2%	79.0% (69.2–86.3%)	95.2% (92.1–97.1%)	16.4 (10.5–25.5)	0.2 (0.1–0.3)	95%	0.014b^b P-value of proximal versus distal of corresponding tests.	74.4
Overall	17	24 190	519	2.1%	80.1% (70.9–87.0%)	93.6% (90.7–95.7%)	12.6 (8.8–18.1)	0.2 (0.1–0.3)	94%	0.014b^b P-value of proximal versus distal of corresponding tests.	59.5

AUC, area under curve; CI, confidence interval; CRC, colorectal cancer; DOR, diagnostic odds ratio; FOBT, faecal occult blood test; gFOBT, guaiac-based faecal occult blood test; iFOBT, immunochemical-based faecal occult blood test.
17 cohorts from 13 studies have reported FOBT accuracy regarding the CRC location.
^a Subjects with CRC or normal colon but without FOBT result were excluded.
^b P-value of proximal versus distal of corresponding tests.

**Figure 2**
Open in figure viewer PowerPoint

Forrest plots for the pooled sensitivity and specificity of guaiac-based faecal occult blood test on the detection of colorectal cancer.

**Figure 3**
Open in figure viewer PowerPoint

Forrest plots for the pooled sensitivity and specificity of immunochemical-based faecal occult blood test on the detection of colorectal cancer.

Diagnostic performance of distally located colorectal cancer

For FOBT diagnostic performance in the distal colon, the pooled sensitivity, specificity, positive LR and negative LR were 80.1% (95% CI = 70.9–87.0%), 93.6% (95% CI = 90.7–95.7%), 12.6 (95% CI = 8.8–18.1) and 0.2 (95% CI = 0.1–0.3) respectively (Table 2). Corresponding findings for six gFOBT cohorts were 75.4% (95% CI = 30.4–95.6%), 90.0% (95% CI = 84.3–93.8%), 7.5 (95% CI = 3.7–15.3) and 0.3 (95% CI = 0.1–0.3) (Figure 2b). For the 11 iFOBT cohorts, the pooled sensitivity, specificity, positive LR and negative LR were 79.0% (95% CI = 69.2–86.3%), 95.2% (95% CI = 92.1–97.1%), 16.4 (95% CI = 10.5–25.5) and 0.2 (95% CI = 0.1–0.3) (Figure 3b).

Comparison of diagnostic performance of faecal occult blood tests for colorectal cancer detection in the proximal vs. distal Colon

The performance of FOBT performance were compared; the confidence regions of the HSROC curve for sensitivity and specificity of the proximal CRC detection were plotted with reference to that of the HSROC curve of distal CRC detection (Figure 4). The pooled summary points of HSROC curves showed comparable specificity between proximal (93.6%) and distal (93.6%), but proximal CRC detection had significantly lower sensitivity than that of distal (71.2% vs. 80.1%). The AUC of the proximal and distal CRC detection were 90% vs. 94% with P = 0.014. For gFOBT, AUC for the proximal and distal CRC detection were 88% vs. 92% P = 0.027, while corresponding AUC for iFOBT were 91% vs. 95% with P = 0.014 (Table 2). The diagnostic performance of overall FOBT, gFOBT and iFOBT were found significantly better in the distal colon than in the proximal colon.

**Figure 4**
Open in figure viewer PowerPoint

Hierarchical summary receiver operating characteristic curve for the detection of colorectal cancer in the proximal and distal colon using faecal occult blood test.

Discussion

This meta-analysis included 13 studies, 17 cohorts and 26 342 patients assessing the diagnostic accuracy of FOBT for the detection of CRC according to anatomical location. To the best of our knowledge, this is the first systematic review to evaluate the accuracy of FOBTs according to different locations of CRC and has only included colonoscopy-verified diagnostic studies. FOBT, both iFOBT and gFOBT, showed better diagnostic performance for the detection of CRC in the distal than in the proximal colon.

Both iFOBT and gFOBT showed higher sensitivity for CRC in the distal colon than the proximal colon. These findings may be explained by the nature of the proximal colon; lesions in the right-sided colon are usually nonpolypoid or flat, which may associate with less bleeding. As FOBT was developed to detect blood in the stool, nonpolypoid lesion in the proximal colon may induce a lower detection rate. In a Taiwanese study, iFOBT produced a high rate of false-negative results for patients with small or nonpolypoid adenomas.42 Moreover, the sensitivity of FOBT would be decayed along with the prolonged time of specimen collection time.43

The inclusion of a colonoscopy-verified studies and large sample size are of the main strengths of this meta-analysis. Verification bias is diminished and overestimation of sensitivity of FOBT is reduced by about 34%.17 Using FOBT results from normal colon as reference to calculate the specificity would better compare diagnostic performance in the proximal and distal colon. We also performed subgroup analysis to compare diagnostic performance of gFOBT and iFOBT in the proximal vs. the distal colon.

This study has some limitations. First, the detection rates of CRC in the proximal or distal colon may be different across different age, gender, race and number of samples. A previous study showed that age had a greater effect on proximal CRC44; it may affect the diagnostic performance in different anatomic location. Secondly, there exist numerous brands of FOBTs.45 Diagnostic performance may be verified from brands to brands; however, subgroup analysis of FOBT brands could not be performed due to lack of studies. Third, this study included both symptomatic and asymptomatic patients; pooled results may not reflect FOBT performance under screening setting. Inclusion of symptomatic patients might overestimate the sensitivity. Also, for studies recruiting cancer patients, stool collection might take place after colonoscopy; endoscopic biopsy of polyps or cancers may cause bleeding and induce overestimation of sensitivity. Fourth, as FOBT aim at CRC detection, this study evaluated diagnostic performance of FOBTs for CRC detection only; detection of advanced adenoma was not performed in this meta-analysis. The stage of cancer affects the performance of FOBT; however, it was not well-documented in primary studies. Subgroup analysis of stage of cancer was not available; results in this meta-analysis do not reflect early stage disease. Finally, some unpublished studies may not have been identified through the literature search in OVID databases and publication bias may exist in this meta-analysis. Lastly, studies showed substantial heterogeneity.

In conclusion, this systematic review and meta-analysis showed that FOBT, both gFOBT and iFOBT, have a better diagnostic performance for the relative detection of CRC in the distal colon than that in the proximal colon.

Authorship

Guarantor of the article: Kelvin KF Tsoi.

Author contributions: Study concept, data search, extraction and analysis and drafting of the manuscript were contributed by Hoyee Hirai; Study concept, design and drafting of the manuscript, study supervision were contributed by Kelvin Tsoi; Data search andextraction were contributed by Joyce Chan; Revision of the manuscript and study supervision were contributed by Siew Ng; Study supervision were contributed by Sunny Wong, Jessica Ching, Martin Wong, Justin Wu, Francis Chan and Joseph Sung.

All authors approved the final version of the manuscript.

Acknowledgement

Declaration of personal and funding interests: None.

References

1Siegel R, Ma J, Zou Z, Jemal A. Cancer Statistics, 2014. CA Cancer J Clin 2014; 64: 9– 29.
Wiley Online LibraryCASPubMedWeb of Science®Google Scholar
2Sung JJ, Ng SC, Chan FK, et al. An updated Asia Pacific Consensus Recommendations on colorectal cancer screening. Gut 2015; 64: 121– 32.
CrossrefCASPubMedWeb of Science®Google Scholar
3Mandel JS, Church TR, Bond JH, et al. The effect of fecal occult-blood screening on the incidence of colorectal cancer. N Engl J Med 2000; 343: 1603– 7.
CrossrefCASPubMedWeb of Science®Google Scholar
4Mandel JS, Bond JK, Church TR, et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. N Engl J Med 1993; 328: 1365– 71.
CrossrefCASPubMedWeb of Science®Google Scholar
5Feinberg EJ, Steinberg WM, Banks BL, Henry JP. How long to abstain from eating red meat before fecal occult blood tests. Ann Intern Med 1990; 113: 403– 4.
CrossrefCASPubMedWeb of Science®Google Scholar
6Songster CL, Barrows GH, Jarrett DD. Immunochemical detection of fecal occult blood. The fecal smear punch-disc test: a new non-invasive screening test for colorectal cancer. Cancer 1980; 45: 1099– 102.
Wiley Online LibraryCASPubMedWeb of Science®Google Scholar
7Brenner H, Tao S. Superior diagnostic performance of faecal immunochemical tests for haemoglobin in a head-to-head comparison with guaiac based faecal occult blood test among 2235 participants of screening colonoscopy. Eur J Cancer 2013; 49: 3049– 54.
CrossrefCASPubMedWeb of Science®Google Scholar
8Park DI, Ryu S, Kim YH, et al. Comparison of guaiac-based and quantitative immunochemical fecal occult blood testing in a population at average risk undergoing colorectal cancer screening. Am J Gastroenterol 2010; 105: 2017– 25.
CrossrefPubMedWeb of Science®Google Scholar
9Parra-Blanco A, Gimeno-Garcı′a AZ, Quintero E, et al. Diagnostic accuracy of immunochemical versus guaiac faecal occult blood tests for colorectal cancer screening. J Gastroenterol 2010; 45: 703– 12.
CrossrefCASPubMedWeb of Science®Google Scholar
10Tao S, Brenner H. Well adjusted qualitative immunochemical faecal occult blood tests could be a promising alternative for inexpensive, high-quality colorectal cancer screening. Eur J Cancer Prev 2013; 22: 305– 10.
CrossrefCASPubMedWeb of Science®Google Scholar
11Allison JE, Tekawa IR, Tansom LH, Adrain AL. A comparison of fecal occult-blood tests for colorectal-cancer screening. N Engl J Med 1996; 334: 155– 16.
CrossrefPubMedWeb of Science®Google Scholar
12Castiglione G, Zappa M, Grazzini G, et al. Immunochemical vs guaiac faecal occult blood tests in a population-based screening programme for colorectal cancer. Br J Cancer 1996; 74: 141– 4.
CrossrefPubMedWeb of Science®Google Scholar
13Frommer DJ, Kapparis A, Brown MK. Improved screening for colorectal cancer by immunological detection of occult blood. BMJ 1988; 296: 1092– 4.
CrossrefPubMedWeb of Science®Google Scholar
14Robinson MH, Marks CG, Farrands PA, Thomas WM, Hardcastle JD. Population screening for colorectal cancer: comparison between guaiac and immunological faecal occult blood tests. Br J Surgery 1994; 81: 448– 51.
Wiley Online LibraryPubMedWeb of Science®Google Scholar
15Lee JK, Liles EG, Bent S, Levin TR, Corley DA. Accuracy of fecal immunochemical tests for colorectal cancer: systematic review and meta-analysis. Ann Intern Med 2014; 160: 171– 81.
CrossrefPubMedWeb of Science®Google Scholar
16Launois R, Le Moine JG, Uzzan B, Fiestas Navarrete LI, Benamouzig R. Systematic review and bivariate/HSROC random-effect meta-analysis of immunochemical and guaiac-based fecal occult blood tests for colorectal cancer screening. Eur J Gastroenterol Hepatol 2014; 26: 978– 89.
CrossrefCASPubMedGoogle Scholar
17Rosman AS, Korsten MA. Effect of verification bias on the sensitivity of fecal occult blood testing: a meta-analysis. J Gen Intern Med 2010; 25: 1211– 21.
CrossrefPubMedWeb of Science®Google Scholar
18Oono Y, Iriguchi Y, Doi Y, et al. A retrospective study of immunochemical fecal occult blood testing for colorectal cancer detection. Clin Chim Acta 2010; 411: 802– 5.
CrossrefCASPubMedWeb of Science®Google Scholar
19Young GP, St John DJB, Cole SR, et al. Prescreening evaluation of a brush-based faecal immunochemical test for haemoglobin. J Med Screen 2003; 10: 123– 8.
CrossrefCASPubMedWeb of Science®Google Scholar
20Kutzner N, Hoffmann I, Linke C, et al. Non-invasive detection of colorectal tumours by the combined application of molecular diagnosis and the faecal occult blood test. Cancer Lett 2005; 229: 33– 41.
CrossrefCASPubMedWeb of Science®Google Scholar
21St. John DJB, Young GP, Mchutchison JG, et al. Comparison of the specificity and sensitivity of Hemoccult and HemoQuant in screening for colorectal neoplasia. Ann Intern Med 1992; 117: 376– 82.
CrossrefCASPubMedWeb of Science®Google Scholar
22Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009; 151: 264– 9.
CrossrefPubMedWeb of Science®Google Scholar
23Leeflang MM, Deeks JJ, Gatsonis C, Bossuyt PM; Cochrane Diagnostic Test Accuracy Working Group. Systematic reviews of diagnostic test accuracy. Ann Intern Med 2008; 149: 889– 97.
CrossrefPubMedWeb of Science®Google Scholar
24Macaskill P, Gatsonis C, Deeks JJ, Harbord RM, Takwoingi Y. Chapter 10: analysing and Presenting Results. In: JJ Deeks, PM Bossuyt, C Gatsonis, eds. Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy Version 1.0. The Cochrane Collaboration, 2010. Available at: http://srdta.cochrane.org/.
Google Scholar
25Whiting PF, Rutjes AW, Westwood ME, et al. ; QUADAS-2 Group. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011; 155: 529– 36.
CrossrefPubMedWeb of Science®Google Scholar
26Bossuyt PM, Reitsma JB, Bruns DE, et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med 2003; 138: W1– 12.
CrossrefPubMedWeb of Science®Google Scholar
27Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J ClinEpidemiol 2005; 58: 982– 90.
PubMedWeb of Science®Google Scholar
28Glas AS, Lijmer JG, Prins MH, Bonsel GJ, Bossuyt PM. The diagnostic odds ratio: a single indicator of test performance. J Clin Epidemiol 2003; 56: 1129– 35.
CrossrefPubMedWeb of Science®Google Scholar
29Rutter CM, Gatsonis CA. A hierarchical regression approach to meta-analysis of diagnostic test accuracy evaluations. Stat Med 2001; 20: 2865– 84.
Wiley Online LibraryCASPubMedWeb of Science®Google Scholar
30Swets JA. Measuring the accuracy of diagnostic systems. Science 1988; 240: 1285– 93.
Wiley Online LibraryCASPubMedWeb of Science®Google Scholar
31DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177– 88.
CrossrefCASPubMedGoogle Scholar
32Bjerregaard NC, Tottrup A, Sorensen HT, Laurberg S. Detection of colorectal cancer in symptomatic outpatients without visible rectal bleeding: validity of the fecal occult blood test. Clin Epidemiol 2009; 1: 119– 24.
CrossrefPubMedGoogle Scholar
33Chiu HM, Lee YC, Tu CH, et al. Association between early stage colon neoplasms and false-negative results from the fecal immunochemical test. Clin Gastroenterol Hepatol 2013; 11: 832– 8.
CrossrefPubMedWeb of Science®Google Scholar
34Hope RL, Chu G, Hope AH, Newcombe RG, Gillespie PE, Williams SJ. Comparison of three faecal occult blood tests in the detection of colorectal neoplasia. Gut 1996; 39: 722– 5.
CrossrefCASPubMedWeb of Science®Google Scholar
35Kaul A, Shah A, Magill FH, et al. Immunological faecal occult blood testing: a discriminatory test to identify colorectal cancer in symptomatic patients. Int J Surg 2013; 11: 329– 31.
CrossrefCASPubMedWeb of Science®Google Scholar
36Kim BC, Joo J, Chang HJ, et al. A predicitive model combining fecal calgranulin B and fecal occult blood tests can improve the diagnosis of colorectal cancer. PLoS ONE 2014; 9: e106182.
CrossrefPubMedWeb of Science®Google Scholar
37Koga Y, Yamazaki N, Yamamoto Y, et al. Fecal miR-106a Is a useful marker for colorectal cancer patients with false-negative results in immunochemical fecal occult blood test. Cancer Epidemiol Biomarkers Prev 2013; 22: 1844– 52.
CrossrefCASPubMedWeb of Science®Google Scholar
38Lee Y, Chiu H, Chiang T, et al. Accuracy of faecal occult blood test and Helicobacter pylori stool antigen test for detection of upper gastrointestinal lesions. BMJ Open 2013; 3: e003989.
CrossrefPubMedWeb of Science®Google Scholar
39Shastri YM, Loitsch S, Hoepffner N, et al. Comparison of an established simple office-based immunological FOBT with fecal tumor pyruvate kinase type M2 (M2-PK) for colorectal cancer screening: prospective multicenter study. Am J Gastroenterol 2008; 103: 1496– 504.
CrossrefCASPubMedWeb of Science®Google Scholar
40Sung JJ, Chan FK, Leung WK, et al. Screening for colorectal cancer in Chinese: comparison of fecal occult blood test, flexible sigmoidoscopy, and colonoscopy. Gastroenterology 2003; 124: 608– 14.
CrossrefPubMedWeb of Science®Google Scholar
41Thomas WM, Hardcastle JD, Jackson J, Pye G. Chemical and immunological testing for faecel occult blood: a comparison of two tests in symptomatic patients. Br J Cancer 1992; 65: 618– 20.
CrossrefCASPubMedWeb of Science®Google Scholar
42Chiu HM, Lee YC, Tu CH, et al. Association between early stage colon neoplasms and false-negative results from the decal immunochemical test. Clin Gastroenterol Hepatol 2013; 11(7): 832– 8. e2
CrossrefPubMedWeb of Science®Google Scholar
43van Rossum LGM, van Rijn AF, van Oijen MGH, et al. False negative fecal occult blood tests due to delayed sample return in colorectal cancer screening. Int J Cancer 2009; 125: 746– 50.
Wiley Online LibraryCASPubMedWeb of Science®Google Scholar
44Nelson RL, Dollear T, Freels S, Persky V. The relation of age, race, and gender to the subsite location of colorectal carcinoma. Cancer 1997; 80: 193– 7.
Wiley Online LibraryPubMedWeb of Science®Google Scholar
45Daly JM, Bay CP, Levy BT. Evaluation of fecal immunochemical tests for CRC screening. J Prim Care Com Health 2013; 4: 245– 50.
CrossrefPubMedGoogle Scholar

Citing Literature

Volume43, Issue7

April 2016

Pages 755-764

Systematic review with meta-analysis: faecal occult blood tests show lower colorectal cancer detection rates in the proximal colon in colonoscopy-verified diagnostic studies