ArticlePDF Available

Computed tomographic (CT) colonography for the detection of colorectal cancer - a Technical Brief

Authors:

Marita Broadstock

Content uploaded by Marita Broadstock

Content may be subject to copyright.

NZHTA TECHNICAL BRIEF

June 2007

Volume 6 Number 6

Computed tomographic (CT)

colonography for the detection of

colorectal cancer – a Technical Brief

Marita Broadstock

New Zealand

Health Technology Assessment

Department of Public Health and General Practice

University of Otago, Christchurch

Christchurch, New Zealand

NEW ZEALAND HEALTH TECHNOLOGY ASSESSMENT (NZHTA)

Department of Public Health and General Practice

University of Otago, Christchurch

Christchurch, New Zealand

Computed tomographic (CT)

colonography for the detection of

colorectal cancer – a Technical Brief

Marita Broadstock

NZHTA TECHNICAL BRIEF

June 2007 Volume 6 Number 6

This report should be referenced as follows:

Broadstock, M. Computed tomographic (CT) colonography for the detection of colorectal

cancer – a Technical Brief. NZHTA Technical Brief 2007; 6(6)

Titles in this Series can be found on the NZHTA website:

http://nzhta.chmeds.ac.nz/publications.htm#tech

2007 New Zealand Health Technology Assessment (NZHTA)

ISBN 978-1-877455-07-0 (Print)

ISBN 978-1-877455-08-7 (Web)

ISSN 1175-7884

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

ACKNOWLEDGEMENTS

This Technical Brief was commissioned by the New Zealand Ministry of Health.

Marita Broadstock (Research Fellow) conducted the review, prepared the report and coordinated the

project. The literature search strategy was developed and undertaken by Susan Bidwell (Information

Specialist Manager), who also coordinated retrieval of documents and maintained the bibliographic

database. Catherine Turnbull (Administrator) provided document formatting. Internal peer review was

provided by Dr Robert Weir (NZHTA Director).

DISCLAIMER

NZHTA takes great care to ensure the accuracy of the information supplied within the project

timeframe, but neither NZHTA nor the University of Otago can accept responsibility for any errors or

omissions that may occur. NZHTA and the University of Otago along with their employees accept no

liability for any loss of whatever kind, or damage, arising from the reliance in whole or part, by any

person, corporate or natural, on the contents of this paper. This document is not intended to be used as

personal health advice; people seeking individual medical advice are referred to their physician. The

views expressed in this report are those of NZHTA and do not necessarily represent those of the

University of Otago, or the New Zealand Ministry of Health.

This work is copyright. Apart from any use as permitted under the Copyright Act 1994 no part may be

reproduced by any process without written permission from New Zealand Health Technology

Assessment. Requests and inquiries concerning reproduction and rights should be directed to the

Director, New Zealand Health Technology Assessment, University of Otago, Christchurch, P O Box

4345, Christchurch, New Zealand.

CONTACT DETAILS

New Zealand Health Technology Assessment (NZHTA)

Department of Public Health and General Practice

University of Otago, Christchurch

PO Box 4345

Christchurch

New Zealand

Tel: +64 3 364 3696 Fax: +64 3 364 3697

Email: nzhta@chmeds.ac.nz

Website: http://nzhta.chmeds.ac.nz

LEVEL OF EVIDENCE CONSIDERED IN TECHNICAL BRIEFS

Technical Briefs are rapidly produced assessments of the best available evidence for a topic of highly

limited scope. They are less rigorous than systematic reviews. Best evidence is indicated by research

designs which are least susceptible to bias according to the National Health and Medical Research

Council’s (NHMRC) criteria (2000; 2005). Where methodologically acceptable and applicable,

appraised evidence is limited to systematic reviews, meta-analyses, evidence based clinical practice

guidelines, health technology assessments and randomised controlled trials (RCTs). Where not

available, poorer quality evidence may be considered.

CONFLICT OF INTEREST

None.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

EXECUTIVE SUMMARY

Aim and scope

This Technical Brief aimed to review evidence for the effectiveness and safety of computed

tomographic (CT) colonography for the early identification and management of colorectal cancer.

Eligible studies were those reporting on the use of CT colonography (CTC) for the early detection of

CRC for primary population screening of asymptomatic patients at average risk of CRC; for secondary

screening of asymptomatic patients at high risk of CRC; and/or for diagnostic testing of patients with

symptoms of bowel disease (for surveillance and management). Relevant comparators were

optical/visual endoscopy colonoscopy and double-contrast barium enema. Studies required a valid

reference standard (results from complete optical colonoscopy or surgery, performed independently and

in a blinded fashion, following CTC). Outcomes of interest were health outcomes, test accuracy in

detection of polyps and CRC; and safety and patient experience outcomes including benefits, harms,

preferences and acceptability.

Research papers were excluded if they: were non-systematic reviews; included fewer than 40

participants; reported solely on cost effectiveness analyses, decision analyses, or on data on

extracolonic outcomes; compared CTC technologies or techniques with each other; used CTC for

diagnosis of non CRC conditions; evaluated capsule endoscopy or MRI colonography; and/or used

CTC for pre-operative evaluation or staging, for evaluation of the proximal colon in patients with

occlusive colon cancer, or in patients with an incomplete colonoscopy.

High quality secondary research (systematic reviews and meta analyses) were considered best evidence

on the topic. These were included where selection criteria overlapped with those established for this

review.

Methods

The search strategy considered original articles published from July 2004 in the English language. The

search included major bibliographic and review databases and secondary sources, and mostly published

and indexed literature. Databases included: Medline, Embase, Current Contents, Cochrane Library

Controlled Trials Register, Cochrane Database of Systematic Reviews, Database of Abstracts of

Reviews of Effectiveness (DARE), NHS Economic Evaluation Database, PubMed, Clinical Evidence,

ACP Journal Club, and Health Technology Assessment Database. Search terms and keywords

included: computed tomographic colonography, colonoscopy, virtual colonoscopy, virtual

colonography. References of retrieved publications were also scanned for eligible reviews.

As sufficiently high quality, relevant and recently published systematic reviews were identified; these

alone were included in the Technical Brief. Summaries of appraisal results were presented in Evidence

Tables which detailed study design, study setting, sample, methods, results, reported conclusions, and

reviewer comments based on the limitations and validity of the review. Results were synthesised and

overall conclusions made.

Key results and conclusions

From the search strategy, 775 potentially relevant articles/abstracts were identified, of which 67 were

retrieved. Of these, eight secondary research articles/reports were identified as eligible for appraisal

and were included in the review. Three were published in 2005, three in 2006, and two in 2007.

Appraised papers included three relatively poor quality, largely narrative reviews, and one limited

quality and four high quality meta-analyses.

The following findings and conclusions were made:

1. CT colonography is a relatively safe procedure compared to DCBE, and at least as safe as, or safer

than, diagnostic colonoscopy. Ionizing radiation exposure is relatively low but a cumulative risk

for regular screening. There is a very small risk of colonic perforation.

2. Generally there have been inconsistent findings regarding preferences for and experiences from

CTC versus colonography. However one meta-analysis of 11 studies of increased risk or

symptomatic patients concluded that CTC may be preferred over colonoscopy, and that the

majority of studies have reported results favouring CTC over colonoscopy with respect to pain and

discomfort.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

iii

3. CT colonography has reasonable test sensitivity and specificity in the detection of large and

medium polyps, but is poorly accurate for small lesions. Whilst specificity has been consistently

high, test sensitivities have varied and pooled statistics need to be considered with caution. There

is some evidence that CTC is highly accurate in the detection of symptomatic cancer.

4. There is some evidence that CTC is more accurate than air-contrast barium enema for detecting

polyps and cancers in increased risk or symptomatic populations. However, conventional

colonoscopy appears to be more accurate than CTC for large polyps, and particularly for smaller

polyps.

5. Limitations of the current evidence base include that there is a lack of evidence about the accuracy

of CTC for primary screening in average risk populations. There is also a need for greater

investigation of the reasons for such wide variations in test accuracy achieved in different trials

with respect to patient and scanner characteristics. Likely sources of variation relate to prevalence

of disease, CTC techniques (such as width of collimation, type of detector, and mode of imaging),

and radiologist experience. The definition of what constitutes a clinically important polyp in size

and morphology also requires evidence-based elucidation. Whilst the methodological quality of

studies is improving, comparisons between studies would be facilitated by more consistency in

reporting and more appropriate statistical analysis and data synthesis techniques.

6. There have been no studies reporting on overall health outcomes of CTC including efficacy in

reducing CRC incidence or mortality.

7. Based on the evidence and conclusions considered in this review CTC is not currently

recommended for generalised screening.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

TABLE OF CONTENTS

ACKNOWLEDGEMENTS.................................................................................................i

DISCLAIMER.................................................................................................................i

COPYRIGHT..................................................................................................................i

CONTACT DETAILS .......................................................................................................i

LEVEL OF EVIDENCE CONSIDERED IN TECHNICAL BRIEFS .............................................i

CONFLICT OF INTEREST................................................................................................i

EXECUTIVE SUMMARY ................................................................................................ii

Aim and scope............................................................................................................ii

Methods......................................................................................................................ii

Key results and conclusions.......................................................................................ii

TABLE OF CONTENTS..................................................................................................iv

RESEARCH QUESTION..................................................................................................1

BACKGROUND.............................................................................................................1

Overview of CT colonography...................................................................................1

METHODOLOGY ..........................................................................................................2

Study inclusion criteria..............................................................................................2

Study exclusion criteria..............................................................................................3

Main search terms......................................................................................................3

Publication date and level of evidence ......................................................................4

Selection and appraisal..............................................................................................4

Data extraction into evidence tables..........................................................................4

RESULTS .....................................................................................................................5

Summary of review findings.....................................................................................22

Conclusions..............................................................................................................25

REFERENCES .............................................................................................................27

APPENDIX 1: SEARCH STRATEGY........................................................................28

Medline/Medline Pending/Cochrane Central Register of Controlled Trials ..........28

Embase ......................................................................................................28

Current Contents ......................................................................................................29

PubMed (last 60 days) .............................................................................................29

APPENDIX 2: EXCLUDED RETRIEVED PAPERS.............................................................30

PPENDIX 3: APPRAISED RETRIEVED PAPERS ............................................................35

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

RESEARCH QUESTION

Dr John Childs (Principal Advisor, Cancer Control) and Sarah Greensmith (Senior Policy Analyst,

Strategic Screening Team, National Screening Unit) from the Ministry of Health, New Zealand

Government, requested this Technical Brief. The focus is on the use of computed tomographic (CT)

colonography for the identification and management of colorectal cancer in three domains:

 population screening,

 screening of increased risk cohorts, and

 management of symptomatic people.

The National Screening Unit is interested in what are the benefits and risks of this procedure, and

whether it has potential to improve morbidity and mortality from CRC.

The full research question is: What is the effectiveness and safety of computed tomographic (CT)

colonography for the early identification and management of colorectal cancer (CRC)?

BACKGROUND

Overview of CT colonography

Computed tomographic colonography or CTC, also known as virtual colonoscopy, simulates

conventional endoscopy colonoscopy (EC). Virtual colonoscopy using computed tomography (CT)

scanners is a modality that has been considered as potentially suitable for colorectal cancer screening

of average risk and above average risk patients, as well as for diagnosis and management of

symptomatic patients. The technique generally involves a similar bowel preparation to conventional

endoscopy colonoscopy. Air or carbon dioxide is insufflated into the colon through a rectal tube and

then data are acquired by the scanner which generates images of the colon. It provides a complete

examination, allowing imaging of the outside and inside of the bowel and neighbouring organs.

Depending on the equipment used, data obtained from rapid helical CT scanning of the abdomen are

presented as two-dimensional images, with three-dimensional images generated of areas identified as

suspicious (Levin et al. 2003). Data can be double read which increases sensitivity without increased

risks for the patient. Sedation is not required, although some mild patient discomfort may be

experienced from the air insufflation (Walsh and Terdiman 2003). According to the NZ Branch of the

Royal Australian and New Zealand College of Radiologists, the procedure has no significant risk of

perforation, and does not require a post-procedure recovery period (Colorectal Cancer Screening

Advisory Group 2006). After radiologist review of results, patients noted to have suspicious-looking

polyps or a colonic mass need to proceed to conventional colonoscopy, to enable biopsy or resection of

the lesion. CTC identifies incidental (extracolonic) abnormalities outside the bowel in a significant

proportion of cases (e.g., liver and kidney cysts), most of which are benign, and these may require

further (sometimes invasive) investigative procedures (Colorectal Cancer Screening Advisory Group

2006).

CTC has several potential advantages over other tests considered for early detection of CRC. It

potentially can identify large-bowel malignancies that are often poorly assessed by conventional

colonoscopy, such as those located within haustral folds, as well as being able to view the entire colon.

For this reason it has also been suggested as the examination of choice for failed or incomplete

colonoscopies in many settings (Levin et al. 2003; Medical Services Advisory Committee (MSAC)

2006). CTC may also allow for small (and probably hyperplastic, low-risk) colonic polyps to be left

in-situ when detected, and for regular reassessment to be carried out to monitor them (Levin et al.

2003), although patients may prefer it to be removed even if considered to be low risk (Banerjee and

Van Dam 2006). Future potential developments for this modality include the ‘patient-friendly’

possibility of avoiding pre-imaging bowel preparation (Iannaccone et al. 2004), through faecal tagging

with oral contrast, electronic cleansing (where the stool density relative to other tissues is calculated

and subtracted during image generation), or using noncathartic protocols. Computer-aided detection

(CAD) software is also being investigated to improve diagnostic accuracy (Perumpillichira et al. 2005).

High resource costs for equipment and radiologist training are currently well-recognised barriers to

widespread use of virtual colonoscopy. By contrast, CTC is less time consuming for the patient and

the relevant medical specialists than conventional colonoscopy and is less expensive. Multislice CT

scanners are already widely available in New Zealand.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

As with any test, virtual colonoscopy also has disadvantages. Patients are exposed to ionizing

radiation, although low-radiation dose protocols are under investigation. False positives can occur as a

result of retained stool in the bowel, diverticular disease (which can produce poorly distensible areas of

the colon), or thickened bowel folds. Virtual colonoscopy may be less sensitive to detecting relatively

rare flat adenomas (Levin et al. 2003), which are defined as having a height that is no more than one-

half of their width (Nicholson et al. 2005a). As polyps cannot be removed during the procedure,

patients with polyps detected at CTC require an additional endoscopic procedure for their removal.

Variable factors such as radiologist experience and training may also influence accuracy (Pignone et al.

2002; Walsh and Terdiman 2003), and training specific to CTC interpretation is being advocated.

CTC is already well established as a reliable diagnostic tool in symptomatic patients who are unable to

undergo complete colonoscopy. Whilst not yet endorsed as a screening test in the general risk

population, there is general consensus that the technology holds promise for the future, and there is

burgeoning research into CTC for colorectal cancer screening. Since 2003, there has been an

exponential growth in studies investigating test performance of CTC, usually using visual colonoscopy

as a reference standard. There is also a growing literature investigating psychological outcomes

including patient discomfort and acceptability.

METHODOLOGY

Study inclusion criteria

Secondary studies were eligible for inclusion where they:

 included a methods section describing how the relevant studies were identified (including

bibliographic database/s)

 considered primary studies with selection criteria overlapping with those presented below

 were published from July 2004.

Population

Studies considering the use of CT colonography for the early detection of CRC in the following

modalities:

 For primary population screening of asymptomatic patients at average risk of CRC

 For secondary screening of asymptomatic patients at high risk of CRC. These include those

with a personal history of CRC, colorectal adenoma, or inflammatory bowel disease; positive

faecal occult blood; family history of CRC; gene carriers of/at risk for hereditary

nonpolyposis colorectal cancer (HNPCC); and/or gene carriers of/at risk for familial

adenomatous polyposis (FAP)

 For diagnostic testing of patients with symptoms of bowel disease (for surveillance and

management).

Intervention

Studies evaluating computed tomographic colonography (CTC) for the detection of precancerous

(adenomatous and benign) polyps and CRC.

Comparators

Other commonly used examinations including:

 optical/visual endoscopy colonoscopy (EC) (rigid or flexible scope)

 double-contrast barium enema.

Reference standard

The reference standard used for verification of test accuracy included results from complete optical

colonoscopy or surgery, performed independently and in a blinded fashion, following CTC.

Note that whilst colonoscopy is the current gold standard, it has a miss rate of 6% for polyps 1cm or

more in diameter, and may be incomplete in 5-15% of examinations (Banerjee and Van Dam 2006;

Nicholson et al. 2005b). To adjust for this, some test accuracy studies have employed “segmented

unblinding” during colonoscopy as the reference standard. After evaluation by colonoscopy is

completed for a segment, the endoscopist is given the CTC results for that segment and the area is re-

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

examined if a polyp is noted on the CTC. This identifies a false negative rate for colonoscopy and a

false positive rate for CTC more accurately (Nicholson et al. 2005b).

Outcomes

Outcome measures including any of the following:

 health outcomes including CRC morbidity or mortality, incidence of CRC, and overall

mortality

 test accuracy in detection of polyps (precancerous adenomatous and benign) and CRC,

presented per-polyp and per-patient, including sensitivity (Se) and specificity (Sp); true

positive, false positive, and false negative results; and positive predictive value, negative

predictive value, and likelihood ratios; ideally presented for different polyp size ranges

 other outcomes relating to the procedure, including benefits and harms (including physical or

psychological sequelae) and acceptability for the intervention compared with the comparator.

Study design

A range of study designs were relevant, dependent on the outcomes considered:

 for health outcomes (mortality and morbidity), randomised and pseudo-randomised controlled

trials

 for test characteristics, prospective studies of test accuracy involving within-subjects,

independent, blinded comparison of CT colonography with a valid reference standard

(performed after the CTC) in people with a defined clinical presentation

 for benefits and harms relating to patient experience (including acceptability, pain, discomfort

and preference), prospective controlled trials with within-subjects comparisons of outcomes

following CTC to those following conventional examination comparators in a clinical setting

 for adverse events, data from national database/s of adverse events.

Sample size

Studies with samples of at least 40 participants.

Study exclusion criteria

Research papers were excluded if they:

 were not published in English,

 were “correspondence”, editorials, comments, book chapters, conference proceedings,

abstracts,

 reported animal studies,

 did not clearly describe their methods and results or had significant discrepancies,

 were narrative (non systematic) reviews,

 were case presentations, and/or reported on:

 participants aged under 18 years of age,

 cost effectiveness analyses or decision analyses,

 data on extracolonic outcomes,

 studies comparing CT colonography technologies or techniques with each other,

including use of bowel preparation methods, oral (e.g., faecal tagging) contrast

techniques, intravenous contrast techniques, colonic insufflation, smooth-muscle

relaxants, image acquisition, bipositional scanning, image processing, or reader

training or experience,

 studies using CT colonography for diagnosis of non CRC conditions, such as colonic

diverticulitis or acute appendicitis,

 studies evaluating capsule endoscopy, or MRI colonography (alone), and/or

 studies using CT colonography for pre-operative evaluation or staging of CRC

tumours in patients with known CRC, or for evaluation of the proximal colon in

patients with occlusive colon cancer, or in patients with an incomplete colonoscopy.

Main search terms

Details of the search strategy are presented in Appendix 1.

MESH headings (Medline subject headings): colonography-computed-tomographic, exp

colonoscopy

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Embase subject headings (where different from Medline): computed tomographic colonography

Additional free text (used in all databases): (virtual adj colonoscopy), (virtual adj colonography), (ct

adj colonograph$), (ct adj colonograph$)

Search sources

The NZHTA CORE Search was employed. Characteristics of the core search include: essential

sources only, major databases and secondary sources, and mostly published and indexed literature.

Principal sources of information

Bibliographic databases

 Medline

 PubMed (last 90 days)

 Embase

 Current Contents

 Cochrane Central Register of Controlled Trials

Review databases

 Cochrane Database of Systematic Reviews

 Clinical Evidence

 DARE database

 NHS Economic Evaluation Database

 Health Technology Assessment Database

 ACP Journal Club

Other sources of information

 National Institute for Clinical Evidence (NICE)

 Institute for Clinical Services Improvement

 Blue Cross Blue Shield Technical Assessment Program

 New Zealand Ministry of Health

 Australian National Health and Medical Research Council

 National Cancer Control Initiative (Australia)

Cited references of retrieved articles were scanned for additional potentially eligible papers.

Extended searching of internet websites, meeting abstracts, hand searching of journals, and contacting

of authors for unpublished data was not undertaken.

Publication date and level of evidence

Following an initial scoping search, several recently published, well-conducted systematic reviews and

meta-analyses were identified on the topic. In recognition that well conducted systematic reviews are

considered “best evidence”, literature considered eligible for critical appraisal were restricted to

secondary research published from July 2004 onwards.

Selection and appraisal

The search strategy identified abstracts and titles for published articles. The reviewer (MB) applied the

inclusion and exclusion criteria to identify those potentially eligible for selection and appraisal and

these were retrieved as full text. The selection criteria were fully applied to the retrieved articles to

identify the final set of papers to be appraised and included in the review.

Data extraction into evidence tables

Systematic reviews and/or meta-analyses were described and critiqued. Summaries of appraisal results

were shown in evidence tables, and included:

 reference (authors, publication date)

 review question or aim

 search method employed by the review (including databases searched, search dates, search

terms, other reference identification processes)

 review inclusion and exclusion criteria, data extraction and appraisal methods

 results of analyses comparing intervention and comparator groups on eligible outcomes

including statistically-tested comparisons and reporting relevant statistical data

 comments on the review’s limitations

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

 authors’ conclusions

 critical comments (strengths and weaknesses of the review).

For appraisal of the methodology of the included secondary studies, specific criteria assessed whether

the review asked a focused question, if the eligibility criteria for included studies were explicit, what

search strategy was used, how the validity of included trials was assessed, and whether results of

included studies were similar. A summary of these criteria is presented in Table 1 below.

Table 1. Validity criteria for appraisal of secondary studies

Is there a focused research question?

i.e. PICO elements: patient, intervention/diagnostic or screening test of interest, comparator,

outcomes

Are inclusion and exclusion criteria for selected studies stated?

Is there an explicit and comprehensive search strategy?

Did review incorporate a search strategy comprehensive enough that it was unlikely to have

missed studies?

Are the included trials appraised for validity?

Are validity criteria stated?

Are results consistent from study to study?

Is homogeneity assessed?

Summary of main results

Strengths and limitations

Adapted from Evidence Based Medicine Toolkit, University of Alberta (http://www.med.ualberta.ca/ebm/ebm.htm)

In addition to evidence tables, the appraised secondary studies were summarised in the report’s text.

Inter-review consistency was considered, results synthesised and overall conclusions drawn.

RESULTS

From the search strategy 775 potentially relevant articles/abstracts were identified of which 67 were

retrieved (including one paper that was found opportunistically when scanning the contents of a

recently published Journal). Retrieved papers included primary research, however once it was

determined that there was sufficient high order (level I) secondary evidence according to NHMRC’s

(2000) hierarchy of evidence, primary studies were excluded. In total, 59 retrieved articles were

excluded and these are listed in Appendix 2.

Eight publications reporting on systematic reviews/meta analyses were eligible for appraisal and

inclusion in the Technical Brief (listed in Appendix 3). Evidence tables of included papers are

presented in Table 2 in chronological order of year of publication (and within each year, in

alphabetical order of first author).

Halligan et al (2005)

The systematic review and meta-analysis of Halligan et al (2005) employed a rigorous search strategy

to identify studies published between 1994 and December 2003. A preliminary search to May 2003

included Medline, Cochrane Controlled Trials Register, EMBASE, Science Citation Index, and hand-

searching of Journals. The search was extended to December 2003 using Medline only as the broader

initial search had not found any articles that were not in Medline. The authors reported that as this

search strategy did not identify any eligible articles not already identified by Medline, the May to

December 2003 search only employed Medline. A broad range of search terms with no language

restrictions were used, with additional references obtained from the bibliographies of the retrieved

articles. Authors were contacted for clarification only.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Inclusion criteria required that the study focused on the detection of colorectal polyps verified with

blinded within-subject reference colonoscopy or surgical findings. Required procedures were the

inclusion of full bowel preparation, acquisition of prone and supine images, and helical scanners using

commercially available software. Studies that used computer-aided detection (CAD) systems were

excluded (reportedly to avoid bias as two authors worked for a company that develops such systems),

as were studies with intravenous contrast material administered to patients. Included studies were to

involve at least 30 patients. Studies where the prevalence of abnormality was broadly known to CT

observers were also excluded. Two researchers conducted searches, applied selection criteria and

abstracted data independently, resolving discrepancies by consensus or in discussion with other authors

where necessary. Explicit quality assessment guidelines were employed.

The reviewers identified 24 eligible studies, including one study which included average risk patients

only, and one multi-centre study. For detection of large polyps (1cm or larger) most studies had good

per-patient sensitivity (93%, 95% CI 73%-98%) and all had excellent per-patient specificity (97%,

95% CI 95%-99%). For detection of medium and larger polyps (6mm and above), again studies had

good per-patient sensitivity (86%, 95% CI 75%-93%) but specificity was variable (86%, 95% CI 76%-

93%).

For detection of polyps of all sizes, studies were too heterogenous in sensitivity (range 45%-97%) and

specificity (26%-97%) to pool results. The number of cancers per study was too small to allow meta-

analysis. However, when treating data as if it were from a single study, sensitivity was 96% (95% CI

91%-99%). There were too few studies to permit sensitivity analyses investigating the effects of either

using a modified reference standard (segmental unblinding of colonoscopy), or involving individual

observer assessment versus consensus assessment.

Halligan et al (2005) concluded that CTC seems sufficiently sensitive and specific in the detection of

large and medium polyps; it is especially sensitive in the detection of symptomatic cancer. They further

argued that CTC should be further investigated as a diagnostic tool for cancer. As only one study

considered patients at average population risk alone, the authors argue that data may not be applicable

to a screening situation. This review provided a detailed discussion of methodological issues in the

field and suggested improvements for future research. In addition to a comprehensive assessment of

study limitations, a minimum data set was proposed for future researchers to observe in reporting

studies of diagnostic accuracy.

Mulhall et al (2005)

Mulhall et al’s (2005) systematic review and meta-analysis considered the test performance of CTC

compared to endoscopy colonoscopy or surgery. The search strategy considered English language

publications between 1975 and February 2005 identified from a range of databases (though without

journal hand-searching, citation searching or contact with authors). Explicit selection criteria specified

that studies included prospective trials of adults undergoing CTC after full bowel preparation, with

colograms interpreted blind to the result of the EC or surgery (as the gold reference standard). Studies

needed to have used state-of-the-art technology, including at least a single-detector CT scanner with

supine and prone positioning, insufflation of the colon with air or carbon dioxide, collimation smaller

than 5mm, and both 2-D and 3-D views used during scan interpretation. Two researchers

independently searched the literature, applied selection criteria, abstracted data and coded aspects of

methodological quality with disagreements resolved by consensus.

Thirty-three studies were identified for inclusion. The pooled per-patient sensitivity of CTC was

heterogenous and improved as polyp size increased, being 0.48 for polyps <6mm (95% CI 0.25-0.70),

0.70 for polyps 6-9mm (95% CI 0.55-0.84), and 0.85 for polyps >9mm (95% CI 0.79-0.91).

Characteristics of the CTC scanner explained some of the heterogeneity, including width of collimation

(thinner led to higher sensitivity), type of detector (multiple scanners more sensitive than single

detectors), and mode of imaging (“fly through” being more sensitive, though based on only two studies

using this method). Other possible sources of false-negative results were discussed but limited

reporting and number of studies limited ability to investigate them systematically. By contrast,

specificity was homogenous, being 0.92 for polyps <6mm (95% CI 0.89-0.96), 0.93 for polyps 6-9mm

(95% CI 0.91-0.95), and 0.97 for polyps >9mm (95% CI 0.96-0.97).

Detailed tables highlighted common sources of bias. These included: differences in disease severity or

prevalence among studies, investigators being aware of baseline risk (clinical review bias), differential

verification of findings, and variability in observer experience. The review considered limitations of

the evidence base, including that the studies differed widely (with the extractable variables explaining

only a small amount of variance), that only a few studies considered the newest CTC technology, that

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

only three studies were designed to evaluate a true screening population of average risk people, and

that there are several limitations of colonoscopy as a gold standard.

Mulhall et al (2005) concluded that whilst CTC is highly specific, the range of reported sensitivities is

wide. Patient or scanner characteristics do not fully account for this variability, but collimation, type of

scanner, and mode of imaging explain some. This heterogeneity raises concerns about consistency of

performance and about technical variability. The authors advised that these issues must be resolved

before CTC can be advocated for generalised screening for colorectal cancer and that in the meantime

CTC should only be used in research protocols or when other accepted screening methods are not

appropriate.

Nicholson et al (2005b)

A systematic review by Nicholson et al (2005b) considered the role of CTC in CRC screening. The

review employed a single database search (Medline), although cross-checking of paper references was

also performed to identify additional papers. The review presented no details on search dates, search

terms, selection criteria, or data extraction or appraisal methods. Appraisal and synthesis was

narrative, with no tabular presentation of findings. Despite these limitations, the paper provides a

useful critical account of research relating to CTC across the following areas: performance

characteristics, flat lesions, extra-colonic findings, indication for CTC, patient preference, ways of

improving the technique, and cost effectiveness.

With respect to test accuracy, the authors reported that some studies reported promising sensitivity (of

more than 90% for detection of polyps at least 10mm in diameter), whilst other studies reported

disappointing results with sensitivity ranging between 55% and 64%. Nicholson et al (2005b) discuss

reasons for variation including the following: inclusion of polyp-enriched populations, inclusion of

faecal tagging with electronic cleansing, varying radiologist experience (and use of single institution

expertise), use of radiologists’ consensus interpretation following double-reading, and primary reliance

on 3D image review. As expected, the authors found that CTC performs better in detecting polyps

greater than 10mm than those around 5mm in diameter. They cited the meta-analysis of Sosna et al’s

(2003) involving 14 prospective studies published to July 2002. This found a pooled per patient

sensitivity of 0.84 for polyps 6-9mm, and of 0.88 (and specificity of 0.95) for polyps 10mm or larger.

Nicholson et al (2005b) argued that the detection of flat polyps has been under-investigated and

therefore it is difficult to draw firm conclusions about this issue. They suggest that what 2D and 3D

imaging combination formats are optimal is also not established.

With respect to research reporting on patient experience outcomes, Nicholson et al (2005b) noted that

there has been conflicting findings, with some studies finding greater preference for CTC, others

favouring visual colonoscopy, and some studies finding similar acceptance for both methods. The

authors argue that the literature on patient experience is limited by a reliance on unvalidated and

subjective questionnaires. Several studies were described suggesting that patients generally find CTC

less painful and embarrassing than colonoscopy. The reviewers concluded that CTC may be favored

by patients compared with other screening tests due to the ease of performance and comfort, and may

become more acceptable once noncathartic preparation methods have been perfected. It is also noted

that radiation exposure for CTC is in the realm of that received with barium enema, an accepted

screening technique.

Nicholson et al’s (2005b) review was supported by industry funding. The reviewers provided a

summary of various medical society recommendations and concluded that although not yet endorsed

for widespread use by major gastroenterological societies, “CTC shows promise as a screening tool”.

Banerjee and Van Dam (2006)

This review considered CTC’s efficacy for colorectal screening in a largely narrative review that

included a systematic process for ranking quality of a subset of studies on test accuracy. The reviewers

considered research published in English between 1960 and March 2005 identified using a Medline

search and a comprehensive range of search terms for CTC. There was no hand-searching of Journals,

or checking of reference lists performed. Papers which compared CTC with colonoscopy in

prospective, blinded trials were examined independently by two investigators and assigned a

categorical rating of quality, A. B or C, in decreasing order of study quality. Evidence A studies

considered over 500 patients, were multicentre, used multidetector CT scanners, and employed

segmental unblinding. Evidence B studies included over 100 patients, were single or multicentre

studies, used single row or multidetector CT scanners, and did not employ segmental unblinding.

Evidence C studies were those with 100 or fewer subjects, were single centre studies, used single row

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

CT scanners, and again did not employ segmental unblinding. Results were presented in tables as well

as described in the text, although there were some minor discrepancies and the Table results are

referred to here.

Six studies were graded as being of the poorest quality (Evidence C). CTC was unequivocally poor at

detecting small lesions in these studies, with per-polyp sensitivity for detecting polyps of at least 5mm

ranging between 11% and 55%. For medium sized polyps (6-9mm), per-polyp sensitivity ranged from

36% to 82%, and increased to between 50% and 91% for polyps at least 10mm in size. Per-patient

sensitivity for 6-9mm sized polyps was 43% to 94% and specificity was between 58% and 92%. The

per-patient sensitivity for polyps of at least 10mm in diameter ranged from 37% to 96%, with

specificity of between 74% and 96%.

For somewhat higher quality designs (Evidence B), seven studies were identified. Results were still

widely variable, with per-polyp sensitivity for detecting 6-9mm sized polyps ranging between 29% and

82%, and between 32% and 93% for polyps of at least 10mm in diameter. Per-patient sensitivity for 6-

9mm polyps were 41% to 93%, with specificity of between 71% and 95%. For the larger polyps of at

least 10mm, per-patient sensitivity ranged from 35% to 100%, and specificity was between 92% and

98%.

Finally, the highest quality group (Evidence A) included three recently published studies. Pickhardt et

al’s (2003) study considered asymptomatic patients at average or increased through family history risk

for CRC. Unlike the other two evidence A studies, this trial involved careful bowel preparation with

solid-stool tagging and electronic cleansing, primary readings using 3D images and 2D images for

problem solving, and experienced readers. The study achieved similar per-polyp sensitivity for CTC

and EC in detection of polyps of 6mm or greater (86% and 90% for CTC and EC respectively) and

those sized 10mm or greater (92% and 88% for CTC and EC respectively). Per-patient sensitivity for

polyps at least 6mm for CTC and EC was 88% and 80% respectively, and for polyps at least 10mm

was 94% and 96% respectively. Test accuracy was poorer for CTC in a more recent study by Cotton et

al (2004) involving symptomatic patients or patients with family history of polyps at nine academic

centres. Per-polyp sensitivity for 6-9mm polyps was 23% for CTC compared with 96% for EC, and for

polyps of at least 10 mm in size, per-polyp sensitivity was 52% for CTC compared with 96% for EC.

The reviewers argued that poor results may relate to inexperienced readers, as the largest centre had

sensitivity for polyps greater than 6 mm of 82%. Most recently, in a sample of increased risk subjects

Rockey et al (2005) found similarly poor results, despite using superior scanners and more experienced

readers than Cotton et al (2004). Sensitivity for polyps 6-9mm was 47% for CTC compared with 99%

for EC, and for larger polyps of at least 10mm in diameter, sensitivity was 53% for CTC compared

with 99% for EC.

The reviewers also considered evidence for other outcomes, though it was unclear whether these papers

were systematically identified and appraised. With respect to patient experience literature, Banerjee

and Van Dam (2006) concluded that it is unclear whether CTC is preferred to EC, and mixed results

were described. Two larger trials found similar levels of discomfort expressed for each procedure, and

no preference for either procedure in a third study. The reviewers described the study by Iannaccone et

al (2004) where CTC with faecal tagging and no cathartic preparation was offered, and preferred by

61% of subjects compared with EC following catharsis. Nevertheless, it is emphasised that a

significant sub-group of 35% still preferred EC, possibly because it allowed for immediate polyp

removal where necessary. With respect to safety, the authors noted that the radiation dose of 0.44 rem

from CTC is similar to that received when undergoing two abdominal radiographs. However they

argued that this level could still be of concern if CTC was used as a regular screening tool. The

reviewers identified only two cases of perforation from CTC arising in patients with diseased colons

due to over-inflation with air, and noted that no cases had occured in average-risk populations to date.

Whilst there were no overall conclusions, the review provided a useful description of results of studies

organised in relation to technique, false negative and false positive results on CTC, effects of training,

difficulties of detecting flat and small polyps, CTC in special situations, extracolonic findings, cost-

effectiveness, and upcoming advances in CTC

Davila, Rajan and Baron (2006)

This review was conducted to update a guideline of the American Society for Gastrointestinal

Endoscopy relating to the use of gastrointestinal endoscopy for colorectal cancer screening and

surveillance. Details on the limited search were scant, however a Medline search was undertaken with

additional references obtained from the bibliographies of the identified articles and from

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

recommendations of expert consultants. Search dates, search terms, selection criteria and appraisal

methods were not described and results were discussed narratively.

From a number of recent studies, the following test accuracy results were described for comparisons

between CTC and colonoscopy. Per-polyp sensitivity for polyps at least 6mm ranged between 39%

and 94%, and specificity between 79% and 92%. By comparison, per-polyp sensitivity for polyps of at

least 10mm ranged between 55% and 100%, and specificity between 94% and 98%. The prospective

trial of Rockey et al (2005) for people at high risk for CRC was reported with sensitivity for at least

10mm diameter lesions being 59% for CTC, compared with 48% for DCBE and 98% for colonoscopy.

The authors noted that there were no studies reporting on efficacy of CTC in reducing CRC incidence

or mortality. They also commented that EC may detect clinically important extracolonic findings, and

that cost effectiveness studies indicate that under most assumptions endoscopy colonoscopy is more

cost-effective than CTC. Comparative studies on patient acceptance data were cited as suggesting that

there was no consistent preference.

Davila et al (2006) concluded that as studies evaluating virtual colonoscopy (and faecal DNA testing)

for CRC screening have yielded conflicting results it therefore cannot be recommended.

MSAC (2006)

This systematic review and meta analysis by the Australian Federal government’s Medical Services

Advisory Committee (2006) was conducted by a team from the NHMRC Clinical Trials Centre.

Clinical and consumer expertise were also involved through the establishment of an advisory panel.

The reviewers aimed to assess the safety and effectiveness of CTC for the diagnosis or exclusion of

colorectal neoplasia in symptomatic patients or in patients that are asymptomatic but at high risk of

colorectal neoplasia due to a personal or family history of colorectal polyps or cancer, versus DCBE

and versus endoscopy colonoscopy. Note that studies of average risk asymptomatic patients were not

considered. Cost effectiveness evidence and evidence relevant to patients who are ineligible for

colonoscopy was also considered but are not reported here as beyond the scope of the current Technical

Brief.

The review considered research published between 1994 and June 2005 through a comprehensive

search of several databases and websites using a detailed list of search terms. Additional references

were obtained from the bibliographies of the retrieved articles. Inclusion criteria were that: studies

were to perform multislice CTC (at least 4-slice CT scanning); use colonoscopy or surgical findings as

the reference standard; and use double contrast barium enema and/or colonoscopy as a comparator.

Studies had to report on at least one of the following: diagnostic accuracy with sufficient data to

calculate sensitivity and specificity; changes in clinical management; and patient outcomes (morbidity,

mortality, adverse events, quality of life, patient preferences). Exclusion criteria were: not being an

appropriate clinical study, studies with average risk asymptomatic patients, studies with fewer than 10

patients undergoing CTC, studies which compared two or more different techniques of CTC without a

reference standard, and studies not published in the English language.

A second researcher scanned abstracts for potentially eligible articles, with discrepancies resolved by

discussion. Data were extracted using a standardised instrument by one reviewer, and checked by a

second, with discrepancies resolved by discussion with a third reviewer. Data on quality was rated

against explicit criteria relevant to patient outcomes using the QUADAS tool.

The review found that no studies compared overall health outcomes following the use of CTC, DCBE

or endoscopy colonoscopy. The review included four systematic reviews, including two appraised in

this Technical Brief (Halligan et al. 2005; Mulhall et al. 2005), 24 clinical studies reported on safety

and accuracy, and 11 studies reported on patient preferences or quality of life outcomes.

With respect to CTC accuracy, CTC was found to be generally highly sensitive and specific for the

diagnosis or exclusion of cancers and polyps greater or equal to 10mm in size in symptomatic patients

and asymptomatic patients at high risk of colorectal neoplasia. This was based on 11 studies of

variable quality, with median CTC sensitivity of 84% (range 55-100%), and median CTC specificity of

97% (range 74-100%). Estimates of CTC accuracy were higher for the detection of cancer alone.

From a meta-analysis of four studies: CTC sensitivity was 97% (95% CI 89-100%) and CTC

specificity was 98% (95% CI 95-99%). These findings were interpreted as being consistent with

results from three published systematic reviews. The authors concluded that the results also indicated

that CTC is only moderately sensitive for the detection of lesions 6-9 mm in diameter (from six studies,

CTC sensitivity range 30-80%, CTC specificity range 93-99%), and is poorly sensitive for lesions less

than 6mm in diameter (from four studies, CTC sensitivity range 14-57%, CTC specificity range 83-

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

97%). Reviewers commented that variation observed between studies demonstrates that CTC is less

accurate in some population subgroups or settings. They further argued that the extent to which patient

characteristics, prevalence of disease, CTC techniques, the experience of those performing and

interpreting the tests or other factors may influence CTC performance has not yet been clearly defined.

Relative accuracy of CTC, DCBE and colonoscopy was investigated in only one eligible study, which

was appraised as being of fair quality (Rockey et al. 2005). The study found that CTC and DCBE

accuracy was lower than found in noncomparative studies and systematic reviews of CTC accuracy.

The study’s results suggested that CTC is a more specific test than DCBE, but less sensitive and

specific than colonoscopy for the detection of cancers and polyps greater than or equal to 10 mm. The

study suggested that CTC may be a more sensitive test than DCBE; this only reached statistically

significance for lesions 6-9 mm in size.

With respect to patient preferences, the evidence reviewed suggested that CTC may be preferred over

colonoscopy. However, comparison of pain and discomfort experienced by patients undergoing both

tests showed mixed results with five of eight studies reporting results in favour of CTC, and three in

favour of colonoscopy.

CTC was found to be a relatively safe procedure compared to DCBE and at least as safe as, or safer

than, diagnostic colonoscopy. Both CTC and DCBE were said to expose patients to ionizing radiation

and to be associated with a very small risk of colonic perforation.

The authors concluded that CTC is a relatively safe test compared to DCBE and colonoscopy.

Evidence about CTC accuracy for the detection of cancers and polyps greater than or equal to 10 mm

was said to compare favourably with DCBE. The reviewers also argued that there is some evidence to

suggest that patients prefer CTC over DCBE, and over colonoscopy. It was concluded that CTC is less

accurate than colonoscopy for the detection of cancers and polyps greater than or equal to 10 mm.

MSAC (2006) recommended that public funding for CTC as a substitute investigation for colonoscopy

should not be supported. Whilst beyond the scope of the current Technical Brief, based on evidence

appraised MSAC also recommended that public funding for CTC should be supported for exclusion of

colorectal neoplasia in symptomatic or high risk patients who are either ineligible for colonoscopy due

to patient contraindications, or where there is an inability to perform or complete a colonoscopy. The

review also reported on additional considerations relating to CTC’s success in visualising the entire

colon in patients following an incomplete colonoscopy, visualising the proximal colon in patients with

a distal obstruction, in detecting extracolonic lesions, and test failure rates.

Purkayastha et al (2007)

The meta-analystic study by Purkayastha et al (2007) aimed to indirectly compare the diagnostic

accuracy of CTC with magnetic resonance colonography (MRC) (when compared with conventional

colonoscopy) for patients presenting with colorectal cancer (CRC). The literature search was relatively

broad considering several databases, and additional references from the bibliographies of the retrieved

articles. There were no language restrictions. Articles were considered to end of October 2005.

Included studies were prospective, blinded trials comparing CTC (or MRC) to visual colonoscopy and

reporting data on CRCs with sufficient information for calculations of sensitivity and specificity. Data

were extracted by three reviewers and discrepancy resolved by consensus with comprehensive

assessment of study quality undertaken using published guidelines.

Twelve studies were included comparing CTC with EC (data on MRC not reported here as beyond the

scope of the current review). Results indicated high sensitivity (0.96, 95% CI 0.92-0.99) and

specificity (1.00, 95% CI 0.99-1.00). There was a high area under the summary receiver operating

characteristic (SROC) curve (0.99) and high diagnostic odds ratio (DOR) (1461.90, 95% CI 135.00-

2448.56). Sensitivity analyses revealed that no factors improved diagnostic accuracy from CTC except

studies with more than 100 patients (AUC=1.00, DOR=2938.35, 95% CI 701.84-12,302.91).

In sum, Purkayastha et al’s (2007) meta-analysis indicated high diagnostic accuracy for CTC compared

with EC for detecting CRC in patients presenting with CRC.

Rosman and Korsten (2007)

A sophisticated meta-analysis of the accuracy of CTC using an sROC approach was recently conducted

by Rosman and Korsten (2007). The search was somewhat limited, considering publications to

November 2005, identified from one database, Medline. There was no reported citation searching or

Journal hand-searching. Inclusion criteria were that all subjects underwent CTC and colonoscopy (as a

reference standard), that studies reported per-patient sensitivity and specificity for polyp detection, and

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

that studies included at least five patients with disease or controls in the sample. Excluded were studies

that only reported per-polyp sensitivity (and not per-patient), studies with overlapping patients, or

studies with samples with excess colorectal cancers without sub-grouping.

Two researchers applied selection criteria but whether this was performed independently was not

reported. Appraisal and data extraction methods were not described. Detailed description of statistical

analyses were described. The reviewers suggested that statistical pooling in meta-analyses such as

employed by Mulhall et al (2005) may not be adequate, arguing that sensitivity and specificity

represent a trade-off and pooling them can give an inaccurate assessment. Rosman and Korsten (2007)

recommend that sROC curves be constructed in addition to pooled results.

Thirty studies were identified as eligible for inclusion in the meta-analysis. Results included that the

pooled per-patient sensitivity of CTC was higher for polyps greater than 10mm (0.82, 95% CI 0.76-

0.88) compared with polyps 6-10mm (0.63, 95% CI 0.52-0.75) and polyps 0-5mm (0.56, 95% CI 0.42-

0.70). The sROC curve analyses supported these findings with higher exact areas under the curve for

the threshold of over 10mm compared with thresholds of greater than 5 mm, and any size. Endoscopy

colonoscopy (EC) had significantly higher sensitivities and specificities than CTC at either a threshold

of greater than 5mm or greater than 10mm. At a threshold of greater than 5mm, the exact area under

the sROC curve was significantly higher for conventional colonoscopy compared with CTC (0.998 ±

0.006 vs 0.884 ±0.033, P < .005).

Other findings reported include that there were no significant differences in the diagnostic

characteristics of 2-dimensional versus 3-dimensional software algorithms (“fly-through”) for initial

analysis of the CT images. Further, based on only two studies, the reviewers concluded that CTC

seems to be more accurate than air-contrast barium enema for detecting polyps regardless of size.

Limitations were discussed, including that the design of studies comparing CTC and EC have an

inherent bias in favour of colonoscopy because EC is used as the “gold standard”. The use of segmental

unblinding attempts to address this but is of limited success because some polyps (such as those in

mucosal folds) may be difficult to locate even after a second unblended attempt.

Rosman and Korsten (2007) concluded from these results that CTC has reasonable sensitivity and

specificity for detecting large polyps but was less accurate than endoscopy colonoscopy for smaller

polyps. Given the limitations of CTC for small polyps, the authors argue that CTC may not be a

reasonable alternative in situations in which a small polyp may be clinically relevant. Therefore they

suggest that CTC should not be considered as a first-line screening test in patients with a strong family

history of CRC, where the progression of small polyps to cancer may require only 2-3 years. Further

the authors argue that if CTC is used for screening for people without a family history, it should be

repeated more frequently than surveillance schedules for conventional colonoscopy.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors, date Aim and search method Inclusion and exclusion criteria Results and authors’ conclusions Comments

(Halligan et al.

2005)

Aim

To assess the methodological

quality of available data in

published reports of computed

tomographic (CT) colonography

by performing a systematic

review and meta-analysis.

Search period: 1994 – December

2003.

Databases searched: Medline

(but a preliminary search to May

2003 included Cochrane

controlled trials register, EMBASE,

Science Citation Index),

handsearching of Journals (the

broader search did not find

articles that were found in

Medline).

Search terms: colonography,

colography, computed

tomographic colonography, CT

pneumocolon, virtual

conoloscopy, and virtual

endocscopy.

No language restrictions.

Additional references were

obtained from the bibliographies

of the retrieved articles. Authors

were contacted to determine

whether there was patient

overlap.

Inclusion criteria: the focus was the detection of

colorectal polyps verified with blinded within-

subject reference endoscopy colonoscopy (EC)

or surgical findings. Methods included full bowel

preparation, prone and supine images acquired,

and helical scanners used. Software used

needed to be (or mimic those) commercially

available, and allow 2-dimensional interpretation

and luminal 3-dimensional rendering for problem

solving, or use primary 3-dimensional

interpretation. Full reports from human in vivo

studies only considered.

Exclusion criteria: studies that used computer-

aided detection (CAD) systems (to avoid bias as

two authors work for a company that develops

such systems). Studies reported only as

Abstracts, studies with fewer than 30 patients.

Studies where the CT observers could guess that

the prevalence of abnormality was excessively

high (eg, where EC was incomplete due to an

obstructing tumour in more than 50% of

patients). Studies with artificially inserted polyps.

Studies where intravenous contrast material was

routinely administered to patients, or during

subsequent CT. Duplicate studies.

Appraisal methods: Two researchers conducted

searches, applied selection criteria and

abstracted data independently, resolving

discrepancies by consensus or monthly meetings

with other authors where necessary. Only

reported results from experienced observers.

Data on methodological quality recorded using

the QUADAS tool.

24 studies eligible of 1398 studies identified

by search strategy. One included

average risk patients only, and one was a

multi-centre study.

For detection of large polyps (1cm or

larger):

Per-patient Se (compared with EC): 93%,

95% CI 73%-98%.

Per-patient Sp (compared with EC): 97%,

95% CI 95%-99%.

For detection of medium and larger

polyps (6mm and above):

Per-patient Se (compared with EC): 86%,

95% CI 75%-93%.

Per-patient Sp (compared with EC): 86%,

95% CI 76%-93%.

For detection of polyps all sizes, studies

were too heterogenous in sensitivity

(range 45%-97%) and specificity (26%-

97%). The number of cancers per study

was too small to allow meta-analysis.

When treating data as if from a single

study, Se was 96% (95% CI 91%-99%)

Author/s Conclusions

CT colonography seems sufficiently

sensitive and specific in the detection of

large and medium polyps; it is especially

sensitive in the detection of symptomatic

cancer. Authors also concluded that CT

colonography should be further

investigated as a diagnostic tool for

cancer.

Comments

 relatively extensive search strategy, range of

databases and list of search terms with citation

checking, and hand-searching of Journals

 very detailed selection criteria provided

 search strategy and selection criteria applied

and quality rated independently by two

reviewers

 brief background section

 quantitative meta-analysis performed

 sensitivities and specificities determined

 several tables presented reasons for exclusion,

forest plots and sROC plots, and suggested

minimum data set

 too few studies to permit sensitivity analyses

investigating the effects of using a modified

reference standard (segmental unblinding of

colonoscopy), and individual observer

assessment versus consensus assessment

 makes point that data may not be applicable

to a screening situation

 an extremely detailed and comprehensive

assessment of study limitations in conduct,

design and reporting and description of

minimum data set based on difficulties with

data extraction.

Key: CAD: computer aided detection; CRC: colorectal cancer; CTC: computed tomographic colonography, EC = endoscopy colonoscopy, Se: sensitivity, Sp: specificity, sROC:

summary receiver operating characteristic

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2 continued. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors, date Aim and search method Inclusion and

exclusion criteria

Results and authors’ conclusions Comments

(Mulhall et al.

2005)

Aim

To systematically review the

test performance of CT

colonography compared to

colonoscopy or surgery and

to assess variables that may

affect test performance.

Search period: Jan 1975 –

February 2005.

Databases searched:

PubMed, Medline, EMBASE,

and the Cochrane

Controlled Trials Register.

English language only.

Search terms: virtual

colonoscopy, CT

colonography, CT

colography, CT

pneumocolon.

Selection criteria: Included

prospective trial of adults

undergoing CTC after full

bowel preparation, with

cologram interpreted blind to

the result of the EC or surgery

(as the gold reference

standard). Studies needed to

have used state-of-the-art

technology, including at least

a single-detector CT scanner

with supine and prone

positioning, insufflation of the

colon with air or carbon

dioxide, collimation smaller

than 5mm, and both 2-D and

3-D views during scan

interpretation.

Appraisal methods: Two

researchers independently

searched the literature,

applied selection criteria,

abstracted data and coded

aspects of methodological

quality (with disagreements

resolved by consensus).

33 studies included.

The pooled per-patient sensitivity of CTC was heterogenous and

improved as polyp size increased: for polyps <6mm (0.48, 95% CI

0.25-0.70), for polyps 6-9mm (0.70 95% CI 0.55-0.84), and for polyps

>9mm (0.85 95% CI 0.79-0.91).

Characteristics of the CTC scanner, including width of collimation

(thinner led to higher sensitivity), type of detector (multiple

scanners more sensitive than single detectors), and mode of

imaging (“fly through” more sensitive (but based on only 2 studies

using this method), explained some of the heterogeneity. Other

possible sources of false-negative results are discussed but limited

reporting and number of studies limited ability to investigate

them.

Specificity was homogenous, for polyps <6mm (0.92, 95% CI 0.89-

0.96), for polyps 6-9mm (0.93 95% CI 0.91-0.95, and for polyps

>9mm (0.97 95% CI 0.96-0.97).

Author/s Conclusions

Computed tomographic colonography is highly specific, but the

range of reported sensitivities is wide. Patient or scanner

characteristics do not fully account for this variability, but

collimation, type of scanner, and mode of imaging explain some

of the discrepancy. This heterogeneity raises concerns about

consistency of performance and about technical variability.

These issues must be resolved before CT colonography can be

advocated for generalised screening for colorectal cancer (from

Abstract). CT colonography should be used in research protocols

or when other accepted screening methods are not appropriate,

Comments

 moderately broad search strategy, several

databases and search terms. No citation

checking, hand-searching of Journals, or

contact with authors reported

 detailed selection criteria provided

 independent quality assessment and data

extraction

 brief introductory section

 thorough description of statistical analyses and

data synthesis

 quantitative meta-analysis with Se’s and Sp’s

weighted for sample size calculated and

heterogeneity explored using stratified analyses

and meta-regression techniques

 performed, pooled per-patient Se and Sp

calculated at various polyp size thresholds. No

sROC plots or characteristics reported

 several tables presented, for included relevant

studies, study design characteristics; per-patient

sensitivities tabulated by polyp threshold size,

scanner type (single slice or multidetector), and

mode of imaging; per-patient specificities

 very detailed table of patient characteristics

and sources of potential biases

 the Discussion considered limitations including

that the studies differed widely, with the

extractable variables explaining only a small

amount of variance. Only a few studies

considered the newest CTC technology.

Discussed limitations of gold standard, and that

only 3 studies were designed to evaluate a true

screening population of average risk people.

Key: CAD: computer aided detection, CRC: colorectal cancer, CTC: computed tomographic colonography, EC = endoscopy colonoscopy, Se: sensitivity, Sp: specificity, sROC:

summary receiver operating characteristic, 2-D: 2-dimensional, 3-D: 3-dimensional

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2 continued. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors, date Aim and search method Inclusion and

exclusion criteria

Results and authors’ conclusions Comments

(Nicholson et al.

2005b)

Aim: To review the role of

CT colonography in CRC

screening (from title).

Search period: not

provided, (paper

submitted March 2005,

accepted April 2005).

Databases searched:

Medline

Search terms: not

provided.

Cross-checking of

reference lists of retrieved

papers.

Considered original

articles and reviews

pertaining to CTC.

Explicit inclusion criteria not

provided.

Appraisal methods: not

described

Number of articles identified by search strategy not stated.

Test accuracy outcomes:

Some studies (lists five) reported promising sensitivity, greater than 90% for

detection of polyps ≥ 10mm in diameter, whilst others reported disappointing

results with Se 55-64% range (3 studies). Variation has been attributed to

inclusion of polyp-enriched population, inclusion of faecal tagging with

electronic cleansing, radiologist experience (and use of single institution

expertise), use of double-reading with 2 radiologists’ consensus interpretation

(unlikely to be available in clinical practice), or primary reliance on 3D image

review. CTC performs better in detecting polyps greater than 10mm than

those around 5mm. Cites Sosna et al’s (2003) meta-analysis of 14 prospective

studies using high-quality protocols (published to July 2002) which found a

pooled per patient Se for polyps 6-9mm as 0.84 and for polyps 10mm or larger

of 0.88 (and specificity of 0.95). Comments that difficult to draw firm

conclusions about how well CTC will detect flat polyps as under-investigated

in large trials of test accuracy. Also comments that it is not yet clear what 2D

and 3D imaging combination formats are optimal. Notes that CAD

technology should reduce the number of images that require careful review

by a radiologist.

Patient experience/safety outcomes:

Describes several studies considering patient preference, and notes that there

has been conflicting findings, with some finding similar acceptance of CTC

and EC, some finding preferences for CTC, and others finding greater

preference for EC. Results are limited by reliance on unvalidated subjective

questionnaires. Several studies suggest that patients generally find CTC less

painful and embarrassing than colonoscopy. Concludes that CTC may be

favored by patients compared with other screening tests due to the ease of

performance and comfort. Suggested that acceptance for CTC would

improve when noncathartic preparation methods perfected. Notes that

radiation exposure in the realm of what is offered with the accepted

screening technique of barium enema.

Author/s Conclusions

“Although not yet endorsed for widespread use by major gastroenterological

societies, CTC shows promise as a screening tool” (from Abstract).

Comments

 minimal detail of search strategy, or

selection criteria

 no detail of dates of search

 searched only one database

 checking of reference lists of retrieved

papers

 no detail of data extraction or appraisal

methodology

 narrative overview of considered

evidence under the following headings:

performance characteristics, flat lesions,

extra-colonic findings, indication for

CTC, patient preference, improving the

technique, and cost effectiveness

 some description of study designs,

sample characteristics and

technology/process variations where

relevant, and reasons for variations

between findings posed

 no tabular presentation review findings

 summary of various medical society

recommendations

 supported by industry funding

(unrestricted grant from Medicsight

PLC).

Key: CRC: colorectal cancer, CTC computed tomographic colonography, EC = endoscopy colonoscopy, Se: sensitivity, Sp: specificity

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2 continued. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors,

date

Aim and search

method

Inclusion and exclusion

criteria

Results and authors’ conclusions Comments

(Banerjee

and Van

Dam 2006)

Aim

To examine the

currently available

data on (CTC)

efficacy for

colorectal

screening.

Search period:

1960 – March 2005

Databases

searched: Medline.

Search terms:

virtual

conoloscopy,

virtual endoscopy,

computed

tomographic

colonography,

computerized

tomographic

colonography, CT

colonography,

computed

tomographic

colography,

computerized

tomographic

colography, CT

colography.

English language

only.

Inclusion criteria: CTC compared with

colonoscopy in prospective, blinded

trials.

Exclusion criteria: abstracts presented at

meeting that were not fully published as

full articles.

Appraisal methods: two investigators

examined each article independently,

and assigned a categorical rating of

quality (differences dealt with by

consensus). The three ratings were:

Evidence A: over 500 patients,

multicentre, multidetector CT scanners

used, segmental unblinding used.

Evidence B: over 100 patients, single or

multicentre, single row or multidetector

CT scanners, segmental unblinding not

used.

Evidence C: 100 patients or fewer, single

centre studies, single row CT scanners,

segmental unblinding not used.

Test accuracy results comparing CTC and endoscopy colonoscopy (EC) in

prospective, blinded trials.

Evidence C (poorest quality, n= 6 studies, published 1997 – 2002).

Per polyp Se for ≤5mm polyp 11%-55%.

Per polyp Se for 6-9mm polyp 36%-82%.

Per polyp Se for ≥10mm polyp, 50%-91%.

Per patient Se for 6-9mm polyp 43%-94%, Sp 58%-92%

Per patient Se for ≥10mm polyp 37%-96%, Sp 74%-96%

Evidence B, 7 studies, published 2000-2004

Per polyp Se for 6-9mm polyp 29%-82%.

Per polyp Se for ≥10mm polyp, 32%-93%.

Per patient Se for 6-9mm polyp 41%-93%, Sp 71%-95%

Per patient Se for ≥10mm polyp 35%-100%, Sp 92%-98%

Evidence A best quality, n=3 studies, published 2003 or later

1) Pickhardt et al (2003) study of asymptomatic patients at average or

increased risk for CRC. Involved careful bowel preparation with solid-stool

tagging and electronic cleansing, primary readings using 3D images and 2D

images for problem solving, and experienced readers. Per polyp Se for CTC

and EC in detection of ≥ 6mm polyps (86% and 90% respectively) and ≥10 mm

polyps (92% and 88% respectively). Per patient Se for ≥6mm polyps for CTC

and EC was 88% and 80% respectively, and for ≥10mm polyps, 94% and 96%

respectively.

2) Cotton et al (2004) – symptomatic patients or those with family history of

polyps at 9 academic centres. Se for 6-9mm polyps was 23% for CTC cf 96%

for EC. Se for ≥10 mm polyps was 52% for CTC cf 96% for EC. Results may

relate to inexperienced readers, as largest centre had Se >6 mm polyps of

82%.

3) Rockey et al (2005) – used superior scanners and more experienced reader

than Cotton et al, but still poor results for sample of increased risk subjects. Se

for 6-9mm polyps was 47% for CTC cf 99% for EC. Se for ≥10 mm polyps was

53% for CTC cf 99% for EC.

Comments

 limited search strategy, with no hand-

searching of Journals, or checking of

reference lists. English language

publications only

 single data base searched

 comprehensive list of search terms used

 selection criteria provided and quality

rated using independent ratings by two

reviewers

 brief background section

 tables present studies included in each

evidence level, the sample size,

population studied, and per patient Se,

per polyp Se and per patient Sp

 narrative summary in the text of the

studies that compare CTC with EC, with

some discussion of why results vary due

to technical and methodological

reasons

 in addition, there was brief description

of results of studies organised in relation

to technique, false negatives and false

positive on CTC, effects of training,

problem with flat and small polyps, CTC

in special situations, extracolonic

findings, patient preference, cost-

effectiveness, and upcoming advances

in CTC

 there was no final discussion or

conclusion section.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2 continued. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors,

date

Aim and search

method

Inclusion and exclusion

criteria

Results and authors’ conclusions Comments

(Banerjee

and Van

Dam 2006)

continued

Patient preferences. Concludes that it is unclear whether CTC is preferred to

EC, and mixed results described. Two larger trials cited with results including

similar levels of discomfort for each procedure, and no preference for either.

Bowel preparation was the worst part for both procedures. In a study by

Iannaccone et al (2004) where CTC with faecal tagging and no cathartic

preparation was offered, 61% preferred “prepless” CTC to EC following

catharsis, yet 35% still preferred EC as it allowed for immediate polyp removal

where necessary.

Safety. Radiation dose of 0.44 rem similar to undergoing 2 abdominal

radiographs. Only two cases of perforation from CTC arisen in patients due to

over-inflation with air in patients with diseased colons. No cases in average-

risk populations to date.

Author/s Conclusions

No overall conclusions made and no abstract summary.

Key: CTC computed tomographic colonography, EC = endoscopy colonoscopy, Se: sensitivity, Sp: specificity

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2 continued. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors, date Aim and search method Inclusion and

exclusion criteria

Results and authors’ conclusions Comments

(Davila et al.

2006)

Aim

To replace and

supplement a previous

guideline of the American

Society for Gastrointestinal

Endoscopy relating to the

use of gastrointestinal

endoscopy for colorectal

cancer screening and

surveillance.

Search period: not

provided.

Databases searched:

Medline.

Search terms: not

described.

Additional references were

obtained from the

bibliographies of the

identified articles and from

recommendations of

expert consultants.

Inclusion criteria: not

described

Exclusion criteria: not

described.

Appraisal methods: not

described.

Synthesises results from a number of studies to provide the following

results for CTC compared with endoscopy colonoscopy:

Per polyp Se for ≥6mm polyp 39%-94%, Sp 79%-92%

Per polyp Se for ≥10mm polyp 55%-100%, Sp of 94%-98%

Cites Rockey et al’s results also for people at high risk for CRC with Se

for

≥10mm lesions being 59% for CTC, compared with 48% for DCBE and

98% for colonoscopy. Notes that there are no studies reporting on

efficacy of CTC in reducing CRC incidence or mortality. Comments

that EC may detect clinically important extracolonic findings, and

that cost effectiveness studies indicate that under most assumptions

colonoscopy is more cost-effective than CTC.

Also reports on patient acceptance data reporting that comparative

studies show no consistent preference.

Suggests that whilst CTC cannot be recommended for primary

screening, it maybe useful for patients who refuse colonoscopy or

who have had an incomplete colonoscopy.

Author/s Conclusions

“Studies evaluating virtual colonoscopy [and faecal DNA testing] for

CRC screening have yielded conflicting results and therefore cannot

be recommended.”

Comments

 limited search strategy of one database

 checking of reference lists and experts

mentioned

 search terms and publication date range not

described

 selection criteria not provided

 brief background section

 no tabular presentation of results

 narrative summary in the text

 notes that an earlier Guideline published in 2003

focuses on CTC explicitly

 brief summary section.

Key: CRC: colorectal cancer; CTC: computed tomographic colonography, DCBE: double contrast barium enema, Se: sensitivity, Sp: specificity

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2 continued. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors,

date

Aim and search method Inclusion and exclusion criteria Results and authors’ conclusions Comments

(Medical

Services

Advisory

Committee

(MSAC)

2006)

Aim: To assess the safety,

effectiveness (accuracy,

patient preferences/quality

of life) and cost-

effectiveness of CTC for the

diagnosis or exclusion of

colorectal neoplasia in

symptomatic patients or in

patients that are

asymptomatic but at high

risk of colorectal neoplasia

due to a personal or family

history of colorectal polyps

or cancer, versus DCBE and

versus colonoscopy.

Evidence relevant to

patients who are ineligible

for colonoscopy was also

considered but not

reported here as beyond

scope of current Technical

Brief.

Search period: 1994 – June

2005.

Databases searched:

Medline, preMedline,

Current Contents,

Cochrane Library, Health

Technology Assessment

databases, and numerous

websites.

Inclusion criteria

Studies were to: perform multislice CTC (at least

4-slice CT scanning); use colonoscopy or surgical

findings as the reference standard; use double

contrast barium enema and/or colonoscopy as

a comparator. Studies had to report on at least

one of the following: diagnostic accuracy with

sufficient data to calculate sensitivity and

specificity; changes in clinical management;

patient outcomes (morbidity, mortality, adverse

events, quality of life, patient preferences).

Exclusion criteria:

Not an appropriate clinical study. Case series

where the use or reporting of a reference

standard was based on the CTC result

(positive/negative). Case-control studies where

patients were selected for inclusion in the study

based on their known disease status.

Retrospective case referent studies (reporting on

subjects all known to have the condition of

interest). Studies with average risk asymptomatic

patients and studies with < 10 patients

undergoing CTC were excluded. Studies which

compared two or more different techniques of

CTC without performing a reference standard

were excluded. Not in English.

Appraisal methods: A second researcher

scanned abstracts for potentially eligible articles,

with discrepancies resolved by discussion. Data

extracted using a standardized instrument by

one reviewer, and checked by a second, with

discrepancies resolved by discussion with a third.

Data on quality rated against explicit criteria

relevant to patient outcomes using the QUADAS

tool.

No studies compared overall health outcomes following the use of CTC,

DCBE or colonoscopy. Four systematic reviews and 24 clinical studies

reported on safety and accuracy, and 11 studies reported on patient

preferences or quality of life outcomes.

CTC accuracy: CTC is generally highly sensitive and specific for the

diagnosis or exclusion of cancers and polyps ≥ 10 mm in symptomatic

patients and asymptomatic patients at high risk of colorectal neoplasia:

11 studies of variable quality, median CTC sensitivity 84% (range 55-100%);

median CTC specificity 97% (range 74-100%). Estimates of CTC accuracy

are higher for the detection of cancer alone [meta-analysis of four

studies: CTC sensitivity 97% (95% CI 89-100%); CTC specificity 98% (95% CI

95-99%)]. These findings are consistent with results from three published

systematic reviews. CTC is only moderately sensitive for the detection of

lesions 6-9 mm (lesions 6-9 mm: six studies, CTC sensitivity range 30-80%,

CTC specificity range 93-99%). CTC is poorly sensitive for lesions < 6 mm

(lesions ≤ 5 mm: four studies, CTC sensitivity range 14-57%, CTC specificity

range 83-97%).

Relative accuracy of CTC, DCBE and colonoscopy:

Evidence limited to one study of fair quality (Rockey et al. 2005) that

found CTC and DCBE accuracy to be lower than noncomparative

studies, and systematic reviews of CTC accuracy. The study indicated

that CTC is a more specific test than DCBE, but less sensitive and specific

than colonoscopy for the detection of cancers and polyps ≥ 10 mm. The

study also suggested that CTC may be a more sensitive test than DCBE;

though this only reached statistical significance for lesions 6-9 mm.

Patient preferences: The evidence reviewed also suggests that CTC may

be preferred over colonoscopy. However, comparison of pain and

discomfort experienced by patients undergoing both tests have shown

mixed results with five of eight studies reporting results in favour of CTC,

and three in favour of colonoscopy.

Safety: CTC is a relatively safe procedure compared to DCBE and at least

as safe as, or safer than, diagnostic colonoscopy. Both CTC and DCBE

expose patients to ionizing radiation and are associated with a very small

risk of colonic perforation.

Comments

 an extremely detailed and

comprehensive review

presented in a 209 page

report

 extensive search strategy,

range of databases and list

of search terms with

citation checking

 no hand-searching of

Journals reported

 very detailed selection

criteria provided

 search strategy and

selection criteria applied

and quality rated by one

reviewer and checked by

a second

 extensive background

section on burden of

disease and compared

procedures

 quantitative meta-analysis

performed, sensitivities and

specificities pooled and

sROC plots fitted

 numerous tables presented

results

 additional considerations

were reviewed relating to

CTC’s success in visualising

the entire colon in patients

following an incomplete

colonoscopy, visualising

the proximal colon in

patients with a distal

obstruction, in detecting

extracolonic lesions, and

test failure rates.

Key: CRC: colorectal cancer; CTC: computed tomographic colonography, DCBE: double contrast barium enema; EC = endoscopy colonoscopy; MRC: magnetic resonance

imaging; MSAC: Medical Services Advisory Committee, Se: sensitivity, Sp: specificity

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2 continued. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors,

date

Aim and search method Inclusion and exclusion criteria Results and authors’ conclusions Comments

(Medical

Services

Advisory

Committee

(MSAC)

2006)

continued

Search terms: full search

terms for each database

were provided, including:

colonography, computed

tomographic, tomography,

X-ray computed, colorectal

neoplasms, colonic polyps,

cancer,

pseudoradiography, virtual

colonoscopy,

pneumocolon, spiral

computer assisted

tomography

No language restrictions on

the search.

Additional references were

obtained from the

bibliographies of the

retrieved articles.

An economic analysis was undertaken but results

are beyond the scope of this Technical Brief.

Clinical and consumer expertise involved

through the establishment of an advisory panel.

Author/s Conclusions

CTC is a relatively safe test compared to DCBE and colonoscopy.

Evidence about CTC accuracy for the detection of cancers and polyps ≥

10 mm compares favourably with DCBE. There is also some evidence to

suggest that patients prefer CTC over DCBE. CTC is less accurate than

colonoscopy for the detection of cancers and polyps ≥ 10 mm. There is

also some evidence to suggest that patients prefer CTC over

colonoscopy.

Recommendations

Evidence in relation to the comparison of CTC with colonoscopy indicates

that CTC is less effective. MSAC recommends that public funding for CTC

as a substitute investigation for colonoscopy should not be supported.

 reviewers commented that

variation observed

between studies

demonstrates that CTC is

less accurate in some

population subgroups or

settings. They further

argued that the extent to

which patient

characteristics, prevalence

of disease, CTC

techniques, the experience

of those performing and

interpreting the tests or

other factors may influence

CTC performance has not

yet been clearly defined.

Key: CRC: colorectal cancer; CTC: computed tomographic colonography, DCBE: double contrast barium enema; EC = endoscopy colonoscopy; MRC: magnetic resonance

imaging; MSAC: Medical Services Advisory Committee; Se: sensitivity, Sp: specificity

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2 continued. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors, date Aim and search method Inclusion and exclusion

criteria

Results and authors’ conclusions Comments

(Purkayastha et

al. 2007)

Aim

To use meta-regression

techniques to indirectly compare

the diagnostic accuracy of (i)

CTC compared with

conventional endoscopy

colonoscopy (EC) with that of (ii)

magnetic resonance

colonography (MRC), compared

with conventional colonoscopy,

for patients presenting with

colorectal cancer (CRC).

Search period: to 1 November

2005.

Databases searched: Ovid,

EMBASE, the Cochrane

database, Medline.

Search terms: computed

tomographic colonography,

magnetic resonance

colonography, virtual

conoloscopy, diagnostic

accuracy, colorectal cancer,

comparative study. Also used

related articles function.

No language restrictions.

Additional references were

obtained from the bibliographies

of the retrieved articles.

Study authors were not

contacted.

Inclusion criteria: all prospective studies

comparing CTC or MRC to EC and

reporting data on CRCs. CTC

compared with colonoscopy in

prospective, blinded trials.

Exclusion criteria: studies where patients

did not undergo EC, studies that did not

appear to report the numbers of people

diagnosed with CRC, and studies

providing insufficient information for

calculations of Se and Sp.

Appraisal methods: All data were

extracted by three reviewers and

discrepancy resolved by consensus.

Assessment of quality undertaken using

the guidelines published by STARD

initiative and the QUADAS tool.

Note that data per lesion or the

accuracy of different sized lesions was

not estimated as these data were not

available for CRCs in the studies

included.

Thirty studies eligible, 12 included studies

comparing CTC with EC (data on MRC not

reported here as beyond scope of current

review).

Overall Se for CTC (compared with EC): 0.96 95%

CI 0.92-0.99.

Overall Sp for CTC (compared with EC): 1.00 95%

CI 0.99-1.00.

Test showed high area under the summary

receiver operating characteristic (SROC) curve =

0.99 and high diagnostic odds ratio (DOR) =

1461.90, 95% CI 135.00-2448.56.

Sensitivity analyses revealed that no factors

improved diagnostic accuracy from CTC except

studies with more than 100 patients (AUC=1.00,

DOR=2938.35, 95% CI 701.84-12,302.91).

Author/s Conclusions

CTC and MRC have similar diagnostic accuracy

for detecting CRC. The accuracy, cost,

availability and practicality of CTC and MRC

have implications for future screening

programmes for CRC.

Comments

 relatively extensive search strategy, range of

databases and list of search terms with citation

checking, but with no hand-searching of Journals

or contact of experts

 selection criteria provided and quality rated using

independent ratings by three reviewers

 brief background section

 quantitative meta-analysis performed

 study quality assessed and sensitivities,

specificities, DORs calculated

 SROC curves and sensitivity analyses utilised

 several Tables presented study characteristics,

quality scores, study results, sensitivity analyses, the

sROC curve for diagnostic accuracy, and meta-

regression analyses

 discussion of the limitations of indirect comparisons

through meta-regression techniques.

Key: CRC: colorectal cancer; CTC: computed tomographic colonography, EC = endoscopy colonoscopy, MRC: magnetic resonance imaging; Se: sensitivity, Sp: specificity, sROC:

summary receiver operating characteristic

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 2 continued. Evidence Table of appraised secondary research relating to computed tomographic (CT) colonography for the detection of colorectal cancer.

Authors, date Aim and search method Inclusion and

exclusion criteria

Results and authors’ conclusions Comments

(Rosman and

Korsten 2007)

Aim

To evaluate the accuracy of CT

colonography for polyp

detection using an sROC

approach.

Search period: 1996 – November

2005.

Databases searched: Medline

Search terms: (virtual or CT or

computed or CAT) near “colon”.

Inclusion criteria: all subjects

who underwent CTC also

underwent colonoscopy (as a

reference standard), studies

reported per-patient sensitivity

and specificity for polyp

detection, and raw data for

determining these to allow

continuity correction.

Exclusion criteria: studies that

only reported per-polyp

sensitivity (and not per patient),

studies with overlapping

patients. Studies with fewer than

5 patients with disease or

controls or had excess

colorectal cancers without sub-

grouping.

Appraisal methods: Two

researchers applied selection

criteria.

30 studies included in the meta-analysis of CTC.

The pooled per-patient sensitivity of CTC was higher for polyps greater

than 10 mm (0.82, 95% CI 0.76-0.88) compared with polyps 6-10mm (0.63,

95% CI 0.52-0.75) and polyps 0-5mm (0.56, 95% CI 0.42-0.70). The sROC

curve analyses supported these findings with higher exact areas under

the curve for the threshold of over 10mm compared with thresholds of >5

mm and any size.

Endoscopy colonoscopy had significantly higher Se’s and Sp’s than CTC

at either a threshold of >5mm or greater than 10mm. At a threshold of

>5mm, the exact area under the sROC curve was significantly higher for

EC compared with CTC (0.998 ± 0.006 vs 0.884 ±0.033, P < .005).

There were no significant differences in the diagnostic characteristics of 2-

dimensional versus 3-dimensional software algorithms (“fly-through”) for

initial analysis of the CT images.

CTC seems to be more accurate than air-contrast barium enema for

detecting polyps regardless of size (based on two studies).

Author/s Conclusions

CT colonography has a reasonable sensitivity and specificity for detecting

large polyps but was less accurate than endoscopy (visual) colonoscopy

for smaller polyps. Given the limitations of CTC for small polyps, the authors

argue that it should not be considered as a first-line screening test in

patients with a strong family history of CRC. CT colonography may not be

a reasonable alternative in situations in which a small polyp may be

clinically relevant.

Comments

 limited search strategy, single

database, limited search terms,

and citation checking, hand-

searching of Journals, or contact

with authors

 detailed selection criteria

provided

 selection criteria applied by two

reviewers (not stated whether

independent)

 no description of quality

assessment or data extraction

methods

 brief introductory section

 thorough description of statistical

analyses

 quantitative meta-analysis

performed, pooled per-patient Se

and Sp calculated at various

polyp size thresholds, and sROC

curves constructed

 several tables presented sROC

plots, diagnostic characteristics

of included studies, 2D CTC, 3D

CTC, and endoscopic

colonoscopy

 the Discussion considers

limitations of statistical pooling

methods in meta-analyses, and

biases in favour of colonoscopy in

study designs.

Key: CAD: computer aided detection, CRC: colorectal cancer, CTC: computed tomographic colonography, EC = endoscopy colonoscopy, Se: sensitivity, Sp: specificity, sROC:

summary receiver operating characteristic, 2-D: 2-dimensional, 3-D: 3-dimensional

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Summary of review findings

Overview

This Technical Brief identified eight eligible secondary research publications, three published in 2005,

three in 2006, and two in 2007. Three were largely narrative reviews involving very limited searching

and no quantitative analysis (Banerjee and Van Dam 2006; Davila et al. 2006; Nicholson et al. 2005b).

The remaining five were systematic reviews including meta-analysis. Four were of high quality

including comprehensive searching and appraisal methods (Halligan et al. 2005; Medical Services

Advisory Committee (MSAC) 2006; Mulhall et al. 2005; Purkayastha et al. 2007), and the other

reported limited searching but employed sophisticated analysis (Rosman and Korsten 2007). Results

will be synthesised according to outcome type.

Safety

Two of the largely narrative reviews mentioned safety data. Nicholson et al (2005b) noted that

radiation exposure for CTC is in the realm of that received with barium enema, an accepted screening

technique. Banerjee and Van Dam (2006) observed that the radiation dose of 0.44 rem from CTC is

similar to that received when undergoing two abdominal radiographs. However they noted that this

level could still be of concern if CTC was used as a regular screening tool. The reviewers identified

only two cases of perforation from CTC arising in patients with diseased colons due to over-inflation

with air, and noted that no cases had been in average-risk populations (to date). Only one of the meta

analyses considered safety rigorously, the high quality systematic review of MSAC (2006). It found

CTC to be a relatively safe procedure compared to DCBE, and at least as safe as, or safer than,

diagnostic colonoscopy. Supporting the narrative reviews, the MSAC reviewers reported that CTC as

well as DCBE were said to expose patients to ionizing radiation and to be associated with a very small

risk of colonic perforation.

Patient experiences

All three of the poorer quality reviews considered patient experience outcomes. Nicholson et al

(2005b) summarised studies with conflicting findings, with some finding greater preference for CTC,

others favouring visual colonoscopy, and some reporting similar acceptance. However the authors also

cited several studies suggesting that patients generally find CTC less painful and embarrassing than

endoscopy colonoscopy. The reviewers suggested that CTC may become more acceptable once

noncathartic preparation methods have been perfected. The largely narrative review of Banerjee and

Van Dam (2006) concluded from mixed results that it was unclear whether CTC is preferred to

endoscopy colonoscopy, and cited two trials finding similar levels of discomfort expressed for each

procedure. Davila et al (2006) cited comparative studies on patient acceptance data which suggested

that there was no consistent preference.

Only one of the high quality systematic reviews and meta analyses reported on patient preferences, the

MSAC (2006) review. From 11 studies reporting on patient experience and quality of life outcomes,

the reviewers concluded that the evidence suggested that CTC may be preferred over colonoscopy.

However, comparison of pain and discomfort experienced by patients undergoing both tests showed

mixed results with five of eight studies reporting results in favour of CTC, and three in favour of

colonoscopy.

Test accuracy

Considering the three largely narrative reviews first, Nicholson et al’s (2005b) industry funded review

concluded that CTC performs better in detecting polyps greater than 10mm than those around 5mm in

diameter. Banerjee and Van Dam’s (2006) review considered evidence graded into quality levels.

From the three rated highest, one found high per-patient sensitivity for both CTC and EC, whereas test

accuracy for CTC was poorer in the other two trials. The third largely narrative review by Davila et al

(2006), which was designed to update a Guideline for the ASGE, described per-polyp sensitivity ranges

for polyps of greater to or equal to 6mm ranging 39% - 94%, and specificity ranging 79% - 92%. For

larger polyps of greater to or equal to 10mm, per-polyp sensitivity ranging 55% - 100%, and specificity

between 94% - 98%. It was noted that there were no studies reporting on efficacy of CTC in reducing

CRC incidence or mortality.

Summary statistics were available from five meta-analytic reviews appraised, although the polyp sizes

used for reporting varied. Data is summarised in Table 3. Note that the reviewers for the MSAC

(2006) report warned that as the sensitivities and specificities were statistically different across the

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

studies considered, pooled results may not provide a valid summary. Median results and/or ranges

were therefore described by the MSAC reviewers as the most appropriate summary statistics available.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Table 3. Summary of per-patient pooled sensitivity and specificity outcomes for CT colonography

Small polyps Medium polyps Large polyps Cancers Study

Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity

Halligan et al

(2005)

86% (95%

CI 75%-

93%)#

86%, (95%

CI 76%-

93%)#

93% (95%

CI 73%-

98%)†

97% (95%

CI 95%-

99%)†

Mulhall et al

(2005)

0.48 (95%

CI 0.25-

0.70)*

0.92 (95%

CI 0.89-

0.96)*

0.70 (95%

CI 0.55-

0.84)§

0.93 (95%

CI 0.91-

0.95)§

0.85 (95%

CI 0.79-

0.91)†

0.97 (95%

CI 0.96-

0.97)†

MSAC (2006)* Range:

14%-57%*

Range:

83%-97%*

Range:

30%-80%§

Range:

93%-99%§

Median

84% Range:

55%-100%†

Median

97% Range:

74%-100%†

97% (95%

CI 89-100%)

98% (95%

CI 95-99%)

Purkayastha et al

(2007)

0.96, 95% CI

0.92-0.99)

1.00 (95%

CI 0.99-

1.00)

Rosman and

Korsten (2007)‡

0.56 (95%

CI 0.42-

0.70)*

NR 0.63 (95%

CI 0.52-

0.75)+

NR 0.82, 95% CI

0.76-0.88)Ω

* <6mm, § 6-9mm, + 6-10mm, # ≥6mm, † ≥10mm, Ω >10mm, ‡ studies of average risk asymptomatic patients were not considered, NR: not reported

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Considering these results and the conclusions of the reviewers, there was general consensus that CTC

has reasonable sensitivity and specificity in the detection of large and medium polyps (Halligan et al.

2005; Medical Services Advisory Committee (MSAC) 2006; Rosman and Korsten 2007), but poorly

accurate for small lesions (2006). However, whilst specificity has been uniformly high, test

sensitivities have varied (based on performance and technical aspects) (Mulhall et al. 2005). The

mixed results have also been emphasised by two of the largely narrative reviews (Banerjee and Van

Dam 2006; Davila et al. 2006).

In comparisons between CTC and air contrast barium enemas based on two reviews, there appears to

be some evidence that CTC is more accurate than air-contrast barium enema for detecting polyps and

cancers (Medical Services Advisory Committee (MSAC) 2006; Rosman and Korsten 2007). However

conventional colonoscopy appears to be more accurate than CTC for large polyps (Medical Services

Advisory Committee (MSAC) 2006; Rosman and Korsten 2007), and particularly for smaller polyps

(Rosman and Korsten 2007).

There is a lack of evidence about the accuracy of CTC in average risk populations, and evidence to date

may not be applicable to a screening situation (Halligan et al. 2005). There are also concerns about use

of CTC for screening in high risk patients such as those with a strong family history given the

limitations of CTC in detecting small polyps which may be clinically relevant in such populations

(Rosman and Korsten 2007). None of the appraised reviews recommended CTC for generalised

screening. It has been suggested that until issues of heterogeneity based on performance and technical

variability are resolved, CTC should only be used in research protocols or when other accepted

screening methods are not appropriate (Mulhall et al. 2005). More positively, the largely narrative,

industry funded review of Nicholson et al (2005b) concluded that CTC shows “promise as a screening

tool”.

There appears to be some evidence that CTC is highly accurate in the detection of symptomatic cancer

(Halligan et al. 2005; Medical Services Advisory Committee (MSAC) 2006; Purkayastha et al. 2007)

and investigation of CTC as a diagnostic tool for cancer has been recommended (Halligan et al. 2005).

Factors affecting test accuracy

It is evident from variation observed between studies that CTC is less accurate in some population

subgroups or settings (Medical Services Advisory Committee (MSAC) 2006). Various factors have

been considered to explain for this heterogeneity. The largely narrative review by Nicholson et al

(2005b) noted that what 2D and 3D imaging combination format is optimal, has been under-

investigated. The meta-analyses appraised in this Technical Brief considered sensitivity analyses to

systematically investigate what factors might improve diagnostic accuracy. There were mixed

conclusions from these tests. In considering the detection of cancer in patients presenting with CRC,

Purkayastha et al’s (2007) sensitivity analyses revealed that no factors improved diagnostic accuracy

from CTC except studies with more than 100 patients. Considering polyp detection more broadly,

Halligan et al (2005) suggested that there were too few studies to investigate factors relating to

reference standard used or whether individual versus consensus assessment affected results. By

contrast, Mulhall et al (2005) reported that whilst patient or scanner characteristics do not fully account

for this variability, some factors explain some, including width of collimation (thinner led to higher

sensitivity), type of detector (multiple scanners more sensitive than single detectors), and mode of

imaging (“fly through” more sensitive, though based on only two studies using this method). More

recently, Rosman and Korsten (2007) concluded that there were no significant differences in the

diagnostic characteristics of 2-dimensional versus 3-dimensional software algorithms (“fly-through”)

for initial analysis of the CT images. In another recent meta analysis, MSAC’s (2006) reviewers

concluded that the extent to which patient characteristics, prevalence of disease, CTC techniques, the

experience of those performing and interpreting the tests or other factors may influence CTC

performance has not yet been clearly defined. However, they suggest that differences in study quality

(low to high), prevalence of lesions, the type of techniques used, and radiologist experience may

explain some of the variation observed.

Conclusions

Eight reviews were identified as eligible for inclusion in the Technical Brief published since 2005, five

of which were published since 2006. Three largely non-systematic reviews (Banerjee and Van Dam

2006; Nicholson et al. 2005b; van Dam et al. 2004) and five meta-analyses were identified as eligible

for inclusion and appraisal (Halligan et al. 2005; Medical Services Advisory Committee (MSAC) 2007,

Mulhall et al. 2005; Purkayastha et al. 2007; Rosman and Korsten 2007):

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Conclusions based on the results from the review are summarised below:

1. CT colonography is a relatively safe procedure compared to DCBE, and at least as safe as, or

safer than, diagnostic colonoscopy. Ionizing radiation exposure is relatively low but a

cumulative risk for regular screening. There is a very small risk of colonic perforation.

2. Generally there have been inconsistent findings regarding preferences for and experiences

from CTC versus colonography. However one meta-analysis of 11 studies of increased risk or

symptomatic patients concluded that CTC may be preferred over colonoscopy, and that the

majority of studies have reported results favouring CTC over colonoscopy with respect to pain

and discomfort.

3. CT colonography has reasonable test sensitivity and specificity in the detection of large and

medium polyps, but is poorly accurate for small lesions. Whilst specificity has been

consistently high, test sensitivities have varied and pooled statistics need to be considered with

caution. There is some evidence that CTC is highly accurate in the detection of symptomatic

cancer.

4. There is some evidence that CTC is more accurate than air-contrast barium enema for

detecting polyps and cancers in increased risk or symptomatic populations. However,

conventional colonoscopy appears to be more accurate than CTC for large polyps, and

particularly for smaller polyps.

5. Limitations of the current evidence base include that there is a lack of evidence about the

accuracy of CTC for primary screening in average risk populations. There is also a need for

greater investigation of the reasons for such wide variations in test accuracy achieved in

different trials with respect to patient and scanner characteristics. Likely sources of variation

relate to prevalence of disease, CTC techniques (such as width of collimation, type of

detector, and mode of imaging), and radiologist experience. The definition of what constitutes

a clinically important polyp in size and morphology also requires evidence-based elucidation.

Whilst the methodological quality of studies is improving, comparisons between studies

would be facilitated by more consistency in reporting and more appropriate statistical analysis

and data synthesis techniques.

6. There have been no studies reporting on overall health outcomes of CTC including efficacy in

reducing CRC incidence or mortality.

7. Based on the evidence and conclusions considered in this review CTC is not currently

recommended for generalised screening.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

REFERENCES

Banerjee, S., & Van Dam, J. (2006). CT colonography for colon cancer screening. Gastrointestinal

Endoscopy, 63, 121-133.

Colorectal Cancer Screening Advisory Group (2006). Report of the Colorectal Cancer Screening

Advisory Group. Wellington: Ministry of Health.

Davila, R. E., Rajan, E., & Baron, T. H. (2006). ASGE guideline: Colorectal cancer screening and

surveillance. Gastrointestinal Endoscopy, 63, 546-557.

Halligan, S., Altman, D. G., Taylor, S. A., Mallett, S., Deeks, J. J., Bartram, C. I., & Atkin, W. (2005).

CT colonography in the detection of colorectal polyps and cancer: systematic review, meta-

analysis, and proposed minimum data set for study level reporting. Radiology, 237, 893-904.

Iannaccone, R., Laghi, A., Catalano, C., Mangiapane, F., Lamazza, A., Schillaci, A., Sinibaldi, G., et

al. (2004). Computed tomographic colonography without cathartic preparation for the

detection of colorectal polyps. Gastroenterology, 127, 1300-1311.

Levin, B., Brooks, D., Smith, R. A., & Stone, A. (2003). Emerging technologies in screening for

colorectal cancer: CT colonography, immunochemical fecal occult blood tests, and stool

screening using molecular markers. Ca: a Cancer Journal for Clinicians, 53, 44-55.

Medical Services Advisory Committee (MSAC) (2006). Computed tomography colonography.

Canberra: MSAC.

Mulhall, B. P., Veerappan, G. R., & Jackson, J. L. (2005). Meta-analysis: computed tomographic

colonography. Annals of Internal Medicine, 142, 635-650.

Nicholson, F. B., Barro, J. L., Bartram, C. I., Dehmeshki, J., Halligan, S., Taylor, S., & Kamm, M. A.

(2005a). The role of CT colonography in colorectal cancer screening. American Journal of

Gastroenterology, 100, 2315-2323.

Nicholson, F. B., Taylor, S., Halligan, S., & Kamm, M. A. (2005b). Recent developments in CT

colonography. Clinical Radiology, 60, 1-7.

Perumpillichira, J. J., Yoshida, H., & Sahani, D. V. (2005). Computer-aided detection for virtual

colonoscopy. Cancer Imaging, 5, 11-16.

Pignone, M. P., Rich, M., Teutsch, S., Berg, A., & Lohr, K. (2002). Screening for colorectal cancer in

adults. Systematic evidence review No. 7. Bethesda, MD: Agency for Health Research

Quality (AHRQ).

Purkayastha, S., Athanasiou, T., Tekkis, P. P., Constantinides, V., Teare, J., & Darzi, A. W. (2007).

Magnetic resonance colonography vs computed tomography colonography for the diagnosis of

colorectal cancer: an indirect comparison. Colorectal Disease, 9, 100-111.

Rockey, D. C., Paulson, E., Niedzwiecki, D., Davis, W., Bosworth, H. B., Sanders, L., Yee, J., et al.

(2005). Analysis of air contrast barium enema, computed tomographic colonography, and

colonoscopy: prospective comparison. Lancet, 365, 305-311.

Rosman, A. S., & Korsten, M. A. (2007). Meta-analysis comparing CT colonography, air contrast

barium enema, and colonoscopy. American Journal of Medicine, 120, 203-210 e204.

Rex, D. K., Cutler, C. S., Lemmel, G. T., Rahmani, E. Y., Clark, D. W., Helper, D. J., Lehman, G. A.,

et al. (1997). Colonoscopic miss rates of adenomas determined by back-to-back

colonoscopies. Gastroenterology, 112, 24-28.

Sosna, J., Morrin, M. M., Kruskal, J. B., Lavin, P. T., Rosen, M. P., & Raptopoulos, V. (2003). CT

colonography of colorectal polyps: a metaanalysis. AJR American Journal of Roentgenology,

181, 1593-1598.

Walsh, J. M., & Terdiman, J. P. (2003). Colorectal cancer screening: scientific review. JAMA, 289,

1288-1296.

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

APPENDIX 1: SEARCH STRATEGY

Medline/Medline Pending/Cochrane Central Register of Controlled Trials

1 (virtual adj colonoscopy).mp.

2 (virtual adj colonography).mp.

3 (ct adj (colonograph$ or colonoscop$)).mp.

4 Colonography, Computed Tomographic/

5 exp Colonoscopy/

6 ct.mp.

7 5 and 6

8 or 1-4,7

9 limit 8 to english

10 animal/

11 animal/ and human/

12 10 not 11

13 9 not 12

14 limit 13 to yr=2005-2007

15 200406$.em.

16 200407$.em.

17 200408$.em.

18 200409$.em.

19 200410$.em.

20 200411$.em.

21 200412$.em.

22 or/15-21

23 13 and 22

24 14 or 23

Embase

1 (virtual adj colonoscopy).mp.

2 (virtual adj colonography).mp.

3 Computed Tomographic Colonography/

4 (ct adj (colonograph$ or colonosc$)).mp.

5 colonoscopy/

6 ct.mp.

7 5 and 6

8 or/1-4,7

9 limit 8 to english

10 limit 9 to yr=2005-2007

11 animal/

12 human/ and animal/

13 11 not 12

14 10 not 13

15 letter.pt.

16 14 not 15

17 200406$.em.

18 200407$.em.

19 200408$.em.

20 200409$.em.

21 200410$.em.

22 200411$.em.

23 200412$.em.

24 or/23-29

25 9 and 30

26 25 not (13 or 15)

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Current Contents

1. Virtual SAME colonoscopy

2. Virtual SAME colonography

3. ((colonograph* OR colonoscop*) SAME tomograph* SAME comput*)

4. CT SAME (colonoscop* OR colonograph*)

5. #1 OR #2 OR #3 OR #4

PubMed (last 60 days)

1. Virtual colonoscopy

2. Virtual colonography

3. CT colonoscopy

4. CT colonography

5. Colonography, computed tomographic[MESH]

6. #1 OR #2 OR #3 OR #4 OR #5

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

APPENDIX 2: EXCLUDED RETRIEVED PAPERS

A. G. A. Clinical Practice and Economics Committee (2006). Position of the American

Gastroenterological Association (AGA) Institute on computed tomographic colonography.

Gastroenterology, 131, 1627-1628.

Narrative review/expert opinion/commentary

Arnesen, R. B., Adamsen, S., Svendsen, L. B., Raaschou, H. O., von Benzon, E., & Hansen, O. H.

(2005). Missed lesions and false-positive findings on computed-tomographic colonography: a

controlled prospective analysis. Endoscopy, 37, 937-944.

Primary study

Ashar, B. H., Hughes, M. T., Marinopoulos, S. S., Prokopowicz, G. P., Berkenblit, G. V., Sisson, S. D.,

Simonson, L. A., et al. (2005). Current evidence for the use of emerging radiologic

technologies for disease screening. American Journal of Managed Care, 11, 385-392.

Narrative review/expert opinion/commentary

Banerjee S., & Van Dam, J. (2006a). CT colonography. Clinical Update, 13, 1-4.

Narrative review/expert opinion/commentary

Barish, M. A., & Rocha, T. C. (2005). Multislice CT colonography: current status and limitations.

Radiologic Clinics of North America, 43, 1049-1062.

Narrative review/expert opinion/commentary

Barish, M. A., Soto, J. A., & Ferrucci, J. T. (2005). Consensus on current clinical practice of virtual

colonoscopy. AJR American Journal of Roentgenology, 184, 786-792.

Narrative review/expert opinion/commentary

Bosworth, H. B., Rockey, D. C., Paulson, E. K., Niedzwiecki, D., Davis, W., Sanders, L. L., Yee, J., et

al. (2006). Prospective comparison of patient experience with colon imaging tests. American

Journal of Medicine, 119, 791-799.

Primary study

Burling, D., Halligan, S., Slater, A., Noakes, M. J., & Taylor, S. A. (2006). Potentially serious adverse

events at CT colonography in symptomatic patients: national survey of the United Kingdom.

Radiology, 239, 464-471.

Primary study

Burling, D., Taylor, S., & Halligan, S. (2004). Computerized tomography colonography. Expert Review

of Anticancer Therapy, 4, 615-625.

Narrative review/expert opinion/commentary

Burling, D., Taylor, S. A., & Halligan, S. (2005). Virtual colonoscopy: current status and future

directions. Gastrointestinal Endoscopy Clinics of North America, 15, 773-795.

Narrative review/expert opinion/commentary

Chung D. J., Huh, K. C., Choi, W. J., & Kim, J. K. (2005). CT colonography using 16-MDCT in the

evaluation of colorectal cancer. AJR American Journal of Roentgenology, 184, 98-103.

Primary study

Cotton, P. B., Durkalski, V. L., Benoit, P. C., Palesch, Y. Y., Mauldin, P. D., Hoffman, B., Vining, D.

J., et al. (2004). Computed tomographic colonography (virtual colonoscopy) - A multicenter

comparison with standard colonoscopy for detection of colorectal neoplasia. JAMA-Journal of

the American Medical Association, 291, 1713-1719.

Primary study

Desch, C. E., Benson, A., Somerfield, M. R., Flynn, P. J., Krause, C., Loprinzi, C. L., Minsky, B. D., et

al. (2005). Colorectal cancer surveillance: 2005 update of an American Society of Clinical

Oncology practice guideline. Journal of Clinical Oncology, 23, 8512-8519.

Ineligible as relates to detection of liver metastases and not CRC per se through the use of CT

scanning.

Fletcher, R. H. (2004). Virtual colonoscopy performed poorly in detecting colorectal neoplasia. ACP

Journal Club, 141, 23.

Narrative review/expert opinion/commentary

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Frentz, S. M., & Summers, R. M. (2006). Current status of CT colonography. Academic Radiology, 13,

1517-1531.

Narrative review/expert opinion/commentary

Gallo, T. M., Galatola, G., Laudi, C., & Regge, D. (2006). CT colonography: screening in individuals

at high risk for colorectal cancer. Abdominal Imaging, 31, 297-301.

Narrative review/expert opinion/commentary

Gollub, M. J., Schwartz, L. H., & Akhurst, T. (2007). Update on colorectal cancer imaging. Radiologic

Clinics of North America, 45, 85-118.

Narrative review/expert opinion/commentary

Hardacre, J. M., Ponsky, J. L., & Baker, M. E. (2005). Colonoscopy vs CT colonography to screen for

colorectal neoplasia in average-risk patients. Surgical Endoscopy, 19, 448-456.

Narrative review/expert opinion/commentary

Heiken, J. P., Peterson, C. M., & Menias, C. O. (2005). Virtual colonoscopy for colorectal cancer

screening: current status. Cancer Imaging, 5 Spec No A, S133-139.

Narrative review/expert opinion/commentary

Iannaccone, R., Catalano, C., Mangiapane, F., Murakami, T., Lamazza, A., Fiori, E., Schillaci, A., et al.

(2005). Colorectal polyps: detection with low-dose multi-detector row helical CT

colonography versus two sequential colonoscopies. Radiology, 237, 927-937.

Primary study

Iannaccone, R., Laghi, A., Catalano, C., Mangiapane, F., Lamazza, A., Schillaci, A., Sinibaldi, G., et

al. (2004). Computed tomographic colonography without cathartic preparation for the

detection of colorectal polyps. Gastroenterology, 127, 1300-1311.

Primary study

Koo, B. C., Ng, C. S., J, U. K.-I., Prevost, A. T., & Freeman, A. H. (2006). Minimal preparation CT for

the diagnosis of suspected colorectal cancer in the frail and elderly patient. Clinical Radiology,

61, 127-139.

Narrative review/expert opinion/commentary

Laghi, A. (2005). Virtual colonoscopy. Current Medical Imaging Reviews, 1, 303-312.

Narrative review/expert opinion/commentary

Lefere, P., Gryspeerdt, S., & Schotte, K. (2006). Virtual colonoscopy: an overview. Onkologie, 29,

281-286.

Narrative review/expert opinion/commentary

Lefkovitz, Z., Shapiro, R., Koch, S., & Cappell, M. S. (2005). The emerging role of virtual

colonoscopy. Medical Clinics of North America, 89, 111-138.

Narrative review/expert opinion/commentary

Leonardou, P., Striggaris, K., Pappas, P., Filippou, D., Bramis, I., Tsavaris, N., Gouliamos, A., et al.

(2006). Screening of patients after colectomy: virtual colonography. Abdominal Imaging, 31,

521-528.

Primary study

Limburg, P. J., & Fletcher, J. G. (2006). Making sense of CT colonography-related complication rates.

Gastroenterology, 131, 2023-2024; discussion 2024.

Letter

Macari, M., & Bini, E. J. (2005). CT colonography: where have we been and where are we going?

Radiology, 237, 819-833.

Narrative review/expert opinion/commentary

MacCarty, R. L., Johnson, C. D., Fletcher, J. G., & Wilson, L. A. (2006). Occult colorectal polyps on

CT colonography: implications for surveillance. AJR American Journal of Roentgenology,

186, 1380-1383.

Primary study

Mang, T., Graser, A., Schima, W., & Maier, A. (2007). CT colonography: techniques, indications,

findings. Eur J Radiol, 61, 388-399.

Narrative review/expert opinion/commentary

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

National Institute of Clinical Evidence (NICE) (2005). Computer tomographic colonography (virtual

colonoscopy). Interventional Procedure Guidance 129. NICE. Available from:

http://www.nice.org.uk/page.aspx?o=IPG129guidance

Accessed on 1.3.07

SR publication period prior to July 2005

Nicholson, F. B., Taylor, S., Halligan, S., & Kamm, M. A. (2005). Recent developments in CT

colonography. Clinical Radiology, 60, 1-7.

Narrative review/expert opinion/commentary

Nio, Y., Van Gelder, R. E., & Stoker, J. (2006). Computed tomography colonography: current issues.

Scandinavian Journal of Gastroenterology Supplement, 139-145.

Narrative review/expert opinion/commentary

O'Hare, A., & Fenlon, H. (2006). Virtual colonoscopy in the detection of colonic polyps and

neoplasms. Best Practice & Research in Clinical Gastroenterology, 20, 79-92.

Narrative review/expert opinion/commentary

Park, S. H., Ha, H. K., Kim, M. J., Kim, K. W., Kim, A. Y., Yang, D. H., Lee, M. G., et al. (2005).

False-negative results at multi-detector row CT colonography: multivariate analysis of causes

for missed lesions. Radiology, 235, 495-502.

Primary study

Perumpillichira, J. J., Yoshida, H., & Sahani, D. V. (2005). Computer-aided detection for virtual

colonoscopy. Cancer Imaging, 5, 11-16.

Narrative review/expert opinion/commentary

Pickhardt, P. J. (2005). Virtual colonscopy for primary screening: the future is now. Minerva

Chirurgica, 60, 139-150.

Narrative review/expert opinion/commentary

Pickhardt, P. J. (2006). Incidence of colonic perforation at CT colonography: review of existing data

and implications for screening of asymptomatic adults. Radiology, 239, 313-316.

Narrative review/expert opinion/commentary

Pickhardt, P. J., Choi, J. R., Hwang, I., Butler, J. A., Puckett, M. L., Hildebrandt, H. A., Wong, R. K.,

et al. (2003). Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in

asymptomatic adults. New England Journal of Medicine, 349, 2191-2200.

Primary study

Pilleul, F., Bansac-Lamblin, A., Monneuse, O., Dumortier, J., Milot, L., & Valette, P. J. (2006). Water

enema computed tomography: diagnostic tool in suspicion of colorectal tumor.

Gastroenterologie Clinique et Biologique, 30, 231-234.

Primary study

Ransohoff, D. F. (2005). Computed tomographic colonography without cathartic preparation performed

well in detecting colorectal polyps. Evidence Based Medicine, 10, 55.

Narrative review/expert opinion/commentary

Renkonen-Sinisalo, L., Kivisaari, A., Kivisaari, L., Sarna, S., & Jarvinen, H. J. (2007). Utility of

computed tomographic colonography in surveillance for hereditary nonpolyposis colorectal

cancer syndrome. Familial Cancer, 6, 135-140.

Primary study

Reuterskiold, M. H., Lasson, A., Svensson, E., Kilander, A., Stotzer, P. O., & Hellstrom, M. (2006).

Diagnostic performance of computed tomography colonography in symptomatic patients and

in patients with increased risk for colorectal disease. Acta Radiologica, 47, 888-898.

Primary study

Rex, D. K. (2005). Update on CT colonography (virtual colonoscopy) trials. Reviews in

Gastroenterological Disorders, 5, 227-229.

Narrative review/expert opinion/commentary

Rockey, D. C. (2005). Colon imaging: computed tomographic colonography. Clinical

Gastroenterology & Hepatology, 3, S37-41.

Primary study

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

Rockey, D. C., Paulson, E., Niedzwiecki, D., Davis, W., Bosworth, H. B., Sanders, L., Yee, J., et al.

(2005). Analysis of air contrast barium enema, computed tomographic colonography, and

colonoscopy: prospective comparison. Lancet, 365, 305-311.

Primary study

Solomon, M. J., Lord, S. J., & Walleser, S. (2006). Review: computed tomographic colonography is

accurate for medium and large colorectal polyps and cancer: Commentary. Evidence Based

Medicine, 11, 153.

Narrative review/expert opinion/commentary

Sosna, J., Kruskal, J. B., Bar-Ziv, J., Copel, L., & Sella, T. (2005). Extracolonic findings at CT

colonography. Abdominal Imaging, 30, 709-713.

Narrative review/expert opinion/commentary

Sosna, J., Sella, T., Bar-Ziv, J., & Libson, E. (2006). Perforation of the colon and rectum--a newly

recognized complication of CT colonography. Seminars in Ultrasound, CT & MR, 27, 161-

165.

Narrative review/expert opinion/commentary

Summers, R. M., Yao, J., Pickhardt, P. J., Franaszek, M., Bitter, I., Brickman, D., Krishna, V., et al.

(2005). Computed tomographic virtual colonoscopy computer-aided polyp detection in a

screening population. Gastroenterology, 129, 1832-1844.

Primary study

Taylor, S. A., Halligan, S., Burling, D., Bassett, P., & Bartram, C. I. (2005). Intra-individual

comparison of patient acceptability of multidetector-row CT colonography and double-

contrast barium enema. Clinical Radiology, 60, 207-214.

Primary study

Taylor, S. A., Laghi, A., Lefere, P., Halligan, S., & Stoker, J. (2007). European society of

gastrointestinal and abdominal radiology (ESGAR): Consensus statement on CT

colonography. European Radiology, 17, 575-579.

Narrative review/expert opinion/commentary

van Dam, J., Cotton, P., Johnson, C. D., McFarland, B. G., Pineau, B. C., Provenzale, D., Ransohoff,

D., et al. (2004). AGA future trends report: CT colonography. Gastroenterology, 127, 970-

984.

Narrative review/expert opinion/commentary

van Gelder, R. E., Birnie, E., Florie, J., Schutter, M. P., Bartelsman, J. F., Snel, P., Lameris, J. S., et al.

(2004a). CT colonography and colonoscopy: assessment of patient preference in a 5-week

follow-up study. Radiology, 233, 328-337.

Primary study

van Gelder, R. E., Nio, C. Y., Florie, J., Bartelsman, J. F., Snel, P., De Jager, S. W., Van Deventer, S.

J., et al. (2004b). Computed tomographic colonography compared with colonoscopy in

patients at increased risk for colorectal cancer. Gastroenterology, 127, 41-48.

Primary study

van Gelder, R. E., Florie, J., & Stoker, J. (2005). Colorectal cancer screening and surveillance with CT

colonography: current controversies and obstacles. Abdominal Imaging, 30, 5-12.

Narrative review/expert opinion/commentary

Wessling, J., Domagk, D., Lugering, N., Schierhorn, S., Heindel, W., Domschke, W., & Fischbach, R.

(2005). Virtual colonography: identification and differentiation of colorectal lesions using

multi-detector computed tomography. Scandinavian Journal of Gastroenterology, 40, 468-

476.

Primary study

Xiong, T., Richardson, M., Woodroffe, R., Halligan, S., Morton, D., & Lilford, R. J. (2005). Incidental

lesions found on CT colonography: their nature and frequency. British Journal of Radiology,

78, 22-29.

Narrative review/expert opinion/commentary

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

You, Y. T., Chang Chien, C. R., Wang, J. Y., Ng, K. K., Chen, J. S., Tang, R., Chiang, J. M., et al.

(2006). Evaluation of contrast-enhanced computed tomographic colonography in detection of

local recurrent colorectal cancer. World Journal of Gastroenterology, 12, 123-126.

Primary study

COMPUTED TOMOGRAPHIC (CT) COLONOGRAPHY FOR THE DETECTION OF COLORECTAL CANCER – A TECHNICAL BRIEF

APPENDIX 3: APPRAISED RETRIEVED PAPERS

Banerjee, S., & Van Dam, J. (2006). CT colonography for colon cancer screening. Gastrointestinal

Endoscopy, 63, 121-133.

Davila, R. E., Rajan, E., & Baron, T. H. (2006). ASGE guideline: Colorectal cancer screening and

surveillance. Gastrointestinal Endoscopy, 63, 546-557.

Halligan, S., Altman, D. G., Taylor, S. A., Mallett, S., Deeks, J. J., Bartram, C. I., & Atkin, W. (2005).

CT colonography in the detection of colorectal polyps and cancer: systematic review, meta-

analysis, and proposed minimum data set for study level reporting. Radiology, 237, 893-904.

Medical Services Advisory Committee (MSAC) (2006). Computed tomography colonography.

Canberra: MSAC.

Mulhall, B. P., Veerappan, G. R., & Jackson, J. L. (2005). Meta-analysis: computed tomographic

colonography. Annals of Internal Medicine, 142, 635-650.

Nicholson, F. B., Barro, J. L., Bartram, C. I., Dehmeshki, J., Halligan, S., Taylor, S., & Kamm, M. A.

(2005b). The role of CT colonography in colorectal cancer screening. American Journal of

Gastroenterology, 100, 2315-2323.

Purkayastha, S., Athanasiou, T., Tekkis, P. P., Constantinides, V., Teare, J., & Darzi, A. W. (2007).

Magnetic resonance colonography vs computed tomography colonography for the diagnosis of

colorectal cancer: an indirect comparison. Colorectal Disease, 9, 100-111.

Rosman, A. S., & Korsten, M. A. (2007). Meta-analysis comparing CT colonography, air contrast

barium enema, and colonoscopy. American Journal of Medicine, 120, 203-210 e204.

ResearchGate has not been able to resolve any citations for this publication.

Computed Tomographic Virtual Colonoscopy to Screen for Colorectal Neoplasia in Asymptomatic Adults

Article

Full-text available

Jan 2004
NEW ENGL J MED

We evaluated the performance characteristics of computed tomographic (CT) virtual colonoscopy for the detection of colorectal neoplasia in an average-risk screening population. A total of 1233 asymptomatic adults (mean age, 57.8 years) underwent same-day virtual and optical colonoscopy. Radiologists used the three-dimensional endoluminal display for the initial detection of polyps on CT virtual colonoscopy. For the initial examination of each colonic segment, the colonoscopists were unaware of the findings on virtual colonoscopy, which were revealed to them before any subsequent reexamination. The sensitivity and specificity of virtual colonoscopy and the sensitivity of optical colonoscopy were calculated with the use of the findings of the final, unblinded optical colonoscopy as the reference standard. The sensitivity of virtual colonoscopy for adenomatous polyps was 93.8 percent for polyps at least 10 mm in diameter, 93.9 percent for polyps at least 8 mm in diameter, and 88.7 percent for polyps at least 6 mm in diameter. The sensitivity of optical colonoscopy for adenomatous polyps was 87.5 percent, 91.5 percent, and 92.3 percent for the three sizes of polyps, respectively. The specificity of virtual colonoscopy for adenomatous polyps was 96.0 percent for polyps at least 10 mm in diameter, 92.2 percent for polyps at least 8 mm in diameter, and 79.6 percent for polyps at least 6 mm in diameter. Two polyps were malignant; both were detected on virtual colonoscopy, and one of them was missed on optical colonoscopy before the results on virtual colonoscopy were revealed. CT virtual colonoscopy with the use of a three-dimensional approach is an accurate screening method for the detection of colorectal neoplasia in asymptomatic average-risk adults and compares favorably with optical colonoscopy in terms of the detection of clinically relevant lesions.

Computed Tomographic Colonography (Virtual Colonoscopy): A Multicenter Comparison With Standard Colonoscopy for Detection of Colorectal Neoplasia

Article

Full-text available

Apr 2004
J Am Med Assoc

Conventional colonoscopy is the best available method for detection of colorectal cancer; however, it is invasive and not without risk. Computed tomographic colonography (CTC), also known as virtual colonoscopy, has been reported to be reasonably accurate in the diagnosis of colorectal neoplasia in studies performed at expert centers. To assess the accuracy of CTC in a large number of participants across multiple centers. A nonrandomized, evaluator-blinded, noninferiority study design of 615 participants aged 50 years or older who were referred for routine, clinically indicated colonoscopy in 9 major hospital centers between April 17, 2000, and October 3, 2001. The CTC was performed by using multislice scanners immediately before standard colonoscopy; findings at colonoscopy were reported before and after segmental unblinding to the CTC results. The sensitivity and specificity of CTC and conventional colonoscopy in detecting participants with lesions sized at least 6 mm. Secondary outcomes included detection of all lesions, detection of advanced lesions, possible technical confounders, participant preferences, and evidence for increasing accuracy with experience. A total of 827 lesions were detected in 308 of 600 participants who underwent both procedures; 104 participants had lesions sized at least 6 mm. The sensitivity of CTC for detecting participants with 1 or more lesions sized at least 6 mm was 39.0% (95% confidence interval [CI], 29.6%-48.4%) and for lesions sized at least 10 mm, it was 55.0% (95% CI, 39.9%-70.0%). These results were significantly lower than those for conventional colonoscopy, with sensitivities of 99.0% (95% CI, 97.1%->99.9%) and 100%, respectively. A total of 496 participants were without any lesion sized at least 6 mm. The specificity of CTC and conventional colonoscopy for detecting participants without any lesion sized at least 6 mm was 90.5% (95% CI, 87.9%-93.1%) and 100%, respectively, and without lesions sized at least 10 mm, 96.0% (95% CI, 94.3%-97.6%) and 100%, respectively. Computed tomographic colonography missed 2 of 8 cancers. The accuracy of CTC varied considerably between centers and did not improve as the study progressed. Participants expressed no clear preference for either technique. Computed tomographic colonography by these methods is not yet ready for widespread clinical application. Techniques and training need to be improved.

Virtual Colonoscopy

Article

Nov 2005

Andrea Laghi

Virtual colonoscopy (VC), also known as CT colonography, is a noninvasive imaging method for the examination of the colon. The technique is easy, less labour-intensive than barium enema and conventional colonoscopy and is inherently safer since no adverse event or procedure-related complications have ever been reported. The success rate of VC is approximately 100%, if bowel preparation and distension are optimal and no sedation is required. From the patient's perspective, the major advantages of VC include the very brief time required to perform the examination, the absence of contrast enemas and the potential for same-day colonoscopy when polyps are detected. To date several studies have reported sensitivity and specificity comparable with conventional colonoscopy in the detection of clinically significant polyps. However, the smaller the polyp size, the lower the sensitivity. The currently accepted clinical uses include the evaluation of patients who have undergone unsuccessful or incomplete conventional colonoscopy, patients with obstructing colorectal cancer, and those whose medical problems make them unsuitable for conventional colonoscopy. Current uses generally do not include the screening of asymptomatic persons, although recently published work has demonstrated no statistically significant differences between VC and conventional colonoscopy regardless of polyp size. A practical approach is to consider virtual colonoscopy 'as a currently credible alternative screening method and as a reasonable alternative to the other colorectal cancer screening tests when a patient is unable or unwilling to undergo conventional colonoscopy'.

Analysis of air contrast barium enema, computed tomographic colonography, and colonoscopy: Prospective comparison

Article

Jan 2005

Screening for Colorectal Cancer in Adults at Average Risk: A

Article

Data Extraction: One reviewer abstracted information from the final set of studies into evidence tables, and a second reviewer checked the tables for accuracy. Discrepancies were resolved by consensus. Data Synthesis: For average-risk adults older than 50 years of

Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies

Article

Jan 1997
GASTROENTEROLOGY

The miss rate of colonoscopy for neoplasms is poorly understood. The aim of this study was to determine the miss rate of colonoscopy by same day back-to-back colonoscopy. Two consecutive same day colonoscopies were performed in 183 patients. The patients were randomized to undergo the second colonoscopy by the same or a different endoscopist and in the same or different position. The overall miss rate for adenomas was 24%, 27% for adenomas < or = 5 mm, 13% for adenomas 6-9 mm, and 6% for adenomas > or = 1 cm. Patients with two or more adenomas at the first examination were more likely than patients with no or one adenoma detected at the first examination to have one or more adenomas at the second examination (odds ratio, 3.3; 95% confidence interval, 1.69-6.46). Right colon adenomas were missed more often (27%) than left colon adenomas (21%), but the difference was not significant. There was evidence of variation in sensitivity between endoscopists, but significant miss rates for small adenomas were found among essentially all endoscopists. Using current colonoscopic technology, there are significant miss rates for adenomas < 1 cm even with meticulous colonoscopy. Miss rates are low for adenomas > or = 1 cm. The results suggest the need for improvements in colonoscopic technology.

Emerging Technologies in Screening for Colorectal Cancer: CT Colonography, Immunochemical Fecal Occult Blood Tests, and Stool Screening Using Molecular Markers

Article

Jan 2003

The American Cancer Society's (ACS) Colorectal Cancer Advisory Group held a workshop on new technologies for the early detection of colorectal cancer and adenomatous polyps as part of a regular review of ACS guidelines for colorectal cancer screening. The Advisory Group formally reviewed CT colonography, immunochemical fecal occult blood tests (FOBT), and stool screening using molecular markers, and also addressed other technologies including capsule video endoscopy. With the exception of immunochemical stool testing, the ACS has determined that at this time there is insufficient evidence to recommend these technologies for routine colorectal cancer screening. Based on recommendations of the Advisory Group, only a minor modification has been made to the ACS's Recommendations for Screening and Surveillance of the Early Detection of Adenomatous Polyps and Colorectal Cancer.

Colorectal cancer screening: Scientific review

Article

Apr 2003

Screening for colorectal cancer clearly reduces colorectal cancer mortality, yet many eligible adults remain unscreened. Several screening tests are available, and various professional organizations have differing recommendations on which screening test to use. Clinicians are challenged to ensure that eligible patients undergo colorectal cancer screening and to guide patients in choosing what tests to receive. To critically assess the evidence for use of the available colorectal cancer screening tests, including fecal occult blood tests, sigmoidoscopy, colonoscopy, double-contrast barium enema, and newer tests, such as virtual colonoscopy and stool-based molecular screening. All relevant English-language articles were identified using PubMed (January 1966-August 2002), published meta-analyses, reference lists of key articles, and expert consultation. Studies that evaluated colorectal cancer screening in healthy individuals and assessed clinical outcomes were included. Evidence from randomized controlled trials was considered to be of highest quality, followed by observational evidence. Diagnostic accuracy studies were evaluated when randomized controlled trials and observational studies were not available or did not provide adequate evidence. Studies were excluded if they did not evaluate colorectal screening tests and if they did not evaluate average-risk individuals. Randomized controlled trials have shown that fecal occult blood testing can reduce colorectal cancer incidence and mortality. Case-control studies have shown that sigmoidoscopy is associated with a reduction in mortality, and observational studies suggest colonoscopy is effective as well. Combining fecal occult blood testing and sigmoidoscopy may decrease mortality and can increase diagnostic yield. The recommendation that all men and women aged 50 years or older undergo screening for colorectal cancer is supported by a large body of direct and indirect evidence. At present, the available evidence does not currently support choosing one test over another.

CT Colonography of Colorectal Polyps: A Metaanalysis

Article

Jan 2004

For proper evaluation of the accuracy of CT colonography, prospective multiinstitutional trials would be ideal. Until these trials are available, data can be collectively analyzed. The purpose of this study is to use metaanalysis to assess the reported accuracy of CT colonography compared with conventional colonoscopy for detecting colorectal polyps. Articles comparing CT colonography and conventional colonoscopy were identified, and a standardized form was used to extract relevant study data. Fisher's exact test and the Mantel-Haenszel test were used for pooling of data. A 95% confidence interval (CI) was selected to determine sensitivity and specificity, and the Kruskal-Wallis exact test was used to identify trends relating to polyp size. Meta-analysis methods were used to test strength of results. Comparisons were made for the percentage of polyps detected grouped by size (> or = 10 mm, 6-9 mm, < or = 5 mm) and the percentage of patients identified who had polyps of the same size. Fourteen studies fulfilled all the study inclusion criteria and gave a total of 1,324 patients and 1,411 polyps. The pooled per-patient sensitivity for polyps 10 mm or larger was (sensitivity [95% CI]) 0.88 (0.84-0.93), for polyps 6-9 mm it was 0.84 (0.80-0.89), and for polyps 5 mm or smaller it was 0.65 (0.57-0.73). The pooled per-polyp sensitivity for polyps 10 mm or larger was 0.81 (0.76-0.85), for polyps 6-9 mm it was 0.62 (0.58-0.67), and for polyps 5 mm or smaller it was 0.43 (0.39-0.47). Sensitivity for detection of polyps increased as the polyp size increased (p < 0.00005). The pooled overall specificity for detection of polyps larger than 10 mm was 0.95 (0.94-0.97). The specificity and sensitivity of CT colonography are high for polyps larger than 10 mm.

Virtual colonoscopy performed poorly in detecting colorectal neoplasia

Article

Jul 2004

Robert H. Fletcher

Computed tomographic (CT) colonography for the detection of colorectal cancer - a Technical Brief

Recommended publications

Accuracy of contrast-enhanced MDCT colononography for preoperative staging of colorectal carcinoma

CT colonography: Its place in the modern imaging diagnostic methods for early detection of colon dis...

Computed tomographic colonography (CTC) performance: one-year clinical follow-up

The Efficacy of Intravenous Contrast-enhanced 16-raw Multidetector CT Colonography for Detecting Pat...

CT Colonography

Two- versus Three-dimensional Colon Evaluation with Recently Developed Virtual Dissection Software f...