FDG PET‐CT imaging for pre operative staging in patients with colorectal cancer
Abstract
This is a protocol for a Cochrane Review (Diagnostic test accuracy). The objectives are as follows:
Primary colorectal cancer (replacement test)
The primary objective is to determine the diagnostic accuracy of integrated FDG PET‐CTas a replacement test to conventional imaging for the pre‐operative staging of primary colorectal cancer. The comparisons of interest are:
1. PET‐CT versus routinely used imaging modalities (e.g. CT chest/abdomen pelvis combined with pelvic MRI imaging) for pre operative staging of primary rectal cancer.
2. PET‐CT versus CT alone for pre‐operative staging of primary colonic cancer.
Recurrent colorectal cancer (add‐on test)
The primary objective is to determine the diagnostic accuracy of integrated FDG PET‐CT as an add‐on test for conventional imaging for the pre‐operative staging of recurrent colorectal cancer. The comparisons of interest are:
1. PET‐CT combined with pelvic MRI imaging versus routinely used imaging modalities (e.g. CT chest/abdomen/pelvis + MRI pelvis) for staging of recurrent rectal cancer
2. PET‐CT plus or minus MRI imaging versus routinely used imaging modalities (e.g. CT chest/abdomen/pelvis + or ‐ MRI pelvis) for staging of patients with recurrent colon cancer
Metastatic colorectal cancer (add‐on)
The primary objective is to determine the diagnostic accuracy of integrated FDG PET‐CT as a replacement test for conventional imaging for the pre‐operative staging of recurrent colorectal cancer. The comparisons of interest are:
1. PET‐CT versus clinical follow‐up including standard imaging as a replacement test for the detection of extra‐hepatic and intra‐hepatic lesions.
Background
Colorectal cancer (cancer of the large bowel) is the third most common cancer worldwide and the second most common cause of cancer death in the Western world. The global incidence was estimated at over one million new cases diagnosed worldwide in 2002 (GLOBOSCAN 2002). Accurate staging to determine the extent of local and distant disease is important to inform the decision about treatment. Curative surgery is an option for most patients with localised disease, some patients with recurrent disease may be suitable for salvage surgery, and surgical removal of metastases (metastectomy) may be an option for some patients with advanced disease.
Target condition being diagnosed
Accurate pre‐operative assessment of tumour stage, lymph node involvement and distant metastases is fundamental to informing clinical decisions about pre‐operative adjuvant radiochemotherapy for localised rectal cancer and to guide patient selection for salvage surgery and metastectomy (Herbertson 2009).
Index test(s)
PET is a type of nuclear medicine imaging that involves introducing a radioactive tracer into the patient’s body either by injection or oral ingestion prior to the scanning procedure. The glucose analogue F‐fluorine‐18‐deoxyglucose glucose (FDG) is currently the tracer most widely used in oncology. Tumours take‐up FDG through the natural process of glucose metabolism; the accumulated tracer emits gamma radiation that can be detected by the PET camera and appear as a bright spots on the PET image. FDG‐PET imaging alone is limited by a lack of anatomical detail, furthermore some colorectal cancers (such as mucinous adenocarcinomas) can be false negative due to limited uptake of FDG and false positive uptake can occur for example in the presence of inflammation (Herbertson 2009). Modern scanners now combine PET and CT in a single machine that performs both imaging studies ‐ these integrated PET‐CT images combine the anatomical detail of CT with the functional detail of PET. Another advantage of integrated PET‐CT is that the addition of the CT camera enables faster attenuation correction of the PET image thereby increasing patient throughput. While most PET‐CT scans are currently performed without contrast agents (non‐contrast enhanced CT) the role of contrast enhanced PET‐CT has been evaluated in more recent studies (Herbertson 2009).
Alternative test(s)
A number of imaging modalities are used in the pre‐operative staging of colorectal cancers including ultrasound (US), computerised tomography (CT), magnetic resonance imaging (MRI) and positron emission tomography (PET). Largely due to its wider availability and relatively low cost contrast enhanced CT is the technique most commonly used in conventional imaging for colorectal cancer diagnosis and tumour, lymph node and metastatic (TNM) staging. CT uses x‐rays to produce images of anatomical (structural) changes due to malignancy and an intravenous contrast agent, such as iodine, is used to increase the contrast between tumour and normal tissue (contrast enhanced CT). Conventional CT provides high spatial resolution but is limited in its ability to accurately distinguish benign from malignant processes on the basis of structural information alone, and image interpretation can be difficult where normal anatomy is distorted, for example by local scarring (Chin 2008). In the UK MRI is recommended for staging rectal cancers. (National Bowel Cancer Audit 2009) Supplementary imaging is often needed to provide sufficient information to inform surgical decisions however, US and MRI like CT produce purely anatomical images. For some patients the information obtained from imaging is still insufficient to avoid operative (surgical) evaluation to determine if curative resection is feasible (Hicks 2006).
Rationale
While the whole body approach to PET‐CT imaging allows broad coverage of multiple organs the radiation dose to the patient is potentially much higher than in conventional CT (Zanzonico 2008). The radiation dose per full body PET‐CT scan is around 15‐25 mSv of which 3‐15 mSv is attributable to CT radiation (Hays 2002). The global average background dose of radiation is approximately 2.4 mSv per person per year (WHO 2004). Reduction in the parameters of the CT scan in PET‐CT considerably reduces the effective radiation dose to the patient without compromising attenuation correction or registration of anatomical detail (Zanzonico 2008). Regulatory agencies in the USA, UK and several other European countries have concluded that the use of FDG is safe for oncology patients undergoing PET (FDA 2000; MHRA 2008). Staff radiation exposure can be maintained below regulatory limits by appropriate design, particularly shielding, and workflow in PET/CT facilities (Zanzonico 2008). Although the patient remains radioactive after the scan this reduces quickly over time (the half‐life of Fluorine‐18 is 109 minutes), however people in contact with the patient once they leave the hospital will receive a small dose of radiation. The associated risk of harm from such low doses is thought to be low although the risk to repeated contacts such as hospital drivers remains uncertain.
Several studies have shown PET‐CT to be more accurate than diagnostic CT and stand‐alone PET for cancer staging including colorectal cancer (Hicks 2006; Bar‐Shalom 2003;Cohade 2003). Replacing diagnostic CT with PET‐CT as the initial imaging investigation has considerable resource and cost implications. Currently, PET‐CT is widely recommended as an add‐on test for the assessment of suspected recurrence of colorectal cancer and in preoperative staging prior to metastectomy, and it may have an up‐and‐coming role in the initial staging of primary rectal cancer (Herbertson 2009; Ell 2006;Heriot 2004). In the UK the Royal College of Radiologists has issued guidance to recommend PET‐CT in four clinically distinct circumstances:
1. restaging potentially resectable local recurrence or limited metastatic disease in conjunction with MRI of the liver for the detection of small liver metastases.
2. Detecting/confirming recurrence where there is rising carcinoembryonic antigen (CEA) with negative or equivocal findings on cross sectional imaging.
3. Evaluation of pre sacral masses post treatment.
4. Restaging of rectal cancer following neo adjuvant chemotherapy.The Royal College of Radiologists 2010
While there is agreement regarding the role of PET‐CT in the identification of recurrent and metastatic disease clinical opinion on the role of PET‐CT in the routine management of primary colon cancer varies: Some investigators suggest that in certain clinical circumstances it should be considered as part of the standard preoperative assessment.Park 2006
Cost and resource considerations currently limit PET‐CT use to an add‐on test in most centres where the technology is available. The UK Royal college of Radiologists recommend that every cancer network should have access to PET‐CT. A systematic review is warranted to inform how diagnostic accuracy is changed by the replacement or addition of PET‐CT in the colorectal cancer staging pathway.
This systematic review aims to evaluate the diagnostic test accuracy of PET‐CT and its value as a replacement test for staging primary colorectal cancer and as an add‐on test for staging recurrent and metastatic colorectal cancer.
Objectives
Primary colorectal cancer (replacement test)
The primary objective is to determine the diagnostic accuracy of integrated FDG PET‐CTas a replacement test to conventional imaging for the pre‐operative staging of primary colorectal cancer. The comparisons of interest are:
1. PET‐CT versus routinely used imaging modalities (e.g. CT chest/abdomen pelvis combined with pelvic MRI imaging) for pre operative staging of primary rectal cancer.
2. PET‐CT versus CT alone for pre‐operative staging of primary colonic cancer.
Recurrent colorectal cancer (add‐on test)
The primary objective is to determine the diagnostic accuracy of integrated FDG PET‐CT as an add‐on test for conventional imaging for the pre‐operative staging of recurrent colorectal cancer. The comparisons of interest are:
1. PET‐CT combined with pelvic MRI imaging versus routinely used imaging modalities (e.g. CT chest/abdomen/pelvis + MRI pelvis) for staging of recurrent rectal cancer
2. PET‐CT plus or minus MRI imaging versus routinely used imaging modalities (e.g. CT chest/abdomen/pelvis + or ‐ MRI pelvis) for staging of patients with recurrent colon cancer
Metastatic colorectal cancer (add‐on)
The primary objective is to determine the diagnostic accuracy of integrated FDG PET‐CT as a replacement test for conventional imaging for the pre‐operative staging of recurrent colorectal cancer. The comparisons of interest are:
1. PET‐CT versus clinical follow‐up including standard imaging as a replacement test for the detection of extra‐hepatic and intra‐hepatic lesions.
Secondary objectives
We will assess adverse effects reported in the included studies.
Investigation of sources of heterogeneity
Several potential sources of heterogeneity have been identified in other (non‐Cochrane) systematic reviews and meta‐analyses of diagnostic imaging techniques in colorectal cancer (Bipat 2004; Bipat 2005; Halligan 2005; Huebner 2000; Purkayastha 2006; Wiering 2005). These were considered by the clinical authors of this review who identified the factors most likely to affect diagnostic accuracy in studies of FDG PET‐CT.
We will investigate the following potential sources of heterogeneity, using subgroup analysis where possible: academic (e.g. university hospital) versus non‐academic setting; indication known or suspected; study conducted up to 2005 and post‐2005 (reflecting differences in PET‐CT technology); blinding of index and reference standard tests and their interpretation (or not). Heterogeniety arising from different reference standards will also be explored.
Heterogeneity in the statistical analysis will initially be assessed graphically and where possible using meta‐regression (see Investigations of heterogeneity, below).
Methods
Criteria for considering studies for this review
Types of studies
Prospective and retrospective patient series (diagnostic cohort) and randomised test accuracy studies will be eligible for inclusion. Both consecutive series and series that are not explicitly reported as consecutive will be included. Diagnostic case‐control studies (two‐gate design) will be excluded because clinically relevant estimates of specificity and sensitivity can only be derived from the clinical population and not healthy controls.
Participants
Adults with known or suspected primary, recurrent and metastatic cancer of the colon or rectum, undergoing pre‐operative staging prior to curative surgery in a secondary care setting, will be eligible for inclusion. Patients with any stage of disease will be included. Studies solely in patients with anal cancer will be excluded because this rare cancer differs from colorectal cancers both biologically and in terms of the treatment pathway. Studies that include colon and rectal and anal cancer patients where data are not reported separately for the colon/rectal and anal cancer groups will be included in the review only if less than 20% had anal cancer (the effect of including any such studies will be explored using sensitivity analysis where possible).
Index tests
Integrated FDG PET‐CT. Studies using both contrast enhanced and non‐contrast enhanced CT will be included.
Comparator tests
Standard imaging tests including diagnostic CT, MRI and PET, alone or in combination.
Target conditions
The accurate pre‐operative assessment of tumour stage, lymph node involvement and distant metastases in primary, recurrent and metastatic CRC.
Reference standards
Surgical histopathology is the gold standard for colorectal cancer staging, however patients who do not undergo surgical resection will have test results verified by an alternative standard. These include histopathology based on biopsy and clinical and imaging follow‐up. Studies using these reference standards singly or mixed will be eligible for inclusion. Any duration of follow‐up and frequency of testing will be included.
Search methods for identification of studies
Studies will be identified through searching a range of electronic databases and a handsearch of key journals. Studies will also be identified through contact with experts in the field. Reference lists of relevant studies and existing review articles will be scanned for relevant studies.
Electronic searches
A wide range of databases to cover the clinical, experimental and technological aspects and shown to yield diagnostic tests accuracy studies (Whiting 2008) will be searched. This selection of databases also covers ongoing studies and the grey literature:
-
Biosis Previews (via ISI Web of Knowledge, 1926 to date of search)
-
CINAHL Plus (via EbscoHost 1937 to date of search)
-
Compendex (via EiVillage 1972 to date of search)
-
Dissertations and Theses (all content to date of search)
-
Embase (via OVIDSP 1980 to date of search)
-
Global Health (via OvidSP 1973 to date of search)
-
Global Health Library Regional databases (comprising LILACS, AFRO, EMRO, PAHO, WHOLIS, WPRO ‐ all content to date of search)
-
Index to theses (all content to date of search)
-
Inspec (via EiVillage 1972 to date of search)
-
Medline (via OvidSP 1950 to date of search)
-
mRCT (all content to date of search)
-
NTIS (all content to date of search)
-
OpenSIGLE (archive ends 2005)
-
UKCRN (all content to date of search)
-
Web of Science, including Conference Proceedings Citation Index, (via Web of Knowledge 1900 to date of search)
The search terms will incorporate a range of relevant database subject headings and text words capturing the concepts of colorectal cancers and PET. Database searches will be as sensitive as possible, in line with guidance on searches for tests of diagnostic accuracy (Ritchie 2007), and with no language or date limits. A Medline search sample is included in Appendix 1.
Searching other resources
Studies will also be identified through contact with experts in the field. In addition, information on studies in progress, unpublished research or research reported in the grey literature will be identified by searching databases including the UKCRN portfolio, the metaRegister of Controlled Trials, Clinicaltrials.gov, the (US) National Cancer Institute trials resource, OpenSIGLE, Digital Dissertations, and Index to Theses. We will also search the reference lists of relevant studies and existing review articles for relevant studies.
Data collection and analysis
Selection of studies
One reviewer (FCr) will screen the titles and abstracts retrieved by the electronic searches and a second reviewer (FCh) will check the decisions on a random sample of 25%. Full papers will be obtained for potentially eligible studies. Two reviewers (HMc, FCr) will then independently apply the inclusion criteria to the full papers and resolve disagreements by discussion. The inclusion criteria will be applied in the same way to the full reports of studies identified through sources other than electronic databases. An overview of the selection process will be summarised in a flow diagram.
Data extraction and management
Data will be extracted by two reviewers (HMc, FCr) independently using a standard form, which will include the quality assessment criteria, and disagreements will be resolved by discussion. Data to populate 2X2 contingency tables consisting of the number of true positives (TP), true negatives (TN), false positives (FP) and false negatives (FN) will be extracted as reported, including both patient‐level and lesion‐level data, and qualitative and quantitative definitions of diagnostic thresholds. Numbers of un interpretable test results will also be extracted.
Data will also be extracted to describe the clinical characteristics of FP and FN PET‐CT findings, and additional cases detected and cases restaged by the use of PET‐CT; actual changes in planned management directed by PET‐CT findings and the clinical consequences of the changes. Data on mortality, adverse events (including how these were monitored and recorded), and technical failures for both index and comparator tests, will also be extracted.
To facilitate interpretation of the findings we will also extract data on co morbidities (e.g. diabetes) and previous treatment in the study population; the PET‐CT system; fasting duration; FDG dose and time between administration of FDG and performance of the scan; comparator imaging test system(s), patient preparation and test interpretation; interval between index and comparator tests (more or less than 3 months); assessors (number, expertise, experience, consensus procedures and learning effect data); and in regard to the reference standard whether histopathology was by surgery or biopsy and the duration, frequency, type and interpretation of clinical and imaging follow‐up tests, and the numbers of patients whose results were confirmed by each type of reference standard.
Assessment of methodological quality
Eleven items from the QUADAS checklist will be used to assess the methodological quality of the included studies. Two reviewers (HMc, FCr) will apply the criteria independently and resolve disagreements by discussion. The results of the quality assessment will be used for descriptive purposes to provide an evaluation of the overall quality of the included studies and to investigate potential sources of heterogeneity (Whiting 2004).
The classification of responses to each of the QUADAS items is summarised in Table 1: Quality Assessment operational definitions).
| 1. Was the spectrum of patients’ representative of the patients who will receive the test in practice? Yes: the study included a consecutive series or random sample of adults undergoing staging for known or suspected primary cancer of the colon or rectum who received PET‐CT at any point in the preoperative staging pathway, either as the first imaging investigation or referred following equivocal or suspicious findings on other tests including imaging, clinical exam and carcinoembryonic antigen (CEA) test No: the study included a non‐consecutive or non‐random sample, or there is clear evidence of selective sampling e.g. restriction to patients with particular findings on PET‐CT Unclear: insufficient information on the method of recruitment and selection criteria. |
| 2. Were selection criteria clearly described? Yes: A clear definition of the inclusion/exclusion criteria is described No: The inclusion/exclusion criteria are not defined Unclear: There is a definition of the inclusion/exclusion criteria but it is ambiguous |
| 3. Is the reference standard likely to correctly classify the target condition? Yes: surgical histopathology if surgical resection in patients who undergo surgery, or clinical/imaging follow‐up in patients who do not receive a surgical resection No: Yes criteria clearly not met Unclear: Insufficient information |
| 4. Is the time period between the index test and reference standard short enough to be reasonably sure that the target condition did not change between the two tests? Yes: if the time between PET‐CT and the reference standard is less than six weeks. The clinical rational being a finding on PET CT in a patient with colon or rectal cancer should be acted upon within 6 weeks of study, or disease progression may occur. Similarly a negative finding should really be confirmed within the same timescale. For patients who do not have a surgical resection clinical/imaging follow‐up should be conducted after at least 6 months. No: Yes criteria clearly not met Unclear: the time lapse between tests was not reported |
| 5. Did the whole sample or a random selection of the sample, receive verification using a reference standard of diagnosis? Yes: it is reported that all or a random sample of the study participants did receive verification of their disease status using the reference standard defined in 2 above. No: all participants did not receive verification using the reference standard and those who did were not selected randomly Unclear: insufficient information |
| 6. Did patients receive the same reference standard regardless of the index test result? Yes: all PET‐CT results were verified by the same reference standard No: some PET‐CT results were verified by a different reference standard Unclear: insufficient information |
| 7. Where a composite reference standard was used did all participants receive all elements of the composite reference standard? Yes: it is clear that all patients received all elements of the composite reference standard No:it is clear that all patients did not receive all elements of the composite reference standard Unclear: it is not reported if all patients received all elements of the composite reference standard or not |
| 8. Was the reference standard independent of the index test (i.e. the index test did not form part of the reference standard)? Yes: PET‐CT was not used in establishing the final diagnosis (i.e. PET‐CT was not a component of the reference standard) No: PET‐CT did form part of the reference standard (NB this does not include the PET‐CT index test result being known when the reference standard diagnosis was made, only the specific incorporation of PET‐CT as part of the reference standard test) Unclear: insufficient information |
| 9.Was the execution of the index test described in sufficient detail to permit its replication? Yes: Details of the PET/CT equipment and patient preparation were clearly described No: Information regarding the PET/CT equipment and patient preparation were missing from the report Unclear:Insuffient information |
| 10.Was the execution of the reference standard described in sufficient detail to permit its replication? Yes:Details of the reference standard(s) (surgically resected specimen +/‐clinical/imaging follow up) were clearly described No:Details of the reference standard(s) (surgically resected specimen +/‐clinical/imaging follow up) were not clearly described Unclear: Insufficient information |
| 11. Were the index test results interpreted without knowledge of the results of the reference standard? Yes: it is clearly stated that PET‐CT results were interpreted blind to the results of the reference standard No: it is clearly stated that PET‐CT results were interpreted with knowledge of the results of the reference standard Unclear: insufficient information |
| 12. Were the reference standard results interpreted without knowledge of the results of the index test? Yes: it is clearly stated that the results of the reference standard were interpreted blind to the PET‐CT results No: it is clearly stated that the results of the reference standard were interpreted with knowledge of the PET‐CT results Unclear: insufficient information |
| 13. Were the same clinical data available when the test results were interpreted as would be available when the test is used in practice? Yes: it is stated that the clinical information, including all previous test results, that would normally be available to the assessors when PET‐CT images are interpreted in practice was available when the index test was evaluated No: it is stated that the clinical information that would normally be available when PET‐CT images are interpreted in practice was not available to the assessors when the index test was evaluated Unclear: insufficient information |
| 14. Were un interpretable/intermediate test results reported? Yes: it is clear that all test results including those that were un interpretable, indeterminate or intermediate are reported No: it is reported that un interpretable, indeterminate or intermediate were excluded with no further information given, or there is reason to believe that such results occurred but were not reported Unclear: it is not clear whether all the study results were reported |
| 15. Were withdrawals from the study explained? Yes: all participants who were entered in the study either completed the study or were accounted for if they did not complete the study (reasons for withdrawal and loss to follow‐up were given) No: some participants who entered the study did not complete the study and were not accounted for Unclear: insufficient information to judge whether all participants who entered the study were accounted for |
Statistical analysis and data synthesis
To ensure that the results of the review are of interest to clinicians, the disease stages will be dichotomised into categories corresponding to surgery versus no surgery. However, we note that there are methodological issues surrounding this. In particular, as we are including patients with suspected colorectal cancer but not confirmed cases, a patient may not be a candidate for surgery because he or she does not have colorectal cancer. Alternatively, the patient may not be a candidate for surgery as the cancer is so extensive that there would be no benefit from surgery. In effect we have three groups of patients, the cancer‐free, patients with cancer that has not spread to the extent to rule out surgery, and patients with extensive disease.
We would not combine data from the first and third patient groups and instead propose two analyses if the data are available. The first would dichotomise the data into cancer‐free versus candidates for surgery. The second would use data only from cancer patients, dichotomised into surgery versus non‐surgery candidates.
Results of the evaluation of the accuracy of PET‐CT will be analysed in the following way:
Due to methodological problems, particularly that caused by the difficulty of estimating within‐study variance where patients contribute more than one data point, and the individual patient data are not available, the 2x2 tables will be reported for the lesion‐level data and the analyses will be restricted to the patient‐level data.
The 2x2 tables for the patient‐level data will be used to calculate sensitivity, specificity with confidence intervals. Data will be plotted graphically in both forest and receiver operating characteristic (ROC) plots. The forest plots will give an indication of the extent of heterogeneity between studies and the ROC plot if any heterogeneity is likely to be due to the threshold effect.
A random effects meta‐analysis will be undertaken to fit the bivariate SROC curve model (Harbord 2006) with the within‐study variances fitted as binomial. The DiagMeta package in R will be used for all meta‐analyses (http://www.r‐project.org/) Gentleman 2011 . If the data are not amenable to bivariate random effects meta‐analyses, separate meta‐analyses each for sensitivity and specificity will be presented instead, using fixed or random effects as appropriate depending on the degree of heterogeneity. Estimates will include the average sensitivity and specificity for each test and of differences in sensitivity and specificity between PET‐CT and each comparator.
Data for outcomes other than diagnostic test accuracy (changes in patient management, mortality) will be presented as relative and absolute risks, and, where appropriate, meta‐analysis will be used to calculate pooled estimates using relative risk with 95% confidence intervals.
Investigations of heterogeneity
Heterogeneity due to individual diagnostic studies using different diagnostic thresholds will be explored as a standard part of fitting the bivariate SROC curve model. Heterogeneity due to other study characteristics will be explored with meta‐regression, but only if the data are adequate for such analyses. In either case, graphical displays of estimates from individual studies grouped according to the prespecified sources of heterogeneity will be provided.
Sensitivity analyses
Where possible we will conduct sensitivity analysis by including only prospective studies (excluding retrospective), studies with explicitly consecutive samples (excluding non‐consecutive or unclear), studies with a histopathology (surgery or biopsy) reference standard for all participants (excluding studies where some or all participants received only clinical or imaging follow‐up as the reference standard), and studies that included only rectal and colon cancer patients (excluding anal cancer).
Assessment of reporting bias
A formal assessment of reporting bias will not be undertaken as there are as yet no accepted methods to do this (Brazzeli 2009). However, the possibility of reporting bias will be highlighted and the results of any analysis interpreted cautiously.
| 1. Was the spectrum of patients’ representative of the patients who will receive the test in practice? Yes: the study included a consecutive series or random sample of adults undergoing staging for known or suspected primary cancer of the colon or rectum who received PET‐CT at any point in the preoperative staging pathway, either as the first imaging investigation or referred following equivocal or suspicious findings on other tests including imaging, clinical exam and carcinoembryonic antigen (CEA) test No: the study included a non‐consecutive or non‐random sample, or there is clear evidence of selective sampling e.g. restriction to patients with particular findings on PET‐CT Unclear: insufficient information on the method of recruitment and selection criteria. |
| 2. Were selection criteria clearly described? Yes: A clear definition of the inclusion/exclusion criteria is described No: The inclusion/exclusion criteria are not defined Unclear: There is a definition of the inclusion/exclusion criteria but it is ambiguous |
| 3. Is the reference standard likely to correctly classify the target condition? Yes: surgical histopathology if surgical resection in patients who undergo surgery, or clinical/imaging follow‐up in patients who do not receive a surgical resection No: Yes criteria clearly not met Unclear: Insufficient information |
| 4. Is the time period between the index test and reference standard short enough to be reasonably sure that the target condition did not change between the two tests? Yes: if the time between PET‐CT and the reference standard is less than six weeks. The clinical rational being a finding on PET CT in a patient with colon or rectal cancer should be acted upon within 6 weeks of study, or disease progression may occur. Similarly a negative finding should really be confirmed within the same timescale. For patients who do not have a surgical resection clinical/imaging follow‐up should be conducted after at least 6 months. No: Yes criteria clearly not met Unclear: the time lapse between tests was not reported |
| 5. Did the whole sample or a random selection of the sample, receive verification using a reference standard of diagnosis? Yes: it is reported that all or a random sample of the study participants did receive verification of their disease status using the reference standard defined in 2 above. No: all participants did not receive verification using the reference standard and those who did were not selected randomly Unclear: insufficient information |
| 6. Did patients receive the same reference standard regardless of the index test result? Yes: all PET‐CT results were verified by the same reference standard No: some PET‐CT results were verified by a different reference standard Unclear: insufficient information |
| 7. Where a composite reference standard was used did all participants receive all elements of the composite reference standard? Yes: it is clear that all patients received all elements of the composite reference standard No:it is clear that all patients did not receive all elements of the composite reference standard Unclear: it is not reported if all patients received all elements of the composite reference standard or not |
| 8. Was the reference standard independent of the index test (i.e. the index test did not form part of the reference standard)? Yes: PET‐CT was not used in establishing the final diagnosis (i.e. PET‐CT was not a component of the reference standard) No: PET‐CT did form part of the reference standard (NB this does not include the PET‐CT index test result being known when the reference standard diagnosis was made, only the specific incorporation of PET‐CT as part of the reference standard test) Unclear: insufficient information |
| 9.Was the execution of the index test described in sufficient detail to permit its replication? Yes: Details of the PET/CT equipment and patient preparation were clearly described No: Information regarding the PET/CT equipment and patient preparation were missing from the report Unclear:Insuffient information |
| 10.Was the execution of the reference standard described in sufficient detail to permit its replication? Yes:Details of the reference standard(s) (surgically resected specimen +/‐clinical/imaging follow up) were clearly described No:Details of the reference standard(s) (surgically resected specimen +/‐clinical/imaging follow up) were not clearly described Unclear: Insufficient information |
| 11. Were the index test results interpreted without knowledge of the results of the reference standard? Yes: it is clearly stated that PET‐CT results were interpreted blind to the results of the reference standard No: it is clearly stated that PET‐CT results were interpreted with knowledge of the results of the reference standard Unclear: insufficient information |
| 12. Were the reference standard results interpreted without knowledge of the results of the index test? Yes: it is clearly stated that the results of the reference standard were interpreted blind to the PET‐CT results No: it is clearly stated that the results of the reference standard were interpreted with knowledge of the PET‐CT results Unclear: insufficient information |
| 13. Were the same clinical data available when the test results were interpreted as would be available when the test is used in practice? Yes: it is stated that the clinical information, including all previous test results, that would normally be available to the assessors when PET‐CT images are interpreted in practice was available when the index test was evaluated No: it is stated that the clinical information that would normally be available when PET‐CT images are interpreted in practice was not available to the assessors when the index test was evaluated Unclear: insufficient information |
| 14. Were un interpretable/intermediate test results reported? Yes: it is clear that all test results including those that were un interpretable, indeterminate or intermediate are reported No: it is reported that un interpretable, indeterminate or intermediate were excluded with no further information given, or there is reason to believe that such results occurred but were not reported Unclear: it is not clear whether all the study results were reported |
| 15. Were withdrawals from the study explained? Yes: all participants who were entered in the study either completed the study or were accounted for if they did not complete the study (reasons for withdrawal and loss to follow‐up were given) No: some participants who entered the study did not complete the study and were not accounted for Unclear: insufficient information to judge whether all participants who entered the study were accounted for |
