Article Text
PDF
Abstract
Objective Colon capsule endoscopy (CCE) has shown promise for colorectal neoplasia detection compared with optical colonoscopy (OC), but has not been compared with other screening tests in average risk screening patients.
Design Patients 50 to 75 years of age (African Americans, 45–75 years) were randomised to CCE or CT colonography (CTC) and subsequent blinded OC. The primary endpoint was diagnostic yield of polyps ≥6 mm with CCE or CTC. Secondary endpoints included accuracy for size and histology, examination completeness, number/proportion of subjects with polyps and adenomas ≥6 mm and ≥10 mm, subject satisfaction and safety.
Results From 320 enrolled subjects, data from 286 (89.4%) were evaluable. The proportion of subjects with any polyp ≥6 mm confirmed by OC was 31.6% for CCE versus 8.6% for CTC (pPr non-inferiority and superiority=0.999). The diagnostic yield of polyps ≥10 mm was 13.5% with CCE versus 6.3% with CTC (pPr non-inferiority=0.9954). The sensitivity and specificity of CCE for polyps ≥6 mm was 79.2% and 96.3% while that of CTC was 26.8% and 98.9%. The sensitivity and specificity of CCE for polyps ≥10 mm was 85.7% and 98.2% compared with 50% and 99.1% for CTC. Both tests were well tolerated/safe.
Conclusion CCE was superior to CTC for detection of polyps ≥6 mm and non-inferior for identification of polyps ≥10 mm. CCE should be considered comparable or superior to CTC as a colorectal neoplasia screening test, although neither test is as effective as OC.
Trial registration number ClinicalTrials.gov no: NCT02754661.
- colorectal cancer screening
- endoscopic procedures
- colonic adenomas
- colonic polyps
Data availability statement
Data may be obtained from a third party and are not publicly available. The study protocol, statistical analysis plan and study results are available on ClinicalTrials.gov (https://clinicaltrials.gov/ct2/show/NCT02754661). Individual participant data are not available for sharing.
Statistics from Altmetric.com
Significance of this study
What is already known on this subject?
Despite multiple diagnostic tests endorsed for colorectal cancer (CRC) screening for average risk adults, compliance with screening recommendations is suboptimal.
There are relatively few comparative trials of imaging tests for CRC screening.
Colon capsule endoscopy (CCE) has shown promise for colorectal neoplasia detection compared with optical colonoscopy (OC), but has not been widely compared to other screening tests.
What are the new findings?
This is the first comparison of CCE to CT colonography (CTC) in an average risk CRC screening population.
CCE was superior to CTC for the detection of colon polyps ≥6 mm and serrated polyps, and non-inferior to CTC for the identification of colon polyps ≥10 mm.
Accuracy of polyp detection in patients undergoing screening with CCE was comparable with colonoscopy.
Both CCE and CTC were safe and well tolerated, but more patients preferred OC to both CCE and CTC.
How might it impact on clinical practice in the foreseeable future?
These results support a sequential approach to CRC screening in which OC is considered the first-line test for patients who desire or require a screening test that requires bowel preparation and maximises polyp detection by imaging, with CCE and CTC reserved for patients who decline OC or in whom OC carries elevated risk.
CCE should be considered a relevant CRC screening option in an average risk population in line with a ranking comparable with or higher than CTC, as well as a first-line imaging test after incomplete OC.
Introduction
Colorectal cancer (CRC) is the second leading cause of cancer death in the USA and Europe.1–3 Recent declines in the incidence of CRC have been attributed to heightened awareness and uptake of screening, which, in the USA, has increased from 21% in 2000 to 74.4% in 2016.2 In Europe, despite the introduction of no-cost or low-cost population-based occult blood test screening programmes in multiple countries within the European Union (EU), less than 50% of adults surveyed in 2014 had undergone CRC screening.3 While the only CRC screening test recommended by the EU is the faecal occult blood test (FOBT),4 multiple CRC screening tests are recommended in various US guidelines.5–8
Offering patients choices among the various CRC screening options can increase adherence.9 Optical colonoscopy (OC) imaging alternatives such as flexible sigmoidoscopy, CT colonography (CTC) and colon capsule endoscopy (CCE) are predicated on the visualisation of neoplasia (polyps and cancer), with the latter two tests permitting evaluation of the entire colorectum. This characteristic is in contrast to non-invasive, non-imaging CRC screening tests such as high-sensitivity FOBT, faecal immunochemical testing (FIT) and the multitarget stool DNA test (mt-sDNA), which primarily detect prevalent CRC. While most of these tests have been compared with OC and are considered cost-effective for CRC screening, there are relatively few comparisons among non-OC imaging tests, limiting determination of test hierarchy and standardisation of guideline recommendations.10
CCE relies on a disposable ingestible capsule designed to acquire video images during natural propulsion through the digestive system. This technology has been extensively studied11 and has demonstrated a favourable safety profile and comparable accuracy with OC for the detection of polyps and adenomas ≥6 mm.12–17 In the USA, the second-generation PillCam COLON 2 CCE is approved by the US Food and Drug Administration (FDA) for use in patients after an incomplete OC with adequate preparation and in patients with major risks for OC or moderate sedation who have evidence of GI bleeding of lower GI origin.18 In Europe, CCE is suggested as an option for a variety of clinical scenarios.19
CT colonography (CTC) is a minimally invasive modality for examining the colon.20 The 2020 joint guideline from the European Society of Gastrointestinal Endoscopy (ESGE) and European Society of Gastrointestinal and Abdominal Radiology (ESGAR) recommend CTC as the radiological examination of choice for the diagnosis of colorectal neoplasia.19 However, while this guideline does not recommend CTC as a primary test for population screening, it does recommend CTC as a primary test on an individualised basis in the absence of an organised FIT-based population CRC screening programme. United States guideline recommendations regarding CTC for CRC screening are variable, with some recommending CTC for average risk screening and others recommending CTC when OC is refused or incomplete.5–8 21 22
The inherent similarity of test characteristics between CCE and CTC (both are non-invasive or minimally invasive, non-therapeutic imaging modalities) make both tests potentially attractive screening tests for patients at average risk for CRC who decline OC or stool-based screening options. Considering the significantly different test requirements and methodologies for CCE and CTC, it is conceivable that one test may have superior performance than the other. Several studies have compared the performance of CCE and CTC in patients with positive FIT or incomplete colonoscopy and have found the diagnostic yield of CCE to be at least comparable, and in some studies superior, with that of CTC.23–28 To date, no studies have compared the two tests in an average risk CRC screening population. In order to further inform CRC screening guideline development and recommendations regarding these options, the TOPAZ study was designed to compare the performance of the PillCam COLON 2 CCE device with CTC in an average risk CRC screening population.
Materials and methods
Study design
This was a randomised, controlled, parallel-group clinical study conducted between September 2016 and August 2018 in 13 sites in the USA. The full study protocol and statistical analysis plan can be found in ClinicalTrials.gov. All authors had access to the study data and reviewed and approved the final manuscript.
Study participants
Eligible subjects between 50 and 75 years of age (African Americans, 45–75 years of age) classified as average risk for CRC and willing and able to participate in the study procedures were included. Patients were excluded if any of the following were present: history of CRC or adenoma, history of CRC screening within testing intervals recommended by the US Multisociety Task Force on Colorectal Cancer Screening (USMSTF), haematochezia, iron deficiency, or positive FOBT or FIT of any variety. Patients were also excluded if they had any current condition associated with an increased risk of capsule retention; dysphagia or any swallowing disorder; any current serious medical conditions so severe that CRC screening would have no benefit; cardiac pacemaker or other implanted electromedical device; were expected to undergo MRI examination within 7 days after ingestion of the capsule; estimated glomerular filtration rate outside of the normal reference range within the past 6 months; diagnosis of gastroparesis, small bowel or large bowel dysmotility; allergies or known contraindications to the medications or preparation agents used in the study procedures; estimated life expectancy of less than 6 months; and were pregnant or actively breast feeding.
Study procedure
After obtaining informed consent, subjects were randomised in a 1:1 ratio to receive CRC screening by either CCE or CTC. Randomisation was performed centrally using an Interactive Web Response System and the randomisation schedule was generated using nquery software for N=400 subjects using blocks of four subjects to ensure that CCE/CTC and central video-reviewing would be continuously balanced during the study. Investigators were unaware of the randomisation schedule until they randomised a patient.
Subjects were to undergo a confirmatory OC within 5 weeks of the CCE or CTC procedure for which the endoscopist was blinded to reader interpretations of CCE or CTC. These interpretations were enclosed in sealed, opaque envelopes and were opened at the conclusion of the OC for endoscopist unblinding. Colonoscopy procedures were video recorded and each polyp photographed with a 9 mm biopsy forceps aligned along its longest axis for subsequent polyp adjudication if required. Immediately after unblinding of CCE or CTC results, an unblinded OC was performed on subjects who were found to have lesions 6 mm or larger by either CCE or CTC that were not identified during blinded OC. All identified polyps were resected and sent for pathology assessment.
Bowel preparation regimens for all three procedure types were standardised across sites (online supplemental figure 1). Any adverse events (AEs) that occurred during CCE, CTC and OC were recorded. Telephone follow-ups were conducted 5–9 days after the CCE or CTC and again 5–9 days after the OC and any additional AEs were recorded. On the second follow-up call, subjects were asked questions regarding procedure preference between OC and CCE or CTC using a non-validated preference questionnaire.
Supplemental material
All CCE videos and CTC images were evaluated by one of two expert central readers who documented study quality as well as location and size of any colorectal polyps or masses. Two sets of central readers were used, one each for CCE videos and CTC images. Readers provided their findings within 2 weeks of CCE or CTC. Both procedures were also evaluated by local readers at each research site. Local CCE readers were experienced gastroenterologists with appropriate training/qualification to read CCE videos. These readers were excluded from performing any of the study related OCs in order to maintain OC blinding. Local CTC readers were radiologists with appropriate training/qualification to interpret CTC images. On study completion, an expert panel adjudicated polyp size measurements of all cases of size discrepancies between CCE or CTC and OC as well as a control group of randomly selected 30% of true positive (TP) cases.
Study outcomes
The primary endpoint of the TOPAZ trial was the diagnostic yield (proportion of subjects shown to have an actionable lesion confirmed by OC, defined as any polyp or mass lesion ≥6 mm) of CCE compared with CTC. Diagnostic yield of CCE or CTC was calculated in relation to OC results.
Secondary endpoints included accuracy measures (sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV)) of CCE versus CTC for the detection of polyps ≥6 mm in relation to OC results and expressed on a per-patient basis. Additional secondary analyses included the diagnostic yield for polyps ≥10 mm by CCE compared with CTC, accuracy measures (sensitivity, specificity, NPV, PPV) of CCE versus CTC for the detection of polyps ≥10 mm in relation to OC results, accuracy measures of CCE versus CTC for the detection of polyps ≥6 mm and ≥10 mm assessed in relation to histopathology results expressed on a per-subject basis (sensitivity, specificity, NPV, PPV), proportion of subjects achieving a complete examination with CCE compared with CTC, proportion of subjects shown to have an adenoma ≥6 mm and ≥10 mm by CCE compared with CTC, number of polyps ≥6 mm and ≥10 mm identified with CCE compared with CTC, number of adenomas ≥6 mm and ≥10 mm identified with CCE compared with CTC, and accuracy according to histology class (conventional adenomas and serrated lesions), subject satisfaction and procedure preference. Safety endpoints were AEs reported by number, severity, timing and relationship to the CCE, CTC or OC.
Sample size
The sample size was estimated based on the primary non-inferiority hypothesis under a Bayesian adaptive design. Interim assessments were performed at every 80 subjects (40 per arm) with a maximum of 320 subjects (160 per arm). Early success was claimed if the pre-defined success criterion was met. The following parameter assumptions were used in the estimation, based on prior literature.15
Predicted CCE diagnostic yield: 25%.
Predicted CTC diagnostic yield: 15%.
Non-inferiority hypothesis with a 5% margin.
Accounting for approximately 10% attrition.
With these assumptions, a total of 320 subjects (160 per arm) were needed to provide more than 80% power to test the primary hypothesis at the significance level of 5%. In addition, if non-inferiority was proven, superiority was also tested. The study was not powered for analysis of ≥10 mm polyps.
Statistical analysis
The primary effectiveness endpoint was based on the per-protocol analysis set (PPAS), which included all randomised subjects without major protocol deviations, those without incomplete CCE or CTC, or those who did not meet any of the following criteria: subject withdrew, CCE remained in the stomach or small bowel during the entire procedure, OC could not be done or there was a system technical failure. In order to assess the potential impact of data exclusion and derive the most conservative test characteristic estimates, a comparative analysis of the primary endpoint and all secondary endpoints and safety analyses were conducted using the full analysis set (FAS). Unless otherwise specified, all statistical comparisons were carried out using two-tailed tests. Type I error level for testing the difference between the two groups was set at 5%, including the use of a t-test or non-parametric Wilcoxon rank-sum test for continuous variables and Pearson χ2 test or Fisher’s exact test for categorical variables. Continuous data were summarised using descriptive statistics, specifically the number of observations (N), mean, SD, median, minimum (Min) and maximum (Max). Categorical data were summarised using frequency counts and percentages.
Polyp matching
The primary analysis used a rigid, but clinically oriented rule termed the whole colon Rex matching rule (online supplemental table 1). According to this algorithm, the polyp with the largest estimated diameter ≥6 mm seen on any imaging modality (CCE, CTC or OC) was referred to as the ‘reference’ polyp. The largest polyp identified on the additional test undergone by the subject was considered a match if the size of the reference polyp and its 50% plus/50% minus range overlapped with the 50% plus/50% minus range of the size measured by the other test, without regard to colon segment. Results using additional matching rules were also evaluated in order to test results across different algorithms (online supplemental tables 2–4).
Pathology
Polyps were classified into four different groups: conventional adenomas, serrated lesions, hyperplastic and other. Lesions interpreted as tubular, tubulovillous or villous adenomas were considered conventional adenomas. Lesions interpreted as sessile serrated polyps, sessile serrated adenomas, serrated adenomas or traditional serrated adenomas were considered serrated lesions. Polyps interpreted as hyperplastic polyps were considered hyperplastic polyps. Lesions interpreted as other included normal mucosal tissue, lipoma or lymphoid aggregates. There was no central review of pathology and no training of study site pathologists.
Results
A total of 323 subjects were screened and 320 were enrolled into the study. One hundred sixty subjects were randomised into each arm (CCE or CTC). The median time from CCE to OC was 4.1 weeks (range 1.9–10.5 weeks) and the median time from CTC to OC was 4.5 weeks (range 1.0–13.3 weeks). Among the 320 subjects enrolled and randomised, 15 in the CCE arm and 19 in the CTC arm withdrew prior to completion of the study procedures. Thus, the FAS included 286 subjects (89.4%) who underwent CTC or CCE and OC procedures. In the FAS, there were a total of 21 subjects with at least one major protocol deviation (8 in the CCE arm and 13 in the CTC arm). In addition, four subjects in the CCE arm had incomplete procedures (>2 segments not visualised) and one subject in the CTC arm had an incomplete procedure (CTC procedure stopped after first acquisition due to AE), resulting in a PPAS of 261 subjects, with 133 in the CCE arm and 128 in the CTC arm. The disposition of subjects and allocation to analysis sets is shown in online supplemental figure 2 and descriptions of major protocol deviations are listed in online supplemental table 5. Among the FAS subjects, 57.7% were female and the mean age was 55.7 years (SD ±5.68). Compared with subjects randomised to CTC, subjects randomised to CCE were more likely to be female (63.4% vs 51.8%, p=0.046) and had a higher body mass index (BMI) (32.2 kg/m2 vs 30.2 kg/m2, p=0.018). Of the 286 subjects, 17.1% had undergone a prior CRC screening test, including 14.7% OC, 1.4% sigmoidoscopy, 1% FOBT and 0.3% FIT. Table 1 depicts subject demographics.
Summary of demographic and baseline characteristics
Overall colon cleansing level was adequate in 84.1% of CCE subjects and 90.1% of CTC subjects (p=0.135). Colon cleansing was adequate in 96.5% of OC procedures and was not different between subjects randomised to CCE or CTC. The proportion of subjects achieving a complete examination in the CCE group was 139/145 (95.9%) and in the CTC group it was 140/141 (99.3%) (p=0.121). In patients randomised to CCE, the capsule was excreted naturally within 12 hours in 121 subjects (83.5%). The median procedure time until excretion was 5 hours and 9 min. Colonoscopy was complete to the cecum in all subjects. Mean OC withdrawal time in subjects with no lesions was 11 min (SD 4 min; range 6–24 min).
Diagnostic and accuracy outcomes
The diagnostic yield for polyps ≥6 mm per subject, or polyp detection rate, of OC was 27.6% (72/261). Using the whole colon Rex matching rule, the proportion of subjects in the PPAS with a polyp ≥6 mm confirmed by OC was 31.6% (42/133) for CCE versus 8.6% (11/128) for CTC (posterior probability for non-inferiority and superiority both=0.999). Table 2 shows the diagnostic yields of CCE and CTC for both the PPAS and the FAS. Results for the diagnostic yields of CCE and CTC using the alternative polyp matching algorithms are presented in online supplemental table 6. Regardless of the polyp matching scheme used or the data set analysed (PPAS or FAS), the superior diagnostic yield of CCE compared with CTC was maintained. In addition, the 31.6% diagnostic yield of CCE for polyps ≥6 mm was similar to the 27.6% diagnostic yield observed with OC.
Diagnostic yield of CCE and CTC for detection of polyps ≥6 mm in relation to confirmatory OC results
Table 3 shows the accuracy measures of CCE versus CTC for the detection of a polyp ≥6 mm, assessed in relation to confirmatory OC results, using the whole colon Rex matching rule for both the PPAS and the FAS. Using the PPAS, the sensitivity of CCE for the detection of a polyp ≥6 mm was 79.2% compared with 26.8% with CTC. Results for specificity were comparable between the two groups (96.3% with CCE vs 98.9% with CTC). Based on OC results, the PPV and NPV of CCE for a polyp ≥6 mm was 93.3% and 87.5%, respectively, while the PPV and NPV of CTC were 91.7% and 74.1%, respectively. Online supplemental table 7 shows the accuracy of CCE and CTC for polyps ≥6 mm according to the other three polyp matching schemes.
Accuracy measures of CCE and CTC for the detection of polyps ≥6 mm, assessed in relation to confirmatory OC results
The diagnostic yield for polyps ≥10 mm per subject of OC was 10.3% (27/261). Colon capsule endoscopy also demonstrated non-inferiority to CTC for the diagnostic yield of a polyp ≥10 mm based on the whole colon Rex matching rule with 13.5% (18/133) in the CCE arm compared with 6.3% (8/128) in the CTC arm (table 4). As with polyps ≥6 mm, the observed 13.5% diagnostic yield of CCE for polyps ≥10 mm was similar to that of OC. Using the PPAS and the whole colon Rex matching rule, the sensitivity of CCE for the detection of a polyp ≥10 mm was 85.7% and the specificity was 98.2%, with a PPV and NPV of 90% and 97.3%, respectively (table 5). The corresponding sensitivity and specificity of CTC for the detection of a polyp ≥10 mm was 50% and 99.1%, respectively, with a PPV and NPV of 88.9% and 93.3%. It should be noted that the study was not powered to assess ≥10 mm polyps (table 5).
Diagnostic yield of CCE and CTC for detection of polyps ≥10 mm in relation to confirmatory OC results
Accuracy measures of CCE and CTC in the detection of polyps ≥10 mm, assessed in relation to confirmatory OC results
Using the PPAS, the detection rate of adenomas ≥6 mm by OC was 17.6% and for adenomas ≥10 mm it was 6.1%. Using the whole colon Rex matching rule and analysing according to polyp histology, CCE identified 84.2% (16/19) of adenomas ≥6 mm found on OC while CTC identified 44.4% (12/27) adenomas ≥6 mm found on OC (p=0.007). For adenomas ≥10 mm, CCE identified 100% (3/3) found on OC while CTC identified 61.5% (8/13) identified on OC (p=0.509) (online supplemental table 8). There were 18 serrated polyps ≥6 mm identified (8 in the CCE subjects, 10 in the CTC subjects) and CCE showed numerically higher levels of agreement with OC for the detection of serrated polyps ≥6 mm compared with CTC (50% vs 20%, respectively, p=0.321) (online supplemental table 9). There were 10 serrated polyps ≥10 mm identified and agreement with OC for these lesions was 40.0% with CCE versus 0.0% with CTC (p=0.464). Expanding this analysis to the FAS or other polyp matching methods did not change the findings or comparisons of CCE with CTC. In fact, regardless of analysis set or polyp matching method used, CTC failed to detect any serrated polyps ≥10 mm. In addition, when comparing unadjudicated versus adjudicated and central versus local reader results, accuracy measures were comparable between the groups (data not shown).
Cancers
Colorectal cancer was detected in two subjects. A 5 mm neuroendocrine tumour was detected in the rectum in a subject in the CCE arm, and a 20 mm adenocarcinoma was detected in the ascending colon in a subject in the CTC arm. Both CCE and CTC identified these cancers correctly, including the size and location of the cancers (figure 1).
Colorectal cancers detected. Cancer (5 mm) detected by colon capsule endoscopy (A) and optical colonoscopy (OC) (B). Cancer (20 mm) detected in the ascending colon by CT colonography (C) and OC (D).
Procedure preference
In both groups, more subjects preferred OC over CCE or CTC. Colonoscopy was preferred over CCE by 68.8% due to less time for examination (76.8%), less bowel prep requirement (42.4%) and the ability to collect tissue during the procedure (42.4%). For the 31.2% of subjects who preferred CCE over OC, reasons included no sedation (62.2%), no need to be driven (42.2%) and the procedure being less invasive (28.9%). In the CTC arm, 59.6% preferred OC over CTC with the main reasons being the ability to collect tissue at the same time of the procedure (44.0%), less discomfort (44.0%) and OC’s ability to identify more polyps (35.7%). Reasons for preferring CTC over OC included shorter examination time (73.7%), no sedation (49.1%) and no need to be driven (42.1%).
Adverse events
There were 15 subjects (15/286, 5.2%) with ≥1 AE and a total of 16 AE reported in the study (table 6). Of the 15 subjects, 11 subjects had AE related to the study procedure: CCE (n=2), CTC (n=5) and OC (n=4). The remaining four subjects had AE that were either not related to the procedure (n=3) or were not possible to determine (n=1). Two subjects in the CCE group reported hemorrhoids after the CCE procedure. Five subjects in the CTC group had AE related to the CTC procedure, presyncope (n=2), dyspnoea (n=1), abdominal pain (n=1) and procedure pain (n=1). Four subjects experienced AE related to OC, including post-polypectomy haemorrhage (n=1), haematochezia (n=1), elevated blood pressure (n=1) and rectal haemorrhage (n=1). The three bleeding AEs related to OC were considered serious and all were resolved with endoscopic therapy.
Procedure-related AE
Discussion
The TOPAZ study is the first comparison to date of CCE with CTC in an average risk CRC screening population. While the study was designed to test for non-inferiority of CCE versus CTC for the detection of colorectal neoplasia, it actually demonstrated superiority of CCE compared with CTC for the diagnostic yield for polyps ≥6 mm; 31.6% with CCE compared with only 8.6% with CTC. Sensitivity, arguably the most important test characteristic for a CRC screening test, of CCE for polyps ≥6 mm was 79.2% compared with only 26.8% with CTC. The corresponding specificity, PPV and NPV of CCE for polyps ≥6 mm were 96.3%, 93.3% and 87.5%, respectively. These characteristics were superior to those observed with CTC and they are also superior to results obtained with other non-colonoscopic CRC screening tests that are considered tier 1 or tier 2 tests in the USMSTF guidelines8 or, in the case of FIT, that serve as the basis for numerous European population-based CRC programmes.4 10
For the detection of polyps ≥10 mm, CCE diagnostic yield was non-inferior to CTC and CCE had higher sensitivity and comparable specificity for these lesions compared with CTC. It is important to note that there were relatively few subjects with large polyps in this trial and the trial was not powered to show these differences. In addition, since OC is known to be an imperfect gold standard,29–31 it was essential to investigate if results of each screening test, based on pathology findings, had an impact on accuracy. As with the primary analysis, CCE showed higher or comparable detection rates for adenomas and serrated lesions compared with CTC across all datasets and polyp matching methods.
The whole colon Rex matching rule was the primary polyp matching method used in this study. Prior studies have shown that matching with this method, which maintains a requirement of size matching within 50% regardless of colon segment, yielded a numerically higher, but not statistically significant, sensitivity compared with other polyp matching rules.13 16 We chose this clinically pragmatic method because identification of at least one polyp ≥6 mm with CCE or CTC, regardless of segment, should prompt a timely follow-up OC, similar to recommendations for positive FOBT or FIT according to current US guidelines.19 The recently updated ESGE/ESGAR guidelines recommend an individualised approach including observation or OC with polypectomy for patients with polyps ≥6 mm found on CTC or CCE and prompt progression to OC for three or more sub-centimetre polyps or polyps ≥10 mm.32 In order to test the clinical relevance and robustness of our findings, we also evaluated whole colon and segmental polyp matching rules and found that CCE non-inferiority and/or superiority compared with CTC was consistently maintained for the a priori primary and secondary endpoints of the study.
The superior polyp detection by CCE compared with CTC observed in average risk patients undergoing CRC screening in this study is consistent with previous reports comparing these modalities in patients with incomplete OC or positive FIT.23–27 Although not the direct focus of the current study, the results presented here have important implications for how to best image the non-visualised colon after an incomplete OC. Screening is only one of the settings in which incomplete OC may occur and, while a recent report of CTC use across a group of community hospitals in patients with prior incomplete OC showed very poor sensitivity,33 there is no evidence that the sensitivity and specificity of OC or CCE are different in screening scenarios compared with surveillance or diagnostic examinations.
Our results indicate that CCE warrants equal or higher ranking than CTC in CRC screening guidelines that rank tests. However, neither modality is as effective as OC since both tests had sensitivities for polyp detection that were below OC to a clinically important extent. In addition, OC was widely preferred compared with CCE and CTC, primarily due to the preparation and time required for CCE and CTC and the CRC prevention afforded through polypectomy with OC. Our results can be reasonably interpreted to support a sequential approach to screening in which OC is considered the first-line test for patients who desire or require a screening test that requires bowel preparation and maximises polyp detection by imaging, with CCE and CTC reserved for patients who decline OC or in whom OC carries elevated risk.
In conclusion, the TOPAZ study, the first large, multicentre, randomised, controlled study of CCE compared with CTC in an average risk CRC screening population, demonstrated that CCE was superior to CTC for the identification of colorectal polyps ≥6 mm and non-inferior to CTC for the identification of colorectal polyps ≥10 mm. Both CCE and CTC had lower sensitivities than OC for polyp detection, although CCE had a diagnostic yield that was comparable with OC while CTC did not. Based on these data, CCE should be considered a relevant CRC screening option in an average risk population in line with a ranking comparable with or higher than CTC, as well as a first-line imaging test after incomplete OC.
Data availability statement
Data may be obtained from a third party and are not publicly available. The study protocol, statistical analysis plan and study results are available on ClinicalTrials.gov (https://clinicaltrials.gov/ct2/show/NCT02754661). Individual participant data are not available for sharing.
Ethics statements
Patient consent for publication
Ethics approval
This study was approved by institutional review boards at each centre.
Acknowledgments
TOPAZ study site contributors included Dr Brooks D Cash (Site and Study PI), Dr Brent Acker, Dr Brad A Steffler, Dr Jack A Di Palma, Dr Phillip K Henderson (University of South Alabama, Mobile, AL); Dr Mark R Fleisher (Site PI, Borland Groover Clinic, Jacksonville, FL); Dr Steven Fern (Site PI), Dr Motoyo Yano (Specialists in Gastroenterology, Clinical Research Professionals, Saint Louis, MO); Dr Elizabeth Rajan (Site PI), Dr David H Bruining, Dr Joel G Fletcher, Dr Louis M Wong Kee Song, Dr Stephanie L Hansel, Dr Badr Al Bawardy, Dr Jeff Fidler, Dr Shannon P Sheedy, Dr John Barlow (Mayo Clinic, Rochester, MN); Dr Robyn Haithcock (Site PI), Dr Melvin Saltzman (MultiCare Medical Division Gastroenterology, Clinical Research Professionals Chesterfield, MO); Dr David M Kastenberg (Site PI), Dr Marianne T Ritchie, Dr Sandeep Deshmukh, Dr Leo C Katz (Thomas Jefferson University Hospital, Philadelphia, PA); Dr David Pound (Site PI), Dr Jason Mullinix, Dr Paul K Haynes (Indianapolis Gastroenterology and Hepatology, Indianapolis, IN); Dr Douglas K Rex (Site PI), Dr Toyia James-Stevenson (Indiana University School of Medicine, Indianapolis, IN); Dr Brian Patrick Garvin (Site PI), Dr William Harlan, Dr Sheri Fleeman, Dr Jason Smith (Asheville Gastroenterology Associates, Asheville NC); Dr Ira J Schmelkin (Site PI, Baystate Medical Centre, Springfield, MA); Dr Pramod Malik (Site PI), Dr Jeff McTavish, Dr Venu Koduri (Virginia Gastroenterology Institute, Suffolk, VA); Dr Suzanne K Lewis (Site PI), Dr Suneeta Krishnareddy, Dr Benjamin Lebwohl (New York Presbyterian Hospital/Columbia University Medical Centre, New York NY). Central CCE and CTC readers included Dr Neofytos P Papageorgiou (American Medical Center, Nicosia, Cyprus); Dr Ignacio Fernandez-Urien (NHC, Pamplona, Navarra, Spain); Dr Alain Hyman (Mount Sinai Hospital/School of Medicine, New York, NY).
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors Study concept: BDC, DKR, MRF, DMK. Study design: BDC, DKR, DMK. Acquisition of data: BDC, DKR. Analysis of data: BDC, DKR. Interpretation of data: BDC, DKR. Drafting of the manuscript: BDC, DKR. Critical revision of the manuscript for important intellectual content: BDC, MRF, SF, ER, RH, DMK, DP, NPP, IF-U, DKR. Approval of the final version of the manuscript: BDC, MRF, SF, ER, RH, DMK, DP, NPP, IF-U, IJS, DKR.
Funding This study was sponsored and funded by Medtronic.
Competing interests These authors disclose the following: BDC: Consultant to Medtronic, Salix, Allergan, Takeda; Speakers’ Bureau for Salix, Allergan, Takeda, AlfaSigma, RedHill, QOL. ER: Consultant to Olympus; IP with Medtronic. DMK: Consultant to Medtronic, Ferring, Salix, MotusGI; Research support from Medtronic, MotusGI. DP: Speakers’ Bureau: Gilead Life Sciences, Merck; Advisory Boards: Gilead, Novartis, Intercept. NPP: Consultant to Medtronic. IF-U received honorarium from Medtronic. IJS: Consultant to Medtronic. DKR: Consultant to Olympus Corporation, Boston Scientific, Medtronic, Aries Pharmaceutical, Braintree Laboratories, Lumendi Ltd, Norgine, Endokey, GI Supply; Research support: EndoAid, Olympus Corporation, Medivators, Erbe USA Inc.; Ownership: Satisfai Health.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.