Clinical indications for computed tomographic colonography: European Society of Gastrointestinal Endoscopy (ESGE) and European Society of Gastrointestinal and Abdominal Radiology (ESGAR) Guideline

CTC after incomplete colonoscopy requires a different approach from primary CTC. When endoscopic biopsy has been done, CTC can be performed on the same day as the endoscopic procedure. An ultralow/low dose pre-CTC scan of the abdomen and pelvis before insertion of the rectal tube may rule out the presence of extraluminal gas that would indicate a colonoscopic perforation. In detail, in 262 patients undergoing CTC after incomplete colonoscopy, 2 perforations were detected (0.8%, 95% confidence interval [95%CI] 0.1–2.7) [47]. In the case of endoscopic resection (i.e. polypectomy/mucosectomy), it is prudent to consider an approximately 2-week delay before performing CTC. However, there is little scientific evidence concerning the interval between endoscopic resection and subsequent CTC, thus for each case there should be a clinical discussion between the endoscopist and the radiologist. However, in a recent study on 65 CRC patients with severe luminal narrowing after incomplete colonoscopy with either polypectomy or biopsy sampling, no extraluminal gas was detected at CTC within 24 hours [48]. Other evidence for the safety of radiologic imaging after endoscopic biopsy comes from barium enema studies, both experimental and clinical [49‐52]. These studies concluded that in a nondiseased colon, barium enema could be performed immediately after endoscopic biopsy without any risk. In the case of endoscopic resection, barium enema could be performed without any risk after 6 days.

Accurate preoperative assessment of the whole colon is required to exclude synchronous CRC. In a recent population-based study of 13 683 Dutch patients diagnosed with CRC, 3.9% were diagnosed with synchronous CRC, and in 34% of these cases the two tumors were located in different surgical segments [53]. These data were in line with those from a previous French study [54] and from other series [55]. Failure to detect synchronous cancer can increase morbidity, and one study has shown that intraoperative palpation can miss up to 69% of synchronous malignancies [56, 57]. Thus, preoperative whole-colon assessment is needed.

CTC appears to be an effective and safe choice when obstructing CRC prevents a complete endoscopic assessment or when cecal intubation fails for other reasons. A recent study including 286 CRC cases after failed colonoscopy showed CTC negative predictive values (NPVs) of 100% and 97% for synchronous cancer and advanced neoplasia, respectively, in a preoperative setting [58]. This is in line with a previous systematic review, showing equivalent sensitivity of colonoscopy and CTC for established cancer [22], and in line with findings from similar cohort studies [44, 59‐63].

In the SIGGAR trial no significant difference in the detection rates for large polyps (≥10 mm) and for colorectal cancer was demonstrated between CTC and colonoscopy [13]. Furthermore, the crude pooled sensitivity of CTC for colorectal cancer in the studies of symptomatic patients was 96% (169 out of 176 colorectal cancers detected) [13]. This is compatible with the 96.1% sensitivity of CTC for colorectal cancer that was reported in a meta-analysis [22] that included both screening and symptomatic/high risk patients. When large polyps (≥10 mm) only were considered, per-patient sensitivity of CTC ranged from 82% to 92% in six meta-analyses that included screening, symptomatic, high risk, and FOBT-positive patients [19‐21, 23, 25, 26]. In the studies specifically investigating symptomatic patients, pooled sensitivity for large ≥10-mm lesions (excluding cancers) was 91.4% (53 of 58 patients).

These data suggest that CTC and colonoscopy have similar sensitivity for detecting CRC and large polyps in symptomatic patients. Small polyps (6–9 mm) and diminutive polyps (≤5 mm) are less relevant in symptomatic patients, since they cannot explain the patient’s symptoms. Nonetheless, the ability to opportunistically detect and remove early precursor lesions and perform histopathologic analysis of diagnosed CRC remains a potential advantage of colonoscopy over CTC.

Abdominal symptoms may be due to non-neoplastic colonic conditions, for which both CTC and colonoscopy may be useful. Diverticulosis is more commonly demonstrated at CTC than colonoscopy [13, 65], although the relationship between diverticulosis and symptoms is less clear. Colonoscopy is more sensitive for the detection of colitis and anal pathology [13]; furthermore it offers the possibility of sampling tissue.

CTC is an abdominal CT examination with the ability to detect extracolonic diseases. Although these extracolonic lesions may occasionally explain the symptoms, on the other hand, incidental findings that ultimately prove unimportant may prompt additional tests that are inconvenient, costly, and even harmful. Few studies of extracolonic findings focus specifically on symptomatic patients, in whom there is a higher prevalence of significant abnormality. The two largest series, of screening [66] and symptomatic [11, 13] patients, respectively reported 0.35% and 1.9% rates of extracolonic malignancy. Importantly, in the paired SIGGAR trials, at 3-year follow-up there was no significant difference in rates of extracolonic malignancy between the two arms of each of the trials (CTC vs. barium enema, and CTC vs. colonoscopy), although all arms showed rates significantly above rates expected for the general population. The latter observation may be explained by subsequent use of CT to investigate persistent symptoms in patients randomized to colonoscopy or barium enema, although this remains unproven.

To date, only guaiac FOBT (g-FOBT) and sigmoidoscopy have been shown to reduce CRC mortality, by 16% and 22%–31% respectively [67‐69]. CTC has not been subjected to randomized trials with CRC incidence or mortality as end points. Therefore, the accuracy of CTC is used as a surrogate end point for CTC efficacy in a screening setting.

CTC accuracy in average risk screening populations has been investigated by a recent meta-analysis [24], which estimated per-patient sensitivity at 88% for advanced neoplasia ≥10 mm. One further primary study published after this review, showed similar results [16]. In six screening studies, none of the 12 CRCs present were missed by CTC in average risk individuals [14, 16‐18, 70‐72]. Individuals with a positive family history of CRC or adenomas should be considered to be at high risk [73]. One recent cohort study showed a 89% sensitivity of CTC for advanced neoplasia ≥10 mm in this setting [74].

The efficacy of a screening program not only depends on the diagnostic accuracy of the screening test that is used, but also on participation. This is illustrated by the results of a large population-based randomized screening trial performed in the Netherlands: participation rates for colonoscopy and CTC of 22% and 34%, respectively, were reported, and detection rates for advanced neoplasia of 8.7 and 6.1 persons per 100 participants, respectively [12]. Despite the higher sensitivity of colonoscopy and the fact that CTC participants were only referred to colonoscopy if they had lesions ≥10 mm detected by CTC, the number of individuals per 100 invitees found to have advanced neoplasia was similar for both screening modalities, namely 1.9 (colonoscopy) versus 2.1 (CTC) per 100 invitees [12]. The poorer sensitivity of CTC compared with colonoscopy was countered by its approximately 1.5 times higher participation rate.

In the case of serrated adenomas the diagnostic yield of colonoscopy was 5 times higher than that of CTC. This is of particular importance, since approximately 10%–20% of CRC develops from the serrated pathway [75].

The diagnostic yield of CTC screening per 100 invitees would appear to be significantly higher than the yield of first-round g-FOBT, but similar to the yield of first-round flexible sigmoidoscopy screening (2.2 per 100 invitees) and fecal immunochemical testing (FIT) screening (2.0 per 100 invitees when using a cutoff of 50 ngHb/mL) [76]. One should however bear in mind that FOBT/FIT screening is repeated at 2-year intervals, whereas 5–10-year intervals are usually recommended for CTC and endoscopic screening.

A recent meta-analysis included articles on preferences and differences in burden for both average risk and high risk individuals who had undergone CTC as well as colonoscopy (tandem design) [77]. Amongst the included studies, 3573 patients reported a preference for CTC, 927 showed a preference for colonoscopy, and 1116 showed no difference in preference.

In a Dutch population-based screening trial, almost half of the nonparticipants made an informed decision on participation as they were provided with adequate knowledge of CRC and CRC screening, and showed a positive attitude towards screening, but nevertheless declined participation, which suggested that additional barriers to participation were present [78]. The reasons cited for declining screening by colonoscopy or by CTC were similar overall [79]. However, colonoscopy invitees who declined most often mentioned ‘unpleasantness of the examination’ as their prime reason, while for CTC invitees ‘no time/too much effort’ and ‘lack of symptoms’ were most often cited. The last finding is consistent with the findings of the study of Ho et al., in which 38% did not participate in CTC screening because of procrastination and 12% because they were too busy [80].

As indicated above, most previous screening studies, using a tandem design to compare perceived acceptability and burden of the two techniques, showed a significant preference for CTC, with 46% to 95% of participants preferring CTC for future investigation [17, 81, 82].

A recent Netherlands study performed within the population-based screening trial mentioned above showed that colonoscopy invitees expected the screening procedure and bowel preparation to be more burdensome than did CTC invitees [83]. CTC participants in the Dutch study however found their screening procedure slightly more burdensome than did colonoscopy participants. Colonoscopy participants gave higher burden scores to ingesting the bowel preparation, while CTC participants gave higher burden scores to related bowel movements (i.e. diarrhoea and bowel cramps). Although these differences were statistically significant, they were mostly small and thus the clinical relevance is limited for a clinical population, but more significant for a primary screening population. This is illustrated by the fact that intended participation in a subsequent screening round exceeded 90% for both colonoscopy and CTC.

Extracolonic findings are common at screening CTC and have been reported to occur in from one quarter to more than one half of screenees [92‐97]. The incidence of extracolonic findings increases significantly with age; one study reported extracolonic findings in 55.4% of screenees younger than 65years and in 74% of those 65 years or older [96]. The large majority of extracolonic findings are irrelevant and can be classified as such at CTC.

Work-up for (potentially) important extracolonic findings occurs in approximately 10% of cases [97‐99]. The prevalence of extracolonic findings of moderate or high importance at CTC is commonly reported to be approximately 10%–15% of screenees [94, 95, 98, 99], although higher prevalence is occasionally reported [92, 100]. This difference is partly caused by variation in the definition of moderate and high importance findings. The proportion of findings of high importance is mostly in the order of 2%–5% [95, 97, 99], and includes approximately 0.5% extracolonic cancers, of which renal cell cancer, lung cancer and lymphoma are most prevalent [66, 97, 99, 100], and are usually localized at the time of diagnosis [66]. Further important extracolonic findings include abdominal aortic aneurysms, adrenal masses, and nonmalignant renal masses.

The costs reported for the additional work-up of extracolonic findings vary substantially and are influenced by the definition of a relevant finding needing work-up and by which costs are included. It appears that the average additional cost for extracolonic findings at CTC is of the order of 20–50 USD averaged over all attendees [94‐96, 100, 101]. No studies report costs that might be saved by earlier detection of disease.

Primary CTC and colonoscopy screening have similar yields for advanced neoplasia per invitee. However, the impact of extracolonic findings, both medically and economically, remains unknown. Although radiation exposure is a drawback, this disadvantage seems to be overemphasised especially given the current reduction in radiation exposure with CTC. Probably the most important factor is the question of whether CTC screening is cost-effective, and this is still unanswered. Based on these considerations, CTC cannot at this stage be recommended as the primary test for population CRC screening or in individuals with a positive first-degree family history. However, it may be suggested as a CRC screening test on an individual basis, providing the screenees are adequately informed about test characteristics, benefits, and risks.

Repeated annual or biennial screening for colorectal cancer (CRC) by guaiac-based fecal occult blood testing (FOBT) reduces disease-specific mortality by approximately 15%–18% [102]. Results of similar repeated screening by means of fecal immunochemical testing (FIT) are awaited. It is assumed that the impact on CRC-related mortality will be considerably higher than with FOBT, because of the higher uptake of FIT testing, and the higher sensitivity for advanced colorectal lesions [103]. This is confirmed by modelling studies [104]. This benefit is contingent on confirmation and treatment of underlying cancer or adenoma after a positive result. Colonoscopy combines sensitive diagnosis with therapy by endoscopic resection and is therefore regarded as the preferred test.

Since most screenees testing FOBT/FIT-positive will not have advanced neoplasia, CTC has been investigated as a possible triage test to select patients with lesions only of greater size for colonoscopy or surgery. The sensitivity of CTC for adenomas ≥6 mm was above 85% in six studies [15, 25, 105‐108] and was over 90% for adenomas ≥10 mm, a finding confirmed by a meta-analysis published after our literature search [25]. A modelling study concluded that the use of CTC as an intermediate after positive FOBT/FIT can only be cost-effective if the costs of CTC were ≤43% of the costs of colonoscopy [109]. Furthermore, despite sensitivity exceeding 85%, lesion prevalence is so high that NPV is less than might be expected, ranging from 85% to 95% in the studies included. These factors mean that CTC should not be offered routinely to those testing FOBT/FIT-positive, and colonoscopy is preferable.

Since CTC does have good diagnostic performance, it may be considered for those unwilling to undergo colonoscopy or in whom colonoscopy is unfeasible or incomplete, although screenees should be informed that sensitivity (particularly for smaller adenomas) is slightly inferior to that of colonoscopy. There is some evidence that offering CTC to those who decline colonoscopy increases uptake [110]. CTC is safe, and therefore may be preferable in those with contraindications to colonoscopy or judged particularly high risk, although observational data suggest absolute detection rates may be lower than in healthy screenees who are fit for colonoscopy [111]. Reasons for differences in detection rates are unknown and only speculative at this stage. If the difference is confirmed, and if it is due to suboptimal CTC practice (CTC technique and/or image interpretation), procedures for guaranteeing high quality of CTC exams within organized population screening programs will be necessary.

Patients with resected colorectal cancer are at a 30% risk of recurrence [112, 113] which can be either colonic or extracolonic. Local recurrence is less common for colonic than rectal cancers [112, 114, 115]. Recurrence can occur either at the site of anastomosis or near the site of the primary resection. In contrast, metachronous lesions are colorectal adenomas and cancers that develop subsequently to the index cancer and do not originate from it. Extracolonic recurrent disease comprises distant metastases in the liver, lung, peritoneum, etc. CTC for postoperative surveillance following potentially curative resection of colorectal carcinoma has the potential to combine both colonic and extracolonic examination, and is therefore an alternative to combined optical colonoscopy and contrast-enhanced abdominal CT [116].

By means of a literature review, we identified eight cohort studies investigating contrast-enhanced CTC as a surveillance tool after resection of colorectal cancer [116‐123]. All of these studies demonstrated a high technical feasibility.

The recent ESGE Guideline recommends endoscopic surveillance only for patients with high risk adenomatous lesions (adenomas with villous histology or high grade dysplasia or ≥10 mm in size, or ≥3 adenomas) or serrated lesions (≥10 mm in size, or any degree of cytological dysplasia) [124]. Colonoscopy is considered to be the method of choice for post-polypectomy surveillance, whose primary aim is to diagnose and remove polyps either missed at initial examination or newly developed during the time interval between the index and follow-up examination. However, compliance with colonoscopic surveillance is relatively low, ranging from 52% to 85%, with the highest levels obtained in research settings [125‐128]. Despite weak evidence supporting CTC for surveillance [15], in patients who are unwilling or unable to undergo colonoscopy, CTC is the best alternative because of its high sensitivity and NPV, outperforming barium enema [11, 22, 29].

To provide same-day endoscopy after CTC, the indications and logistics concerning patient selection, timing, patient transportation, availability of endoscopists and endoscopy suites etc. must be pre-planned jointly by radiology and endoscopy units. This modality also requires that CTC findings are reviewed by a radiologist immediately in order to identify patients in whom same-day colonoscopy is needed, and in order to identify the rare but well-recognised perforations that occur during CTC.

It is possible that colorectal lesions reported at CTC may not be detected at colonoscopy, either because they are CTC false positives or colonoscopic false negatives. Clinical consequences include progression of colonoscopic false-negative polyps towards invasive CRC or anxiety due to CTC false-positive findings. In a recent prospective multicenter study of symptomatic patients, the PPV of CTC for large polyps was about 60%, indicating that colonoscopic inability to confirm CTC findings occurs frequently [11]. The sensitivity of colonoscopy for >10-mm polyps is higher [152], and may be presumed to be substantially increased when – as occurs in daily practice – the endoscopist is searching specifically for a CTC finding. Therefore, the possibility of missing large lesions at such colonoscopies may be considered too low to warrant a further endoscopic examination. However, it is well known that colonoscopy is not 100% sensitive even for large lesions that are present at CTC, a phenomenon that has been explained by the existence of colonoscopic “blind spots” [153]. Most post-colonoscopic interval cancers are related to missed rather than new lesions. In contrast to 6–9-mm polyps, the risk of established cancer in larger lesions is relevant [149]. Thus if, after negative colonoscopy findings, confidence in the CTC diagnosis remains high, an early repetition of colonoscopy should be considered, especially if the abnormality appears to be related to flexures or to be on the proximal side of colonic haustra.

ESGE guidelines represent a consensus of best practice based on the available evidence at the time of preparation. They may not apply in all situations and should be interpreted in the light of specific clinical situations and resource availability. Further controlled clinical studies may be needed to clarify aspects of these statements, and revision may be necessary as new data appear. Clinical consideration may justify a course of action at variance to these recommendations. ESGE guidelines are intended to be an educational device to provide information that may assist endoscopists in providing care to patients. They are not rules and should not be construed as establishing a legal standard of care or as encouraging, advocating, requiring, or discouraging any particular treatment.