Eur Radiol Suppl (2005) 15[Suppl 4]:D138–D141 DOI 10.1007/s10406-005-0125-6 © Springer-Verlag 2005
Andrea Laghi
Virtual colonoscopy: clinical application
A. Laghi (✉) University of Rome La Sapienza, Polo Didattico Pontino – Latina, Via Franco Faggiana 34, 04100 Latina, Italy E-mail:
[email protected] Tel.: +39-06-4455602 Fax: +39-06-490243
Abstract Virtual colonoscopy (VC), also known as computed tomography Colonography (CTC), is a non-invasive test for the examination of the colon based on volumetric, thin-collimation CT acquisition of a cleansed and air-distended colon. The technique is easy, less labour-intensive than barium enema and conventional colonoscopy, and is inherently safer. Several studies demonstrate the ability of VC in the detection of colonic neoplastic lesions, not only large carcinomas, but also polyps. Currently, the most widely accepted clinical indication is incomplete or unsuccessful colonoscopy, which may be the result of redundant colon, patient intolerance to the procedure, spasm not resolving even with the use of spasmolytics, obstructing colo-rectal cancer. VC is
Introduction Virtual colonoscopy (VC), also known as computed tomography colonography (CTC), is a noninvasive test for the examination of the colon first proposed by Vining et al. in 1994 [1]. The examination is based on volumetric, thin-collimation CT acquisition of a cleansed and air-distended colon; CT datasets are edited off-line in order to produce multiplanar reconstructions (coronal and sagittal images) as well as three-dimensional (3-D) modelling, including endoscopic-like views. The technique is easy, less labour-intensive than barium enema and conventional colonoscopy, and is inherently safer since no adverse event and only anecdotic
also used to detect cancer in frail and immobile patients to avoid sedation during colonoscopy or the turning required during barium enema. The use of VC in patients under surveillance following colo-rectal cancer surgery is under investigation. Further studies are necessary in order to assess the cost-effectiveness of this approach. For colo-rectal cancer screening, a practical approach is to consider VC as a currently credible alternative screening method and as a reasonable alternative to the other colo-rectal cancer screening tests when a patient is unable or unwilling to undergo conventional colonoscopy. Keywords CT colonography · Colon, polyps · Colon, neoplasms · Colorectal cancer · Screening
procedure-related complications have ever been reported. VC allows full colonic evaluation, if bowel preparation and distension are optimal. From the patient’s perspective, the major advantages of VC include the absence of sedation and water/contrast enema, the short examination time and the potential for same-day colonoscopy when polyps are detected. A brief review of the results will be provided and current clinical indications will be discussed.
Results Several studies demonstrate the ability of VC in the detection of colonic neoplastic lesions, not only large carci-
D139
Fig. 1 a-c Colonic carcinoma of the ascending colon. Axial scan (a) shows circumferential thickening of the colonic wall (arrow) indicative of colonic neoplasm. Tissue transition projection (or
virtual double contrast barium enema) (b) demonstrates the applecore morphology of the lesion as well as the longitudinal extension. Endoluminal 3-D view (c) shows the narrow colonic lumen
Fig. 2 a, b Sessile polyps on axial 2-D (a, arrow) as well as endoluminal 3-D (b) images
nomas (Fig. 1), but also polyps [2–4] (Fig. 2). A good overview of the results of VC is offered by a recent meta-analysis [5] evaluating 33 publications from 454 potentially eligible papers. Inclusion criteria included: prospective studies; full bowel preparation; colonoscopy or surgery as the gold standard; at least single-slice spiral CT scanner; supine and prone positioning; air or CO2 insufflation; collimation < 5mm; use of 2-D and 3-D views. Per-patient sensitivity for polyps larger than 9 mm was 85% (CI, 0.79–0.91), for polyps between 6 and 9 mm it was 70% (CI, 0.55–0.84) and for polyps smaller than 6 mm it was 48% (CI, 0.25–0.70). Specificity was 97% (CI, 0.96–0.97), 93% (CI, 0.91–0.95) and 91% (CI, 0.89–0.95), for polyps larger than 9 mm, between 6 and 9 mm and smaller than 6 mm, respectively. Another meta-analysis [6], considering 24 studies from 1398 potentially eligible, with 4181 patients and a prevalence of abnormality ranging between 14.7 and 72.2%, showed similar results for polyps. The authors
reported also a high sensitivity for cancer (95.9%; CI, 0.91–0.98). These data confirm that the smaller the polyp size, the lower the sensitivity of VC; some authors for this reason do not even consider polyps smaller than 5 mm. Variability of the results among different studies is probably due to different examination techniques (in particular in terms of slice collimation and image reconstruction thickness) and readers’ experience, as demonstrated by wide inter-observer and intra-observer variability for inexperienced readers [7]. As a matter of discussion, most of the results have been obtained in so-called polyp-enriched populations, that is in populations with a high prevalence of the disease, up to 72.2% [6]. Thus, the results cannot be translated into a screening population, where the prevalence of the disease is much lower. Only few studies, at the moment, were performed in a true screening population, with controversial results.
D140
The first paper based on a real screening population (asymptomatic subjects with no increased risk of cancer development) was published in 1999 by Rex et al. [8]. This study shows a very low sensitivity in detection of clinically significant polyps (> 1 cm), leading the authors to conclude that “VC…is not adequate as a colo-rectal cancer screening test”. However, some limitations should be recognized: first, the paper reflects a very preliminary experience with VC (1995-96), although it was published only in 1999, with most of the false negative examinations to be attributed to perceptual errors (demonstrating reader inexperience and a learning curve of the technique); second, poor colon preparation was another major cause of false positive and false negative studies; third, poor colonic distension was recognized as a major limitation for a correct image interpretation. In another paper, Yee et al. [9] reported the results of a series of more than 300 patients who underwent both VC and conventional colonoscopy. Among these 300 patients, 96 were asymptomatic, thus representing a typical screening population. Data analysis showed similar results among the two patient populations (symptomatic patients and asymptomatic subjects), demonstrating no statistically significant differences. In particular, overall polyp sensitivity, independently of size and histology (hyperplastic or adenomatous polyps), was 69.0 and 69.7%, respectively, in symptomatic and asymptomatic subjects. Similar results were demonstrated for per-patient sensitivity (88.0 and 82.0%, respectively, in symptomatic and asymptomatic subjects) and specificity (90.9 and 67%, respectively, in symptomatic and asymptomatic subjects). If considering only adenomatous polyps, which are the target lesions in the case of screening, no statistically significant difference was demonstrated between symptomatic and asymptomatic subjects. In a series of 703 asymptomatic subjects at higherthan-average risk for colo-rectal cancer who underwent faecal occult blood test (FOBT), flexible sigmoidoscopy, colonoscopy and double-read VC, Johnson et al. [10] showed disappointing results for VC, with a sensitivity for lesions 10 mm or larger ranging between 34 and 73%. Reasons for these results are unclear, although study technique and reader fatigue might be two possible explanations; in fact, due to the low prevalence of adenomas larger than 1 cm (5%) in the series, a single observer had to read over 13,000 images to find one single 1-cm polyp! A landmark study by Pickhardt et al. [11] served to further increase interest in VC, re-establishing the view that it is now ready for widespread clinical implementation as an optional screening tool. The authors evaluated a total of 1,233 asymptomatic adults (mean age 57.8 years) who underwent same-day virtual and optical colonoscopy with segmental unblinding. The latter means that colonoscopist is blinded to VC results, but once he has cleared any colonic segment a nurse would read the
VC report for that segment. If any lesion is detected at VC the scope is re-introduced and the segment re-evaluated. If no lesion is detected, the finding is a false positive of VC; if a lesion is found, this is a false negative of colonoscopy. Using this method, the sensitivity of the two techniques was compared. The sensitivity of virtual colonoscopy for adenomatous polyps was 93.8% for polyps at least 10 mm in diameter, 93.9% for polyps at least 8 mm in diameter, and 88.7% for polyps at least 6 mm in diameter. The sensitivity of optical colonoscopy for adenomatous polyps was 87.5, 91.5 and 92.3% for the three sizes of polyps, respectively. The specificity of virtual colonoscopy for adenomatous polyps was 96.0% for polyps at least 10 mm in diameter, 92.2% for polyps at least 8 mm in diameter and 79.6% for polyps at least 6 mm in diameter.
Current clinical indications Currently, the most widely accepted clinical indication for VC is incomplete or unsuccessful colonoscopy, which may be the result of redundant colon, patient intolerance to the procedure, spasm not resolving even with the use of spasmolytics or obstructing colo-rectal cancer. The major advantage of VC, also compared with barium enema, is that this exam can be performed on the same day as the colonoscopy without additional bowel preparation. VC can complete the examination in most of the cases, being also able to provide the cause of endoscopic failure. In the case of occlusive carcinoma, VC can detect synchronous carcinomas, occurring in 4.9% of the cases; if i.v. injection of iodinated contrast medium is used, a complete staging of the patient can be performed [12, 13]. VC is also used to detect cancer in frail and immobile patients to avoid sedation during colonoscopy or the turning required during barium enema [14]. The use of VC in patients under surveillance following colo-rectal cancer surgery was also investigated [15, 16]. Further studies are necessary in order to assess the cost-effectiveness of this approach. Current uses of VC generally do not include the screening of asymptomatic persons, as also suggested by the American Cancer Society [17] and the American Gastroenterological Association [18], both of which decided that it should not yet be used for colo-rectal cancer screening, because data on true screening populations are missing. Nevertheless, VC is, in any case, the secondbest imaging modality for colonic evaluation, highly superior to double-contrast barium enema for both cancer and polyp detection [19] For this reason, a practical approach is to consider VC as a currently credible alternative screening method and as a reasonable alternative to the other colorectal cancer screening tests when a patient is unable or unwilling to undergo conventional colonoscopy [20].
D141
References 1. Vining DJ, Gelfand DW, Bechtold RE, et al. (1994) Technical feasibility of colon imaging with helical CT and virtual reality. AJR 62[Suppl]:104 (abstract) 2. Fenlon HM, Nunes DP, Schroy PC, et al. (1999) A comparison of virtual and conventional colonoscopy for the detection of colorectal polyps. N Engl J Med 341:1496–1503 3. Cotton PB, Durkalski VL, Pineau BC, et al. (2004) Computed tomographic colonography (virtual colonoscopy). A multicenter comparison with standard colonoscopy for detection of colorectal neoplasia. JAMA 291:1713–1719 4. Rockey DC, Paulson E, Niedzwiecki D, et al. (2005) Analysis of air contrast barium enema, computed tomography colonography, and colonoscopy: prospective comparison. Lancet 365:305–311 5. Mulhall BP, Veerappan GR, Jackson JL (2005) Meta-analysis: computed tomography colonography. Ann Intern Med 142:635–650 6. Halligan S, Bartram C, Taylor S, Mallett S, Atkin W (2004) CT colonography for detection of colorectal neoplasia: meta-analysis. Eur Radiol 14[Suppl 6]:N15 7. Johnson CD, Toledano AY, Herman BA, et al. (2003) Computerized tomographic colonography: performance evaluation in a retrospective multicenter setting. Gastroenterology 125:688–695
8. Rex DK, Vining D, Kopecky KK (1999) An initial experience with screening for colon polyps using spiral CT with and without CT colography. Gastrointestinal Endosc 50:309–313 9. Yee J, Akerkar GA, Hung RK, Steinauer-Gebauer AM, Wall SD, McQuaid KR (2001) Colorectal neoplasia: performance characteristics of CT colonography for detection in 300 patients. Radiology 219:685–692 10. Johnson CD, Harmsen WS, Wilson LA, et al. (2003) Prospective blinded evaluation of computed tomographic colonography for screen detection of colorectal polyps. Gastroenterology 125:311–319 11. Pickhardt PJ, Choi JR, Hwang I, et al. (2003) Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 349:2191–2200 12. Morrin MM, Farrell RJ, Raptopoulos V, et al. (2000) Role of virtual computed tomographic colonography in patients with colorectal cancers and obstructing colorectal lesions. Dis Colon Rectum 43:303–311 13. Fenlon HM, McAneny DB, Nunes DP, et al. (1999) Occlusive colon carcinoma: virtual colonoscopy in the preoperative evaluation of the proximal colon. Radiology 210:423–428 14. Burling D, Halligan S, Taylor SA, Usiskin S, Bartram CI (2004) CT colonography practice in the United Kingdom: a national survey. Clin Rad 59:39–43
15. Laghi A, Iannaccone R, Bria E, et al. (2003) Contrast-enhanced computed tomographic colonography in the follow-up of colorectal cancer patients: a feasibility study. Eur Radiol 13:883–889 16. Fletcher JG, Johnson CD, Krueger WR, et al. (2002) Contrastenhanced CT colonography in recurrent colorectal carcinoma: feasibility of simultaneous evaluation for metastatic disease, local recurrence, and metachronous neoplasia in colorectal carcinoma. AJR 178:283–290 17. Smith RA, Cokkinides V, Eyre HJ (2003) American Cancer Society Guidelines for the early detection of cancer, 2003. CA Cancer J Clin 53:27–43 18. Winawer S, Fletcher R, Rex D, et al. (2003) Colorectal cancer screening and surveillance: clinical guidelines and rationale – update based on new evidence. Gastroenterology 124:544–560 19. Johnson CD, MacCarty RL, Welch TJ, et al. (2004) Comparison of the relative sensitivity of CT colonography and double-contrast barium enema for screen detection of colorectal polyps. Clin Gastroenterol Hepatol 2:314–321 20. Barish MA, Soto JA, Ferrucci JT (2005) Consensus on current clinical practice of virtual colonoscopy. AJR 184:786–792