Volume 4, Issue 1 , Pages 36-37, January 2006
Detecting Dysplasia With Optical Coherence Tomography
Article Outline
In this issue of Clinical Gastroenterology and Hepatology, Evans et al1 present evidence that optical coherence tomography (OCT) can detect high-grade dysplasia or intramucosal carcinoma in patients with Barrett’s esophagus. OCT is one of several optical methods that have recently been used to detect dysplasia in Barrett’s esophagus,2, 3 colonic polyps, and chronic ulcerative colitis. These optical methods share a common goal of being capable to detect and confirm neoplastic change in the background of a large surface area at risk for cancer. The major theoretical advantages of optical methods, as compared with random biopsy, are the potential to reduce sampling error, reduce intraobserver disagreement, and reduce the cost and risk of biopsy by eliminating unnecessary sampling of non-neoplastic tissue. These are regarded as the “holy grail” in early detection of cancer in these conditions. Although this potential has yet to be reached, the article by Evans et al makes an important contribution to the quest.
In this study, Evans et al1 evaluated 55 patients with Barrett’s esophagus including 242 biopsy sites. Each site was evaluated by an OCT probe immediately before biopsy, and the OCT “image” was compared with a consensus histologic diagnosis. OCT produces images based on light scattered back to the detector by dense objects in the tissue, particularly nuclei; thus sites with increased nuclear number and density are detectable. In dysplastic sites, in which enlarged and crowded nuclei are present closer to the surface (poor surface maturation), OCT images appear “darker” at the surface relative to below the surface. The opposite is seen in nondysplastic sites. In addition, OCT provides images of sufficient resolution (approximately 10 μm) to allow visualization of gland architecture so that regular and irregular gland architecture can be seen. They developed a scoring system based on these 2 parameters and classified each site according to the dysplastic or nondysplastic appearance.
Of the 242 sites appraised, 65 had images unsuitable for analysis; thus 177 were evaluated. The OCT parameters correlated well to the degree of dysplasia. By using a cutoff score from the maturation and gland irregularity index, OCT was able to detect high-grade dysplasia or intramucosal carcinoma with a sensitivity of 83% and specificity of 75%.
Before evaluating this article, it is important to describe the basic methods and applications of OCT in gastrointestinal disease. Although OCT has been around for many years, primarily in the field of ophthalmology, it has only recently been applied to gastrointestinal disease. Endoscopic OCT is a method that provides 2-dimensional cross-sectional anatomic images corresponding to the 4 layers of the gastrointestinal tract. Although the images are similar in orientation to endoscopic ultrasound, by using light instead of ultrasound waves, the resolution of OCT is increased to nearly 10-fold greater than high-frequency endoscopic ultrasound and approaches that of light microscopy. Like ultrasound, OCT uses an energy source to deliver a signal, light in this case, to an organ or tissue and a detector to collect the signal if or when it returns. Because light travels exponentially faster than ultrasound waves, most detectors are unable to determine precisely the time of flight of a photon. OCT overcomes this technical limitation by delivering a light signal via 2 separate pathways. One light beam is delivered to the tissue, and an identical light beam is delivered to a reference mirror at a known distance away from the detector.
Light reflected back from the tissue and the light from the reference mirror both return to a detector. However, a device called an interferometer only allows the 2 beams of light that return at exactly the same time to pass through, thus creating an image slice. Light from closer or more distant regions of tissue is discarded. By varying the distance of the reference mirror, structures of varying depth in the tissue can be imaged, and a 2-dimensional image can be produced by stacking the slices together.
OCT is typically performed with near-infrared light because tissue is relatively transparent at these frequencies. OCT uses catheters passed through the accessory channel of standard endoscopes. Radial scanning and linear scanning catheters have been described. Unlike endoscopic ultrasound, OCT can be performed through air, so tissue contact or coupling is not required. Scanning depth is limited to 1–2 mm because of scattering of light by tissue. Most of the systems described achieve a resolution of about 10–20 μm, which is sufficient for visualizing mucosal glands, crypts, and villi but not cellular features such as nuclear dysplasia.
The first clinical publications of OCT reported on its use to image the coronary vasculature and retina.4 In the gastrointestinal tract, catheter-based OCT systems were used in the endoscopic examination of the esophagus, the pancreaticobiliary tract, and the colon.
Barrett’s esophagus has been the focus of intense OCT research in the gastrointestinal tract. A large-scale study formulated objective OCT criteria that are highly sensitive and specific for the diagnosis of specialized intestinal metaplasia.5 Loss of regular crypt-and-pit architecture of the esophageal mucosa was identified as the main OCT feature in the diagnosis of Barrett’s epithelium. Two features seem to be characteristic for dysplasia and cancer: focal (dark) areas of decreased light scattering and focal loss of mucosal structure and organization. Dysplasia was identified with an accuracy of 70% and a negative predictive value of 91%. Other preliminary data also suggest that OCT signals contain information that can be used to identify dysplasia within Barrett’s epithelium with a high degree of accuracy.6
OCT of the biliary system is feasible in patients with biliary pathology.7, 8 Interpretable images are obtainable, but clinical use needs further assessment. Because current OCT probes and processors do not yet provide optimal resolution, further generations of equipment with improved image quality are required.
OCT can visualize the mucosa, muscularis mucosa, and submucosa of the colon.9, 10 Similar to the OCT data on esophageal cancer, the OCT image of colon cancer revealed complete loss of the normal tissue morphology.10 OCT has also the ability to differentiate adenoma from non-adenomatous polyps and normal colon mucosa.11 Other potential applications might be the surveillance of patients with longstanding inflammatory bowel disease by identifying mucosal versus transmural inflammatory change.12, 13 (I am not aware of any published work on microscopic colitis, and I did not find any on Medline. I am less enthusiastic about this as a potential target of OCT because there are fewer gross/microscopic structural changes that could be identified at this level of resolution. I think it would require fluorescence [to detect collagen].)
The article by Evans et al1 provides the first large-scale, prospective, blinded evaluation of OCT for the detection of high-grade dysplasia and intramucosal carcinoma in patients with Barrett’s esophagus. The strengths of the article lie in the rigorous design (double blinding, confirmation of pathology by consensus of 2 expert gastrointestinal pathologists, use of established criteria for OCT diagnosis). There are several important limitations, most of which are discussed by the authors. The accuracy of this OCT system, even under the best circumstances, is marginal for clinical applications. A sensitivity of 83% is likely insufficient for routine use. The authors could lower their threshold (eg, to a score of ≥1 on the dysplasia index). This should increase sensitivity at the expense of specificity, a tradeoff that is probably acceptable because all abnormal sites would be confirmed by biopsy. The accuracy of this system is also likely to be somewhat lower when applied to the general population. Sixty-five of the 224 images were excluded because of poor quality. In addition, the diagnostic algorithm was developed and applied to the same data set. This overestimates true accuracy. An ideal study should develop the algorithm in one group and then test the accuracy in a second independent group.
The algorithm for diagnosing dysplasia was performed offline with static images, carefully reviewed outside the time pressures of endoscopy. The interpretation is subjective and is likely subject to interobserver variation. Although appropriate for initial “proof-of-principle” trials, the algorithms should ideally be automated, so they can render an objective diagnosis in real time, thus allowing image-directed biopsy.
Overall, this article is an important first step for OCT and provides proof of principle that it can detect changes associated with dysplasia in Barrett’s esophagus. This, as well as most of the other optical technologies for Barrett’s esophagus, are still in development and warrant further research and optimism that we might someday have a highly accurate, simple to use optical system for guiding biopsy. With the careful study and development evidenced in this article, these goals will be less elusive than the pot of gold at the end of the rainbow.
References
- Optical coherence tomography to identify intramucosal carcinoma and high-grade dysplasia in Barrett’s esophagus . Clin Gastroenterol Hepatol . 2006;4:38–43
- Fluorescence, reflectance, and light-scattering spectroscopy for evaluating dysplasia in patients with Barrett’s esophagus . Gastroenterology . 2001;120:1620–1629
- Endoscopic detection of dysplasia in patients with Barrett’s esophagus using light-scattering spectroscopy . Gastroenterology . 2000;119:677–682
- Imaging of coronary artery microstructure (in vitro) with optical coherence tomography . Am J Cardiol . 1996;77:92–93
- Diagnosis of specialized intestinal metaplasia by optical coherence tomography . Gastroenterology . 2001;120:7–12
- . Diagnosis of Barrett’s esophagus using optical coherence tomography . Gastrointest Endosc Clin North Am . 2004;14:573–588
- First in vivo optical coherence tomography in the human bile duct . Endoscopy . 2001;33:1018–1021
- Optical coherence tomography of the biliary tree during ERCP . Gastrointest Endosc . 2002;55:84–88
- High-resolution endoscopic imaging of the GI tract using optical coherence tomography . Gastrointest Endosc . 2000;51(Pt 1):474–479
- In vivo endoscopic optical coherence tomography of the human gastrointestinal tract (toward optical biopsy) . Endoscopy . 2000;32:743–749
- Criteria for the diagnosis of dysplasia by endoscopic optical coherence tomography . Gastrointest Endosc . 2003;58:196–202
- In vivo colonoscopic optical coherence tomography for transmural inflammation in inflammatory bowel disease . Clin Gastroenterol Hepatol . 2004;2:1080–1087
- Ex vivo histology-correlated optical coherence tomography in the detection of transmural inflammation in Crohn’s disease . Clinical Gastroenterol Hepatol . 2004;2:754–760
Dr Wallace has received honoraria from Olympus for endobronchial ultrasound. Olympus also makes OCT, but not the ones in this trial.
PII: S1542-3565(05)00997-3
doi:10.1016/j.cgh.2005.10.005
© 2006 American Gastroenterological Association. Published by Elsevier Inc. All rights reserved.
Volume 4, Issue 1 , Pages 36-37, January 2006
