SIMON SCHLACHTER and PATRICK MACCARTHY
The ability of optical coherence tomography (OCT) to perform non-contact, high-resolution, cross-sectional imaging in real time enabled the technique to revolutionize the field of ophthalmology.1,2 Although relatively new, OCT has become the gold standard for the diagnosis and treatment of debilitating eye diseases. It is now also being used to image coronary arteries, lungs, skin, bladder, and cervix in humans, and promises similar impact on those disciplines.
Esophageal disease and OCT
Since its development, OCT has been recognized as a potential solution to many clinical challenges in the esophagus—including the visual delineation of mucosa, muscularis mucosa, and submucosa of the esophagus—that other imaging modalities cannot address.3 OCT meets an imaging need, from a resolution and depth perspective, that is not met by ultrasound, confocal microscopy, or white-light endoscopy (see Fig. 1).
Of the two main types of esophageal cancer, squamous cell carcinoma is linked to the prevalence of smoking, which is declining in western countries (but climbing in Asia). Adenocarcinoma has grown more than sevenfold in the U.S. over the past few decades, however, and the 5-year survival rate is less than 20%.4,5 When a person is suspected of adenocarcinoma, the doctor makes an assessment using standard endoscopy to visualize the surface of the tissue, and then biopsies areas of disease that can be visualized as well as other random areas. Endoscopy involves passing a flexible imaging instrument through the mouth and down the esophagus to the point where it meets the stomach. A biopsy device is then passed through the endoscope to remove small pieces of mucosa. These tissue samples are examined microscopically (histopathology) to determine whether they are normal, metaplastic, dysplastic (precancerous), or cancerous. In the U.S., patients with confirmed precancerous lesions in the esophagus are treated through the use of ablation methods such as radiofrequency ablation (RFA), photodynamic therapy (PDT), or mucosal resection. Post-treatment, the patient is followed with ongoing routine endoscopy and random biopsy.
This paradigm is limited by a number of factors. First, endoscopy is performed with white-light imaging, which can assess only the surface of the esophagus. It is not able to view potential areas of concern beneath the surface of the mucosa. That is one reason for taking biopsies. However, random biopsies assess less than 3% of the esophageal surface, and are subject to sampling error. Because esophageal diseases do not generally exist in clearly defined regions, it is hard to know whether an area of disease was just missed by a random biopsy. Beyond these concerns, endoscopy with multiple biopsies is costly—not to mention invasive.6
OCT has been studied extensively in humans for imaging various diseases of the esophagus, including squamous cell carcinoma and Barrett's esophagus, a precursor to adenocarcinoma. However, technical limitations, related primarily to imaging speed, have prevented OCT from being able to image a large area of the esophagus quickly enough to make it viable as a clinical imaging tool.
To meet this clinical challenge, the Wellman Center for Photomedicine developed an advanced form of OCT that can image a large area of the esophagus—6 cm long and 3 mm deep into the mucosa, with 7 μm resolution—in approximately 90 s.7 NinePoint Medical (Cambridge, MA) has licensed the technology and developed it into a commercial-ready imaging platform. This platform, the NvisionVLE Imaging System, is the first OCT system commercially available to address esophageal imaging. It is being launched at Digestive Diseases Week (DDW; May 18–21, Orlando, FL), the world's largest educational forum for gastroenterology.
Volumetric laser endomicroscopy
The NvisionVLE Imaging System from NinePoint Medical has been cleared by the FDA for use in the U.S. as an imaging tool in the evaluation of human tissue microstructure, including esophageal tissue microstructure, by providing two-dimensional, cross-sectional, real-time depth visualization. The safety and efficacy of this device for diagnostic analysis (i.e., differentiating normal vs. specific abnormalities) in any tissue microstructure or specific disease has not been evaluated.
The system consists of the NvisionVLE Imaging Console and the NvisionVLE Optical Probe (see Fig. 2). The former houses the computer, OCT system, and fiber-optic rotary junction, and supports two touch-screen monitors. The optical probe connects to the imaging console with single-handed operation and automated coupling of the optical and mechanical connections. The probe consists of an outer balloon catheter and an inner optical fiber that carries near-infrared (NIR) light to the distal optics, where it is focused into the esophageal wall. Surrounding the optical fiber is a drive shaft that transmits torque from the console to the probe optics, creating the helical scan that images the 6 cm segment of the esophagus.
Inside the imaging console, light reflected back from the tissue (via the optical probe) is combined (interfered) with light from a reference reflector. The beat frequency between these two interfering beams is directly proportional to the depth of the reflection from the tissue. Therefore, taking the Fourier transform of the interferogram yields an axial plot of the amount of backscatter from the tissue. These axial plots, or A-lines, are acquired 4,096 times per rotation of the probe. The probe undergoes 1,200 rotations over the course of its 6 cm pullback for a total of just under 5 million A-lines in approximately 90 s.
Once the data is acquired, the next challenge is to display the 5 GB datasets with an intuitive interface that allows the physician to extract the most clinically useful information. The NvisionVLE software is controlled via two touch-screens: one 30 in. main display and a 12 in. support screen. The interface shows transverse and longitudinal sections through the datasets (see Fig. 4), with the ability to export videos of the pullback and zoom regions.
Clinical experience
The volumetric laser endomicroscopy (VLE) procedure is quick and uses techniques that are familiar to gastroenterologists. With the patient sedated, the clinician introduces the NvisionVLE Optical Probe through the instrument channel of a standard endoscope. A balloon at the distal end of the probe is then inflated while positioned across the gastroesophageal junction. The balloon helps to stabilize and center the rotating optics within the lumen of the esophagus. Imaging begins in the stomach and then continues 6 cm in the proximal direction back up into the esophagus. The VLE procedure adds approximately 7 min to a standard endoscopy exam.
The images produced by the NvisionVLE Imaging System reveal light-scattering physiological structures up to 3 mm deep in tissue. For esophageal applications, the imaging depth penetrates several layers of the healthy esophagus, including the squamous epithelium, lamina propria, muscularis mucosa, and submucosa. Often, the muscular outer layers of the esophagus, called the muscularis propria, and the adventitia securing the organ in the mediastinum are visible. Each of these layers is approximately 300 to 800 μm thick. The different makeup of the layers (e.g., squamous cells in the epithelium, collagen in the lamina propria, and smooth muscle cells in the muscularis mucosa) produce distinct differences in the scattering from each layer, which is detectable using OCT. Within these layers, it is possible to observe further scattering structures, such as blood vessels, ducts, and esophageal submucosal glands.
These physiological structures give healthy tissue a distinctive signature on NvisionVLE images. Abnormal tissue tends to have a less organized, non-layered architecture, and is often associated with irregular dilated glands and ducts in the submucosa. And abnormally large nuclei close to the surface of the tissue tend to scatter back a large amount of light, resulting in a high OCT signal at the surface of the tissue that decays rapidly with depth.
A recent study published by investigators at Massachusetts General Hospital has revealed that physicians can be quickly trained on how to use OCT images to identify esophageal tissue features.8 These promising results suggest that clinicians can potentially use previously validated criteria for image review to identify areas of suspicion.
Future
The balloon-based NvisionVLE system is currently being used in an IRB-approved clinical trial to evaluate its ability to assist the physician in visualizing the esophagus. However, with the increasing incidence of adenocarcinoma of the esophagus, a large population of patients would benefit from screening for disease without resorting to costs and complexities of endoscopy, which generally requires anesthesia. For this purpose, a capsule-based VLE probe has been developed by researchers at Massachusetts General Hospital and initial results from an IRB-approved clinical trial have been published.9
In the capsule design, the VLE optics are contained within a medical-grade plastic capsule measuring 12.8 mm in diameter and 25 mm long. The capsule is connected to the console via a 2.5 m tether that houses the optical fiber and drive shaft. The string-like tether is highly flexible and, according to the study, patients found swallowing the capsule far more comfortable than undergoing an upper endoscopy.
Once swallowed, the esophageal wall grasps the capsule and peristaltic motion moves it along the esophagus. Good contact between the esophageal tissue and the capsule allows the entire esophagus to be imaged as the capsule moves towards the stomach. The tether is then used to gently retrieve the capsule, and scanning continues on the way back. The researchers found that 94% of the frames captured in this way were of "high quality," suggesting that the microscopic imaging of the entire esophagus could one day be available at every local doctor's office.
Beyond the esophagus, there are a number of other areas in the body that may be shown to benefit from advanced OCT, including the lungs, the biliary and pancreatic system, and the bladder. Perhaps one day, OCT will be as common a term, and a procedure, as MRI or x-ray.
REFERENCES
1. D. Huang et al., Science, 254, 5035, 1178–1181 (1991).
2. E. A. Swanson, "Estimates of Ophthalmic OCT Market Size and the Dramatic Reduction in Reimbursement Payments," OCT News, http://bit.ly/17X2287 (Dec. 1, 2012).
3. A. Das et al., Gastrointest. Endosc., 54, 2, 219–224 (2001).
4. H. Pohl, B. Sirovich, and H. G. Welch, Cancer Epidemiol. Biomarkers Prev., 19, 6, 1468–1470 (2010).
5. http://www.cancer.org/acs/groups/cid/documents/webcontent/003098-pdf.pdf.
6. P. T. Chandrasoma and T. R. DeMeester, GERD: Reflux to Esophageal Adenocarcinoma, Academic Press, Waltham, MA (2010).
7. M. J. Suter et al., Gastrointest. Endosc., 68, 4, 745–753 (2008).
8. J. Sauk et al., Dig. Dis. Sci., doi.10.1007/s10620-013-2625-x (March 19, 2013).
9. M. J. Gora et al., Nat. Med., 19, 2, 238–240 (2013).
Simon Schlachter, Ph.D., is clinical systems engineer and Patrick MacCarthy is vice president of marketing at NinePoint Medical, Cambridge, MA; www.ninepointmedical.com. Contact Mr. MacCarthy at [email protected].