Image enhanced endoscopies, such as narrow band imaging (NBI) and autofluorescence imaging (AFI), were recently developed and clinically applied for the diagnosis of gastrointestinal diseases. From the results of recent clinical studies, these novel technologies appear to be useful for detecting, as well as evaluating, gastrointestinal disorders. While NBI can emboss the vessel structure, AFI can capture fluorescence emitted from intestinal tissues and also describe the area in which the fluorescence intensity is reduced as a magenta color on the monitor. This device produces an excitation light source of 390-470 nm and delivers it to the tissue surface through an endoscope. The low light level autofluorescence emitted from the tissue and white light reflection are detected and amplified by two high sensitivity CCDs. The fluorescence image is then transformed into a green hue, while the reflection image is transformed into red and blue hues. Thereafter, the images are displayed on a monitor as color images in real-time (Figure 1) (Nakashima A, et al. 2001).
AFI is characterized as a tool for simply describing the abnormal findings as color changes, whereas the diagnosis of conventional endoscopy, NBI and other technologies are based on the empirical judgment of morphological features. Moreover, the characteristics that AFI provides are based on the strength of fluorescence, so the quantification of the fluorescence intensity (proposed as the “F index”) using an analytical image software program can provide an objective diagnosis for intestinal disorders (Figure 2).
This chapter focuses on the efficacy of AFI for the detection and differentiation of colon neoplasms, as well as the evaluation and surveillance of ulcerative colitis, and reviews the likely future progression of this technology in the diagnosis of intestinal disorders.
2. Endoscopic features of colonic lesions obtained by AFI
The typical images of intestinal disorders including neoplasms, inflammatory bowel disease, other types of colitis, Behcet disease, amyloidosis and other diseases are described in this section.
2.1. Colon cancer and adenoma
Almost all colon cancers and adenomas are described as magenta in AFI (Figure 3). The invasion depth of the cancer does not appear to influence the intensity of the fluorescence captured by AFI (Figure 4). The margin of the tumor is clear in AFI, even for flat and depressed types of tumors (Figure 5). Interestingly, AFI can detect only the tumor area, but not the reactive changes surrounding the tumor (Figure 6). This suggests that AFI detects some histological component of the tumor. The intensity of the fluorescence captured by AFI appears to be inversely proportional to the dysplastic grade of adenoma (Moriichi, et al. in submission).
2.2. Carcinoid tumors
Carcinoid tumors form in the neuroendocrine system. This type of tumor is thought to initiate from the deep portion of the intestinal epithelia (Modlin IM, et al. 2003). The tumor is therefore likely to be a submucosal tumor in the early stage. A typical finding of carcinoid tumors by white light endoscopy is a yellowish elevation with a smooth surface. AFI detects the tumor as a slight magenta area, but not strong magenta as is seen for typical colon cancer (Figure 7).
The intestinal tract is a major organ presenting with extranodal lymphoma lesions (Groves FD, et al. 2000)(Gurney KA, et al. 2002)(Morton LM, et al. 2006). The most frequently observed types of lymphoma are non-Hodgkin’s, B cell lymphomas including diffuse large B cell lymphoma, follicular lymphoma and mucosa-associated lymphoid tissue (MALT) lymphoma. Diffuse large B cell lymphoma frequently forms a submucosal tumor which is detected as a magenta area by AFI (Figure 8). Conversely, Follicular, MALT and T cell lymphomas are sometimes detected as diffusely spread magenta spots or areas by AFI (Figures 9, 10 and 11).
2.4. Ulcerative colitis
Ulcerative colitis is a chronic refractory colitis. While clinical symptoms are essential to assess the activity of ulcerative colitis, endoscopic assessment helps to predict the relapse or evaluate the grade of mucosal healing (Fujiya M, et al. 2002). Typical endoscopic features of ulcerative colitis are an absence of vessel permeation, multiple erosions and ulcers, and hemorrhage which diffusely appears from the rectum to the oral side of colon. We previously reported that the activity of ulcerative colitis is inversely proportional to the intensity of fluorescence captured by AFI (Figures 12 and 13) (Fujiya M, et al., 2007).
Enterocolitis is an inflammatory disease of the digestive tract caused by various microbes, parasites and chemicals. While most cases can be diagnosed without the need for endoscopy, some enteritis that persists in spite of antibiotic treatments require endoscopic evaluation. The presence of edematous mucosa with multiple erosions, ulcers and bleeding are typically observed. These findings are enhanced as magenta areas by AFI (Figure 14).
2.6. Aphthoid colitis
Aphthoid colitis is a form of intestinal inflammation which cannot be categorized into any other category of intestinal diseases. Sometimes, aphtoid colitis is an early step in the process of developing Crohn’s disease, Behcet disease or other inflammatory disorders. In AFI, aphtoid ulcers exhibit a clear magenta area surrounded by a blurred magenta area that reflects edema and the infiltration of immune cells (Figure 15).
2.7. Behcet disease
Behcet disease is a systemic vasculitis that frequently presents with mucosal injury and ocular involvement (International Study Group for Behçet's Disease, 1990). The small intestine, particularly the ileocecal valve, is a frequently involved site. Punched-out ulcers and aphtha are typical lesions in the intestinal tract (Figure 16).
Amyloidosis is defined as an extracellular deposit of fibril proteins, P-components, or other glycoproteins in organs and tissues, causing mild to severe pathophysiological changes (Westermark P. 2005). The intestinal tract is a frequently involved site of amyloid deposition (Koppelman RN, et al. 2000). The endoscopic features of intestinal amyloidosis include elevated lesions, ulcerations, nodules, petechial mucosal hemorrhage. AFI enhances these changes as magenta areas (Figure 17).
2.9. Phlebosclerotic colitis
Phelebosclerotic colitis is characterized as the presence of sclerosis with calcification in the tributaries of the superior mesenteric vein (Arimura Y, et al. 1998). This leads to chronic venous insufficiency and congestion, causing abdominal pain, diarrhea and intestinal obstruction. The endoscopic feature of phelebosclerotic colitis is a dark purple mucosa with a marked thickness and absence of the haustra at the right side colon. AFI typically reveals a strong green area with magenta spots at the corresponding site (Figure 18).
2.10. Disuse atrophy
When the intestinal tract does not function for a long time due to a bowel rest or an intestinal operation with a stoma, disuse atrophy of the intestine develops. The typical endoscopic features are an edematous mucosa with easy bleeding. AFI detects magenta areas in the corresponding site (Figure 19).
3. The diagnostic efficacy of AFI for colon disorders
The significance of AFI for diagnosing intestinal disorders, including colon cancer, lymphoma and ulcerative colitis are reviewed in this section.
3.1. Adenoma and cancer
During the process of colon carcinogenesis, normal epithelia are thought to initially turn into benign adenomas, accumulate gene alterations, and then transform into advanced adenocarcinomas (Fearon ER, et al. 1990) (Lengauer C, et al. 1998). All adenomas are thus considered to be premalignant lesions. Indeed, several trials on endoscopic resection for colon adenoma successfully decreased the mortality of colon cancer (Winawer SJ, et al. 1993). However, some patients under close colonoscopic surveillance still develop colorectal cancer (Robertson DJ, et al. 2005.). This discrepancy may be caused by the rapid progression of adenomas, as well as the overlooking of colorectal adenoma by endoscopists with different levels of experience. Indeed, systematic reviews of back-to-back colonoscopies showed that 15% to 32% of colorectal adenomas were possibly missed by colonoscopy (van Rijn JC, et al. 2006), particularly flat and depressed adenomas (Rembacken BJ, et al. 2000)(Saitoh Y, et al. 2001).
Several studies have been conducted to assess the efficacy of AFI for detecting colon adenoma. Matsuda et al. investigated the detection rate of colon polyps at the proximal colon by AFI using a modified back-to-back method. They showed a higher detection rate of colon neoplasms by AFI than that by white light endoscopy (Matsuda T, et al. 2008). In contrast, van den Broek et al. compared the detection rate of AFI with that of HRE using a procedure to inspect the entire colon twice during withdrawal: once with AFI and once with HRE by the same endoscopist. They concluded that AFI showed no significant effect on reducing the adenoma miss-rate in comparison to HRE (van den Broek FJ, et al. 2009). We previously investigated the efficacy of AFI for the detection of colon adenoma by either less-experienced physicians or endoscopic experts, and found that AFI improved the detection rate of colon adenoma, particularly by less-experienced physicians. The effectiveness of AFI for detecting colon adenoma thus appears to depend on the endoscopist and his/her level of experience.
Of interest, AFI does not detect hyperplastic polyps as a clear magenta area (Figure 20). Based on this property, AFI is expected to be used for discriminating colon neoplasms from non-neoplasms. However, the practical usefulness of AFI in the differential diagnosis of colon adenoma remains controversial (van den Broek FJ, et al. 2009)(Boparai KS, et al. 2009)(van den Broek FJ, et al. 2009). We recently investigated the effectiveness of HRE, NBI, AFI and chromoendoscopy for differentiating colon adenoma from hyperplastic polyps. From the numerical analysis of the intensity of fluorescence captured by AFI, the strength of the fluorescence was significantly lower for colon adenoma than that in hyperplastic polyp. Furthermore, our prospective study indicated that AFI improved the diagnostic accuracy to distinguish colon adenoma from hyperplastic polyps, particularly in the resident group (Satoh R, et al. in press). Therefore, it is suggested that AFI captures a lower intensity of fluorescence from colon adenoma in comparison to that from hyperplastic polyps, and AFI can therefore contribute to the differential diagnosis of colon polyps if endoscopists accurately assess the intensity of fluorescence.
While the intestinal involvement of lymphoma cells is enhanced as magenta by AFI as mentioned above, the efficacy of AFI for diagnosing lymphoma has not been thoroughly investigated. Our study concerning the capacity of AFI (Ueno et al. 2010) for diagnosing intestinal lymphoma demonstrated the usefulness of AFI (Figure 21). A numerical analysis of the fluorescence intensity showed that AFI captured a stronger fluorescence from lymphoma than lymphoid hyperplasia (LH) (Figure 22). A histological analysis of intestinal lymphoma and LH revealed that the cell density, but not the histological type, is a significant factor that is inversely proportional to the intensity. Consequently, AFI is a useful tool to diagnose intestinal lymphoma, but cannot be used for the differentiation of histological types.
3.3. Ulcerative colitis
Ulcerative colitis is a chronic refractory colitis whose etiology is still unknown. Various therapeutic strategies for ulcerative colitis are performed according to the type and activity of the disease, thus, the evaluation of the activity is important to choose an appropriate treatment. While clinical symptoms are essential to assess the activity of ulcerative colitis, endoscopic assessment helps to predict the relapse or evaluate the grade of mucosal healing (Fujiya M, et al. 2002).
Our previous investigation proposed that the intensity of fluorescence captured by AFI was inversely proportional to the histological activity in ulcerative colitis (Fujiya M, et al. 2007). Even a small lesion with slightly active inflammation can be clearly detected by AFI (Figures 23 and 24). This suggests that when a numerical analysis of the fluorescence intensity is performed, AFI can evaluate the activity of ulcerative colitis with quite a high reproducibility and inter-observer consistency. Furthermore, it has been reported that AFI may be used to detect dysplasia in the inflamed mucosa of ulcerative colitis (Matsumoto T, et al. 2007). The future analysis of the efficacy of AFI to assess the disease activity and detect dysplasia is therefore expected to confirm the significance of AFI in the diagnosis of ulcerative colitis.
This review describes the typical findings of AFI and the significance of AFI in the diagnosis of intestinal disorders including colon cancer and adenoma, lymphoma, inflammatory bowel diseases, intestinal autoimmune diseases and other conditions. While the efficacy of AFI is still being explored, AFI can definitely be used an efficient tool for objectively assessing intestinal diseases, particularly by less-experienced physicians. In the near future, a numerical analysis of the fluorescence intensity will provide a new diagnostic strategy for intestinal disorders with both high reproducibility and inter-observer consistency.