How to screen
Previous studies have found a poor correlation between the endoscopic determination of gastric atrophy and the histologic diagnosis.42 Several studies also found that gastric cancer was missed on initial endoscopic examinations. Sensitivity of endoscopy to detect gastric cancer has ranged from 77% to 93%.43,44 In the United States, there is a lack of standardized quality indicators for upper endoscopy exams. The ESGE has suggested several performance measures to ensure a quality endoscopy exam, including accurate photo documentation, sufficient procedure time of at least 7 minutes, adherence to biopsy protocols, and low complication rates.45 In Asia, a systematic screening protocol is used for photo documentation, and simple techniques such as adequate air insufflation and irrigation to remove mucus are routinely used to improve the endoscopy exam.46,47 The mean time of an endoscopy exam has also been found to increase the detection of neoplastic lesions, as slow endoscopists – with a mean exam duration of 8.6 ± 4.2 min during upper endoscopy – detected threefold more neoplastic lesions than did fast endoscopists.48
A standardized biopsy approach is also important when screening patients. The updated Sydney protocol has been suggested for mapping the stomach to screen for atrophy and GIM. This protocol recommends two biopsies from the antrum (at the lesser and greater curvature), two from the body (at the lesser and greater curvature), and one from the incisura.23 This biopsy protocol was also suggested in the recent MAPS II update, with the biopsy of the incisura felt to be an additional biopsy left to the discretion of the endoscopist. Notably, abnormal appearing mucosal areas should be biopsied separately from the mapping biopsies.
High-definition endoscopy with virtual chromoendoscopy is felt to be better than white-light endoscopy alone at detecting precancerous gastric lesions.38 (See Figure 2.)
Courtesy Diana Curras-Martin, MD, and Susana Gonzalez, MD
Figure 2. A. High definition white light endoscopy of patient with diffuse gastric intestinal metaplasia. B. NBI image of patient with diffuse GIM shows ridge and villous appearance. C. High powered H&E of biopsy shows intestinal metaplasia.
In particular, narrow-band imaging (NBI) has been studied and found to increase the diagnostic yield of GIM and dysplasia compared with white light alone.49 Several studies have shown an increased accuracy for the detection of GIM with magnification NBI.50-52 An unfortunate limitation is the geographic availability of magnification NBI: It is not available in the United States. A multicenter study in Portugal developed a new classification system for the appearance of precancerous lesions with NBI and tested its accuracy in endoscopists with a wide range of NBI experience. An abnormal mucosal pattern that showed light blue crests/regular ridge or a tubulovillous appearance and a regular mucosal pattern was found with GIM. An irregular vascular pattern with a white opaque substance and an absent or irregular mucosal pattern was most often found with dysplasia. Furthermore, the reproducibility of these patterns was high between endoscopists.53 Multiple studies have been performed on additional imaging technologies to enhance the detection of gastric neoplasia; however, these technologies are still investigational and currently not recommended for screening.54-57
Serum pepsinogens have been studied in Europe and Asia as noninvasive indicators of gastric atrophy to determine who should be screened with endoscopy.58 A low serum pepsinogen I level below 70 ng/mL and pepsinogen I/II ratio below 3 has generally been used to detect atrophic gastritis and at-risk populations. However, the studies performed in Europe and Asia used different methods for quantifying pepsinogen levels. Therefore, cutoff values cannot be generalized for all assays and should be validated for the specific tests used.38
