Inability to noninvasively diagnose gastric intestinal metaplasia in Hispanics or reverse the lesion with Helicobacter pylori eradication.

BACKGROUND: Helicobacter pylori infection has been linked with the development of gastric adenocarcinoma and its precursor lesion, intestinal metaplasia (IM). The presence of gastric IM is not associated with symptoms, which makes identification of individuals with this lesion difficult. It is not clear whether eradication of H. pylori infection leads to reversal of gastric IM or the potential decrease in the risk of cancer in these patients. GOALS: The purpose of this pilot study was to define the prevalence of gastric IM in a population at high risk for gastric cancer (Southwestern Hispanics), examine the ability of noninvasive testing to identify individuals with the lesion, and determine whether eradication of H. pylori infection reverses gastric IM in this population. STUDY: Subjects from the Tucson metropolitan area were recruited, and baseline data, including the presence of upper gastrointestinal (UGI) symptoms, urinary sodium, and serum pepsinogen levels, were obtained. Upper endoscopy was performed and six gastric biopsies from specific anatomic sites were obtained, followed by methylene blue staining with targeted biopsies from blue-stained mucosa. Biopsies were evaluated for the presence of H. pylori infection and gastric IM. A subset of patients with gastric IM were treated to eradicate H. pylori infection. Follow-up exams with methylene blue staining, including biopsies for histology and rapid urease testing, were performed for up to 48 months. RESULTS: There were 84 subjects with a mean age of 53.0 years; 24 (29%) had gastric IM and 65 (77%) had H. pylori. There was no significant association between gastric IM and age, gender, UGI symptoms, H. pylori, or urine sodium. There was an association identified between gastric IM and a decreased pepsinogen I:II ratio (p = 0.03). Of the 11 individuals with gastric IM treated for H. pylori infection, 9 had successful therapy and underwent at least 2 follow-up examinations. The mean length of follow-up was 3.3 years. Eight of the nine (89%) had gastric IM identified histologically at the final endoscopic exam. CONCLUSIONS: H. pylori infection and gastric IM are frequent findings in Southwestern Hispanics, a high-risk population for gastric cancer. Noninvasive testing is not clinically useful in distinguishing individuals within this group who harbor gastric IM. Although eradication of H. pylori infection may lead to a decrease in the amount of gastric IM in some individuals, the lesion may be detected in the majority of individuals after more than 3 years of follow-up. These data suggest that therapy for H. pylori may not eliminate the risk of gastric cancer once IM has developed.

Long-term follow-up of intestinal metaplasia of the gastric cardia.

OBJECTIVE: Recent studies have found a relatively high prevalence of gastric cardia intestinal metaplasia in individuals presenting for elective upper endoscopy. It has been hypothesized that this lesion may be a precursor of gastric cardia cancer. Our objective was to identify the incidence of dysplasia in patients with gastric cardia intestinal metaplasia. METHODS: Twenty-eight patients who had previously been identified with cardia intestinal metaplasia had follow-up examinations performed. None of the patients had dysplasia at the time of diagnosis. All had an examination at 1 yr, and 20 patients had an examination at 3 yr after diagnosis. During follow-up examinations all patients underwent vital staining with methylene blue to help identify areas of intestinal metaplasia in the cardia. Two to four biopsies were taken from blue-stained mucosa. Histological specimens were stained using a combination of hematoxylin and eosin with Alcian blue at pH 2.5. RESULTS: There were 27 men and one woman with a mean age of 69.8 yr (range, 48-83 yr). The mean length of follow-up was 2.5 yr (range, 12-46 months). Only one patient was diagnosed with dysplasia (low-grade) during the study, for an incidence of 1.4% per yr. CONCLUSIONS: The prevalence (0%) and incidence (1.4%/yr) of dysplasia in cardia intestinal metaplasia are low. Although further studies are needed, screening and surveillance for gastric cardia intestinal metaplasia is unlikely to be clinically useful for the prevention of gastric cardia cancer.

Barrett's esophagus: update on screening, surveillance, and treatment.

The last 2 decades have seen dramatic advances in Barrett's esophagus. The definition has evolved; the rising incidence of adenocarcinoma has been recognized; and effective therapy to control gastroesophageal reflux disease has been developed. Both proton pump inhibitor therapy and laparoscopic fundoplication represent major developments. Studies of patients with dysplasia have helped to clarify appropriate surveillance intervals and treatment strategies for these patients, although controversy still exists. The possibility of reversing Barrett's esophagus in selected high-risk patients offers major hope for the future prevention of adenocarcinoma of the esophagus.

Short segment Barrett's esophagus--the need for standardization of the definition and of endoscopic criteria.

There has been a recent increase in abstracts and publications reporting intestinal metaplasia in the distal esophagus and cardia. The terms "short segment Barrett's esophagus," "intestinal metaplasia of the esophagogastric junction," and "intestinal metaplasia of the cardia" are being used to describe either similar or different entities. This review article deals with the current data on these issues, the definition of short segment Barrett's esophagus including the endoscopic and histologic criteria, the rationale for separating short segment Barrett's esophagus from intestinal metaplasia of the cardia, and a simple classification of intestinal metaplasia.

Methylene blue staining for intestinal metaplasia of the gastric cardia with follow-up for dysplasia.

BACKGROUND: Previous studies by our group have identified a 23% prevalence of intestinal metaplasia involving the gastric cardia in patients undergoing elective esophagogastroduodenoscopy (EGD). The current study evaluates the potential clinical utility of vital staining with methylene blue for detecting intestinal metaplasia of the cardia. In addition, the prevalence of Helicobacter pylori infection and incidence of dysplasia in cardia intestinal metaplasia was assessed. METHODS: Twenty-four of 104 patients undergoing elective EGD were previously identified with cardia intestinal metaplasia. Twenty-one of these 24 returned for follow-up EGD with methylene blue staining and cardia biopsies. During EGD four random biopsies were taken in the cardia. After staining with methylene blue, four targeted biopsy specimens were taken from any stained areas. H. pylori infection was assessed by rapid urease testing and serologic evaluation. RESULTS: Staining with methylene blue was positive in 16, equivocal in 4, and negative in 1. Overall, cardia intestinal metaplasia was identified in 17 of 21 patients (81%) by at least one of the 8 biopsies. The sensitivity of detecting intestinal metaplasia by targeted biopsies to stained areas was 67% as opposed to 38% for random biopsies (odds ratio 3.0: 95% CI [0.88,10.24]). The prevalence of H. pylori infection in patients with cardia intestinal metaplasia was 76%. One patient developed low-grade dysplasia, whereas the others remained negative. CONCLUSIONS: Intestinal metaplasia of the gastric cardia is associated with a high prevalence of H. pylori infection and is a reproducible finding in the majority of patients. Methylene blue staining and an increasing number of biopsies improve the sensitivity of detecting this potentially neoplastic lesion.

Squamous islands in Barrett's esophagus: what lies underneath?

OBJECTIVE: Squamous islands are frequently visualized at the time of upper endoscopy in patients with Barrett's esophagus, especially those on proton pump inhibitor therapy (PPI). The significance of these islands is not clearly understood. The aim of this study was to systematically biopsy macroscopic squamous islands and to examine their histologic characteristics. METHODS: Patients with Barrett's esophagus undergoing surveillance had squamous islands documented and biopsied at the time of endoscopy. Barrett's esophagus was defined as the presence of a columnar lined esophagus on endoscopy with intestinal metaplasia on biopsy. All biopsies were obtained by a single senior endoscopist and were stained with alcian blue at pH 2.5. Biopsy samples with inadequate tissue quantity were not included in the study. RESULTS: A total of 39 biopsies were obtained from 22 patients. Twenty of the 22 patients were male, with a mean age of 65.4 yr (range 47-80 yr). The mean length of Barrett's mucosa was 5.6 cm (range 1-11 cm). Eleven of 22 patients were on omeprazole (mean dose 29.1 mg/day), whereas seven patients were on lansoprazole (60 mg/day). The mean duration of PPI therapy was 2.3 yr (range 9-71 months) at the time of biopsy of the squamous islands. Three patients were on H2-blocker therapy whereas the remaining patient had not been started on acid suppression therapy. On histology, 24 biopsy specimens (61.5%) revealed only squamous epithelium, whereas 15 (38.5%) showed the presence of intestinal metaplasia underlying the squamous epithelium. There was no significant difference between the patients with and without underlying intestinal metaplasia in regard to age, Barrett's length, dose, and duration of PPI therapy. CONCLUSION: In more than one-third of biopsies of macroscopic squamous islands within Barrett's esophagus, microscopic intestinal metaplasia is detected. The presence of squamous islands should not be equated with regression of Barrett's esophagus or with decreased cancer risk.

Dysplasia in short-segment Barrett's esophagus: a prospective 3-year follow-up.

OBJECTIVE: Short segments of intestinal metaplasia in the distal esophagus are being recognized with increasing frequency. Both long and short segments of Barrett's esophagus can progress to dysplasia and cancer. However, the risk of short-segment Barrett's esophagus (SSBE) for the development of dysplasia and adenocarcinoma of the esophagus is not yet known. Our purpose, therefore, was to determine the frequency with which dysplasia occurs in patients with SSBE. METHODS: Patients with SSBE were followed prospectively for the development of dysplasia. SSBE was defined as <3 cm of Barrett's-appearing epithelium above the gastroesophageal junction at endoscopy, with intestinal metaplasia on biopsy as documented by alcian blue stain at pH 2.5 on at least two endoscopic biopsies 6 months apart. Patients had interval upper endoscopy with systematic biopsy of the Barrett's segment. RESULTS: Fifty-nine SSBE patients were identified. The mean length of Barrett's mucosa was 1.5 +/- 0.1 cm; the mean age of the patients was 63.1 +/- 1.3 yr. Five patients had low-grade dysplasia (LGD) at initial endoscopy, for a prevalence of 8.5%; none had high grade dysplasia (HGD). Thirty-two patients had follow-up endoscopy over a mean period of 36.9 +/- 5.4 months. Five of these patients developed dysplasia on follow-up, three with LGD and two with HGD, the incidence of any dysplasia being 5.7% per year. One patient with HGD that developed during surveillance progressed to adenocarcinoma of the esophagus over a 2-yr period. The other patient with HGD had LGD on follow-up endoscopy. Six patients with initial LGD had no evidence of dysplasia on follow-up. CONCLUSIONS: The prevalence of dysplasia was 8.5% with an incidence of 5.7% per year in this group of SSBE patients, followed prospectively. Although dysplastic changes may not be identified on follow-up examination, some patients progress to adenocarcinoma. Therefore, we recommend surveillance endoscopy and biopsy in patients with SSBE just as in those with long-segment Barrett's esophagus.

Is Barrett's esophagus associated with intestinal metaplasia of the gastric cardia?

OBJECTIVE: Barrett's esophagus has been associated with adenocarcinoma of the esophagogastric junction and gastric cardia. The purpose of this study was to determine whether patients with Barrett's esophagus have a higher prevalence of intestinal metaplasia involving the gastric cardia than those without Barrett's esophagus. METHODS: Two groups of patients were compared for the prevalence of intestinal metaplasia of the gastric cardia. Group 1 included 50 patients with well-defined Barrett's esophagus who were being followed in an endoscopic surveillance program. Group 2 consisted of 104 individuals participating in a separate study identifying the prevalence of cardia intestinal metaplasia in patients undergoing elective upper endoscopy. Both groups had biopsy specimens taken from the gastric cardia. Eleven patients in group 2 were found to have Barrett's esophagus and were excluded from the analysis. Histological evidence of intestinal metaplasia was defined as specialized columnar epithelium containing goblet cells staining with Alcian blue at pH 2.5. RESULTS: The prevalence of cardia intestinal metaplasia in the 50 patients in group 1 (Barrett's esophagus) was 22%, whereas the prevalence in the 93 patients in group 2 (no Barrett's esophagus) was 24%, which was not a statistically significant difference. A significant difference between groups still could not be identified when the results were examined with regard to equal number of biopsy specimens taken. None of the patients in either group had dysplasia identified within the cardia intestinal metaplasia. CONCLUSIONS: Patients with Barrett's esophagus do not have a higher prevalence of intestinal metaplasia of the gastric cardia than those presenting for routine endoscopy. Although intestinal metaplasia of the gastric cardia is a relatively common finding, dysplasia is uncommon. Therefore, we suggest that screening biopsy specimens of the gastric cardia in patients with Barrett's esophagus be limited to study protocols at this time. In addition, we believe that these data raise the question of whether a true association exists between Barrett's esophagus and gastric cardia cancer.

Yield of routine endoscopy beyond the duodenal bulb.

The authors determined the clinical yield, endoscopic time, and patient tolerance of routine upper endoscopy beyond the duodenal bulb. From May through October 1994, all patients undergoing routine esophagogastroduodenoscopy (EGD) were recruited for study. Each procedure was timed from start to finish by the endoscopy nurse, and, in addition, the time of the postbulbar examination was recorded. The endoscopy nurse assessed the patient's comfort level when the endoscope was advanced into the duodenal bulb and again at the postbulbar region. A total of 250 EGDs were performed. There were 152 males and 98 females, with a mean age of 57.1 (range, 23-91) years. Indications for the procedure were as follows: gastroesophageal reflux disease symptoms 82, epigastric pain 64, dysphagia 46, Barrett's surveillance 25, anemia 23, other research study 16, and other 61. The mean time for the procedure was 11 min and 54 s, whereas the mean time for the postbulbar examination was 46.6 s. Patients tolerated endoscope insertion well both before and during examination of the postbulbar duodenum. The only postbulbar finding that affected clinical management was a postbulbar ulcer in a patient without other ulcers who was positive for Helicobacter pylori. Although routine endoscopic examination beyond the duodenal bulb involves minimal time and is well tolerated by patients, the yield of pathologic findings is low (3.6%) and the yield of findings that alter clinical management even lower (0.4%). In patients without prior GI surgery undergoing routine EGD for indications other than suspected small bowel pathology or active upper GI bleeding, examination of the postbulbar duodenum can be considered an elective part of the procedure.

Intestinal metaplasia of the gastric cardia.

OBJECTIVE: Although the incidence of gastric adenocarcinoma in the U. S. is declining, the incidence of cancer localized to the gastric cardia has risen dramatically. It is not yet clear whether cancer of the gastric cardia arises from a premalignant lesion such as intestinal metaplasia (IM). The purpose of this study was to determine the prevalence of IM involving the cardia in patients presenting for elective EGD, and evaluate potential associated factors. METHODS: During a 7 month period patients referred for elective EGD at the Tucson VA Medical Center were invited to participate in the study. Prior to EGD each patient was surveyed with regard to GERD symptoms and smoking and alcohol history. During EGD note was made of the presence of esophagitis, hiatus hernia, and Barrett's-appearing mucosa. The esophagogastric junction (EGJ) was defined as the end of the tubular esophagus coinciding with the proximal heads of the gastric folds. Seven biopsies were taken from specific sites in the antrum, angularis, cardia, and EGJ. Biopsy specimens were stained with a combination of H & E and Alcian blue at a pH of 2.5. Histologic evidence of IM was defined as columnar-type epithelium including goblet cells staining with Alcian blue. Patients were also evaluated for H. pylori infection by histologic and serologic examinations. RESULTS: There were 104 patients (99 male, 5 female) with a mean age of 61.6 years. Twenty-four (23%) were found to have IM involving the gastric cardia, although none of these had dysplasia. Eleven patients (11%) had Barrett's esophagus; however, only 2 of these had concomitant IM of the cardia. On the other hand, 9 of the 24 with IM of the cardia had concomitant IM elsewhere in the stomach. Forty-nine patients (47%) were found to be positive for H. pylori infection and there was a significant association between H. pylori infection and IM of the gastric cardia (p = 0.03). CONCLUSIONS: These data show that IM of the gastric cardia is a relatively common finding and is associated with H. pylori infection. Although dysplasia was not identified, long-term follow-up studies will be necessary to determine the incidence of dysplasia or adenocarcinoma developing from IM of the cardia.

Adenocarcinoma of the gastric cardia.

Although adenocarcinoma of the stomach has decreased in incidence over the past several decades, cancer of the gastric cardia has increased rapidly over this time frame. There are several differences between adenocarcinoma of the cardia and distal stomach with respect to epidemiology, risk factors, and prognosis. In addition, recent data raise questions with regard to possible associations of cardia cancer with Barrett's esophagus, intestinal metaplasia of the cardia, and Helicobacter pylori. This article will review the current literature with regard to this important tumor and explore these potential disease associations.

Endoscopy of the normal stomach.

Diagnostic upper endoscopy is the best tool to obtain the most information regarding potential gastric abnormalities that may lead to clinical symptoms. Its usefulness is related to its application for appropriate indications by an experienced endoscopist. The endoscopist should not only be familiar with the appropriate techniques to allow for complete visualization of the stomach, but with the normal appearance of the different regions of the stomach. In addition, optimal performance of upper endoscopy requires a knowledge of the subtle variations from normal that do not require further investigation as a potential source of clinical symptoms. This knowledge is often as important to the patient as the ability to recognize obvious abnormalities.

The difference in colon polyp size before and after removal.

BACKGROUND: Accurate knowledge of polyp size is important in assessing cancer risk in both clinical studies and individual patients. We sought to determine if a difference exists between the endoscopic estimation of colon polyp size and the actual measurement after removal. METHODS: We measured polyps in a systematic fashion. Using open biopsy forceps as a guide, the largest diameter of 31 pedunculated polyps was estimated endoscopically. The polyp was then removed by snare polypectomy and directly measured by a technician who was blinded to the endoscopic estimate. Each polyp was also measured after formalin fixation by a pathologist who was blinded to previous measurements. RESULTS: There was a significant difference between the endoscopic estimates and the postpolypectomy measurements. Endoscopic estimates on average were 1.6 mm greater than the postpolypectomy measurements (p< 0.05), representing an 18% difference. Twenty-three of the 31 (74%) endoscopic estimates were larger than the postpolypectomy measurements. There was not a significant difference between the postpolypectomy and postfixation measurements. CONCLUSIONS: The size of polyps measured endoscopically is significantly larger on average than postpolypectomy measurements. This is most likely due to factors involved in the removal of the in vivo polyp.

Liver histology in anti-HCV-positive persons with normal or minimally elevated aminotransferases.

The significance of a positive hepatitis C virus (HCV) screening test in asymptomatic blood donors with normal or near normal aminotransferases was studied along with the usefulness of HCV RNA polymerase chain reaction (PCR) testing for predicting chronic hepatitis in these individuals. One hundred and thirty-nine volunteer blood donors who were found positive by second generation ELISA for antibodies to HCV agreed to participate in the study. Thirty-one of them were supplemental test positive, had ALT values less than twice normal, and were followed over a minimum of 12 months. Thirteen consented to percutaneous liver biopsy and also had HCV RNA determination by PCR. Ten of the 13 subjects were positive for HCV RNA by PCR. Of the nine who were positive for HCV RNA and had adequate tissue for evaluation, seven had evidence of chronic hepatitis, three with limiting plate necrosis. Lobular inflammation was similar in severity to that found in the portal region. In addition, two had periportal fibrosis, and one had bridging fibrosis. Of the three subjects who were negative for HCV RNA, only one had portal inflammation which was limited to the portal region. None of these three had lobular changes, or periportal or bridging fibrosis. Of the three normal biopsies, two were from subjects who were negative for HCV RNA. The sensitivity and specificity of HCV RNA testing for chronic hepatitis was 87.5% and 50%, respectively, yielding an overall accuracy of 75%. We conclude that asymptomatic blood donors with antibodies to HCV, normal or mildly elevated liver tests, and HCV RNA may have abnormal liver histology indicating the potential for progressive liver disease. HCV RNA testing by PCR may be clinically useful as a noninvasive means to discriminate between those with and without chronic liver disease.

Screening for Barrett's esophagus by balloon cytology.

OBJECTIVE: to determine whether balloon cytology would allow a nonendoscopic diagnosis of Barrett's esophagus. METHODS: In 10 unselected patients with Barrett's esophagus, balloon cytology was performed with a Brandt esophageal cytology balloon catheter prior to endoscopy/esophageal biopsy. Specimens were cytologically evaluated for the presence of goblet cells and/or dysplasia. RESULTS: No patient had goblet cells or definitive columnar cell dysplasia. CONCLUSION: This technique with a commercially available balloon catheter does not allow for a nonendoscopic diagnosis of Barrett's esophagus.