- Email: [email protected]
Enhanced magnification endoscopy: A new technique to identify specialized intestinal metaplasia in Barrett's esophagus
Enhanced magnification endoscopy: a new technique to identify specialized intestinal metaplasia in Barrett’s esophagus Moises Guelrud, MD, Idamys Herrera, MD, Harold Essenfeld, MD, Julio Castro, MD Caracas, Venezuela
Background: Specialized intestinal metaplasia (SIM) in Barrett’s esophagus (BE) is not identifiable by standard endoscopy. Acetic acid instillation enhances the ability to detect columnar epithelium at the squamocolumnar union. Enhanced magnification endoscopy involves the combined use of magnification endoscopy with acetic acid. This study assessed the value of enhanced magnification endoscopy in detecting SIM in patients with BE. Methods: Patients undergoing endoscopic surveillance because of short segment BE without dysplasia underwent enhanced magnification endoscopy with 1.5% acetic acid instillation. Standard endoscopy was followed by magnification endoscopy and repeated after acetic acid spraying. Surface patterns were characterized prior to and after acetic acid spraying. The observed surface patterns were compared with histology results. Results: Forty-nine patients, 9 women and 39 men, with a mean age of 50.5 years were studied. One was excluded because of unclear definition of the surface pattern. Enhanced magnification endoscopy detected 4 different mucosal surface patterns: I, round pits; II, reticular; III, villous; and IV, ridged. A total of 129 areas were examined. Standard endoscopy identified an endoscopic pattern in 1.5% of the areas, standard endoscopy and acetic acid in 8.5%, magnification endoscopy alone in 38%, and enhanced magnification endoscopy in all 129 endoscopic areas. The yields for detecting SIM according to endoscopic patterns were as follows: pattern I, 0%; II, 11% (odds ratio 0.5, p = 0.54); III, 87% (odds ratio 36, p = 0.001); and IV, 100% (odds ratio 14, p = 0.015). Conclusions: Enhanced magnification endoscopy is an accurate method of predicting SIM in BE. The simplicity of the technique and its ability to identify characteristic endoscopic patterns with outstanding clarity and resolution that correlate with histologic identification of specialized intestinal metaplasia make enhanced magnification endoscopy an excellent method for the evaluation of patients with BE. (Gastrointest Endosc 2001;53:559-65.)
Barrett’s esophagus (BE) is an acquired condition in which the squamous epithelium of the distal esophagus undergoes metaplasia to specialized intestinal epithelium.1,2 Specialized intestinal metaplasia (SIM) in BE is not identifiable by standard endoscopy. The finding of any columnar-appearing mucosa in the distal esophagus on endoscopy raises the suspicion of BE, but does not make the diagnosis by itself.3 Endoscopic biopsies are necessary for the definitive diagnosis of BE. The drawback of the endoscopic biopsy is that sampling error influences sensitivity. In an effort to avoid this type of error, multiple biopsy specimens are often obtained. Received May 15, 2000. For revision August 15, 2000. Accepted January 5, 2001. From the Departments of Gastroenterology, Anatomic Pathology, and Biostatistics and Epidemiology, Policlinica Metropolitana, Caracas, Venezuela. Reprint requests: Moises Guelrud, MD, Policlínica Metropolitana, Urb. Caurimare, Calle A1, Caracas 1060, Venezuela. Copyright © 2001 by the American Society for Gastrointestinal Endoscopy 0016-5107/2001/$35.00 + 0 37/1/114059 doi:10.1067/mge.2001.114059 VOLUME 53, NO. 6, 2001
A technique that improves mucosal visualization would provide a more accurate endoscopic diagnosis of BE. To obtain a better visual recognition of BE, several endoscopic techniques have been used, alone or in combination, including chromoendoscopy, acetic acid instillation, and magnification endoscopy. Chromoendoscopy involves the use of dyes to enhance the accuracy of endoscopic examination. The two most frequently used dyes in the esophagus are Lugol’s solution and methylene blue. Lugol’s is an iodine-based absorptive dye with an affinity for the glycogen in nonkeratinized squamous epithelium. The normal esophageal mucosa turns a greenbrown color, leaving any columnar epithelium unstained. Thus, it is a negative stain for BE. By increasing the demarcation between brownish squamous mucosa and unstained, pink metaplastic columnar epithelium, Lugol’s staining can increase the accuracy of endoscopic diagnosis of BE to 91%.3 Methylene blue stains actively absorbing cells such as those of intestinal metaplasia and is not absorbed by normal squamous esophageal cells. Methylene blue staining has a 95% sensitivity and GASTROINTESTINAL ENDOSCOPY
559
M Guelrud, I Herrera, H Essenfeld, et al.
Enhanced magnification endoscopy in Barrett’s esophagus
A
B
C
D
Figure 1. Short segment Barrett’s esophagus. A, Standard endoscopic view. Short extensions of columnar-appearing mucosa with squamous islands within Barrett’s epithelium. Arrowhead indicates the field to be magnified. B, Magnification endoscopic view of the same field. A villiform appearance is imperfectly observed. C, Standard endoscopic view after acetic acid instillation. The border of the squamous-columnar junction is better visualized as well as the squamous islands within the Barrett’s epithelium. Arrowhead indicates the field to be magnified. D, Enhanced magnification endoscopic view of the same field. A fine villiform appearance is clearly evident.
97% specificity for the biopsy finding of SIM.4 In two recent studies, however, methylene blue staining with targeted biopsies failed to improve the diagnostic yield for BE.5,6 Moreover, the technique is messy and substantially lengthens the procedure time. Acetic acid produces reversible intracellular cytoplasmic protein denaturation.7 Currently, it is used during colposcopy as an aid to detect small lesions in the uterine cervical mucosa, a mucosa similar to that at the squamocolumnar junction.8 It has been shown that acetic acid instillation enhances the ability to detect small or indistinct remnant islands of columnar epithelium after endoscopic treatment of BE, and that this method is safe, rapid, clean, and inexpensive.9 Standard endoscopy is limited in terms of capability to evaluate columnar-appearing epithelium because the superficial mucosal layer is visualized at 560
GASTROINTESTINAL ENDOSCOPY
low magnification. Magnification endoscopy, which involves use of a special endoscope, has been used with and without dye spraying to better evaluate the columnar epithelium in BE.10,11 Enhanced magnification endoscopy involves the combined use of magnification endoscopy and acetic acid. Our goal was to determine the value of enhanced magnification endoscopy in detecting SIM in patients with BE. MATERIAL AND METHODS Patients A total of 49 patients with previously diagnosed short segment BE without dysplasia who were being followed in an endoscopic surveillance program were included in the study. Short segment BE was defined as a length of 3 cm or less of columnar-appearing mucosa proximal to the esophagogastric junction at endoscopy with histologic conVOLUME 53, NO. 6, 2001
Enhanced magnification endoscopy in Barrett’s esophagus
A
C
M Guelrud, I Herrera, H Essenfeld, et al.
B
D
Figure 2. Endoscopic views after acetic acid instillation; 4 different patterns of the mucosal surface were observed. A, Pattern I: round pits with a characteristic pattern of regular and orderly arranged circular dots. B, Pattern II: reticular pits that are circular or oval and are regular in shape and arrangement. C, Pattern III: villous with no pits present but a fine villiform appearance with regular shape and arrangement is evident. D, Pattern IV: ridged with no pits present but a thick villous convoluted shape with a cerebriform appearance with regular shape and arrangement is evident. firmation of SIM on biopsy.12 The study protocol was approved by the ethics committee of our institution. All participating patients gave written informed consent. Magnification gastroscope All procedures were performed with a magnification videoendoscope (GIF-200Z, Olympus America Inc., Melville, N.Y.), which has the capability of both standard videoendoscopy and an adjustable image magnification in a continuous range up to ×35. The instrument has a depth of field in the wide position of 7 to 100 mm and a narrow depth of field of 4 to 11 mm. The diameter of the endoscope is 10.5 mm with a working length of 103 cm and a 2.8 mm accessory channel. Tip angulation is slightly less in the upward direction than a standard gastroscope (180° vs 210°), but the downward and left-right deflection ranges are the same. Endoscopic procedure Endoscopic procedures were performed by a single gastroenterologist (MG) experienced in the technique of VOLUME 53, NO. 6, 2001
enhanced magnification endoscopy using acetic acid instillation. Patients first underwent standard endoscopy followed by magnification endoscopy. Then 10 to 15 mL of 1.5% acetic acid were sprayed onto the distal esophagus with a spray catheter (Olympus, PW-5V). Immediately afterward, the distal esophagus was irrigated with 3 to 5 mL of water by using the lens-cleaning water channel of the endoscope. There was no attempt to wash away the excess acetic acid. In the following 2 to 3 minutes, standard endoscopy and magnification endoscopy were repeated (Fig. 1). Acetic acid procedure time was recorded from the point at which acetic acid was first sprayed to the time at which biopsy specimens were obtained. All examinations were videotaped and photographed. Frequently the endoscopic pattern was diagnosed while the image was captured on the monitor to obtain a photograph. After acetic acid instillation, 4 different mucosal surface patterns were observed (Fig. 2). Pattern I: round pits with a characteristic pattern of regular and orderly GASTROINTESTINAL ENDOSCOPY
561
M Guelrud, I Herrera, H Essenfeld, et al.
Enhanced magnification endoscopy in Barrett’s esophagus
Table 1. Frequencies of endoscopic patterns observed with each endoscopic technique
Pattern I II III IV
Number of total areas 48 18 46 17
SE N areas (%) 0 0 1 1
SE+AA
ME
EME
N areas (%)
N areas (%)
N areas (%)
(0) (0) (2) (6)
0 0 6 5
(0) (0) (13) (29)
10 7 21 11
(21) (39) (46) (65)
48 18 46 17
(100) (100) (100) (100)
SE, Standard endoscopy; AA, acetic acid; ME, magnification endoscopy; EME, enhanced magnification endoscopy.
arranged circular dots. Pattern II: reticular. Pits were circular or oval and were regular in shape and arrangement. Pattern III: villous. No pits were present but there was a fine villiform appearance with regular shape and arrangement. Pattern IV: ridged. No pits were present but there was a thick villous convoluted shape with a cerebriform appearance with regular shape and arrangement. Biopsy protocol One targeted biopsy specimen was taken from each representative pattern with a standard biopsy forceps (FB24U, Olympus) under low magnification. Occasionally, more than one biopsy sample was obtained. In all cases, a single specimen was taken from a pattern I area located distal to the esophagogastric junction. Formalin-fixed samples were stained with H&E. Additional alcian blue staining at pH 2.5 was performed to detect SIM, which was defined by the presence of goblet cells. Biopsy specimens were considered positive for SIM even if they contained only minute, localized SIM. Two pathologists blinded to the method of obtaining the sample independently examined all specimens. Results of histologic examinations were compared with endoscopic findings. Statistical analysis Histologically confirmed SIM was used as the main outcome, and univariate logistical regression was used to calculate the odds ratios (OR) in enhanced magnification endoscopy, using pattern I as reference group. Sensitivity, specificity, positive predictive value, negative predictive value, and ROC (receiver operator characteristic) were calculated for the enhanced magnification technique to assess the yield of the model. A model with no predictive power has an area of 0.5; a perfect model has an area of 1. All values were calculated by using 0.05 significance criteria and 2 tails. STATA 5.0 (Stata Corp., College Station, Tex.) statistical software was used to perform the analysis.
RESULTS Patients Between August 1999 and November 1999, 49 patients with short segment Barrett’s esophagus were studied. One patient was excluded because of unclear definition of the surface pattern after acetic acid instillation. Of the 48 patients, 9 (19%) were women and 39 (81%) were men with a mean age of 50.5 years (range 24–76). Barrett’s epithelium was 562
GASTROINTESTINAL ENDOSCOPY
present in tongue-like or circumferential-type patterns. Twenty-one of the 48 patients were being treated long-term with lansoprazole 30 mg every day; 10 patients took this drug irregularly. The other 17 patients were not taking anti-secretory medication. No adverse clinical events were caused by the acetic acid instillation procedure. It was estimated that the technique added between 5 to 8 minutes to the length of the examination. Endoscopic patterns In the first 2 to 3 minutes after acetic acid spraying, there was a whitish discoloration of the esophageal mucosa, the BE, and the gastric mucosa. Thereafter the esophagus remained white and the columnar epithelium of both the BE and gastric epithelium became reddish. A total of 129 areas from 48 patients were examined. Table 1 shows the percentage of endoscopic patterns observed with each endoscopic technique. Standard endoscopy identified an endoscopic pattern in 1.5% (2 of 129) of the areas. When acetic acid was added, standard endoscopy increased the detection yield to 8.5% (11 of 129) areas. When magnification endoscopy alone was used, the detection yield increased to 49 of 129 areas (38%). In contrast, enhanced magnification endoscopy allowed recognition and classification of the 129 endoscopic patterns. Table 2 shows the yield of detecting SIM according to the endoscopic pattern observed with enhanced magnification endoscopy. In 48 biopsy samples obtained from the same number of endoscopic areas with pattern I located in the stomach (hernia sac) distal to the esophagogastric junction, all had fundic columnar epithelium without SIM. This endoscopic pattern served as a control group for the analysis. Frequently, small areas of pattern III or pattern IV were observed within a pattern I area. In order to have a homogeneous group of pattern I areas, these areas were avoided as sites for performing biopsies. Analysis showed correct identification of SIM in 11% (odds ratio 0.5, p = 0.54) of the biopsy specimens of pattern II, in 87% (odds radio 36, p = 0.001) of pattern III, and in 100% (odds ratio 14, p = 0.015) of pattern IV. VOLUME 53, NO. 6, 2001
Enhanced magnification endoscopy in Barrett’s esophagus
By using histology as the main outcome (Table 3), the sensitivity, specificity, positive predictive value, and negative predictive value of enhanced magnification endoscopy for patterns III and IV for detecting SIM were 96.5%, 88.7%, 87.5%, and 96.9%, respectively. The overall accuracy was 92.2% with an ROC of 0.92. DISCUSSION The enhanced magnification endoscopy technique revealed characteristic patterns in SIM and identified SIM with a high degree of accuracy. The fine surface structure of Barrett’s epithelium has a characteristic villous pattern, regular in shape and orderly in arrangement, that suggests a histologic diagnosis of intestinal metaplasia. Currently, the endoscopic diagnosis of BE is based entirely on gross changes in the mucosal appearance. Endoscopic examination usually can distinguish the pale, glossy squamous epithelium of the esophagus from the red, velvet-like columnar epithelium of BE. However, the endoscopic appearance of the fine surface structure of BE is too subtle to be detected with a standard endoscope. A normal mucosal appearance does not necessarily imply normal histology. Thus the different types of columnar epithelium can be distinguished only by histologic examination of tissue specimens.13 Because metaplasia is patchy, the use of random biopsies to assess the presence of SIM in the columnar epithelium is likely to suffer from sampling error. To avoid this type of error, multiple biopsy specimens are usually obtained. Even with this approach, BE was detected in follow-up examinations in only 80% of patients 6 weeks after the initial examination.14 Moreover, inaccurate biopsy targeting may worsen diagnostic accuracy. It can be difficult to precisely obtain biopsy samples given the tangential orientation of the wall in relation to the forceps. The use of regular forceps might sample more mucosa than desired, leading to heterogeneity of different types of columnar epithelium within a single sample. Careful endoscopic technique with targeted biopsies can prevent this problem. To improve the diagnosis of BE, chromoendoscopy, acetic acid instillation, and magnification have been used. Lugol’s solution has been used to define the squamocolumnar junction accurately,15 to identify areas of possible BE,16 and to improve the diagnosis of BE.3 It also has been used in combination with magnification endoscopy and indigo carmine instillation to diagnose the villiform appearance of BE.11 However, by using two different staining methods, the procedure is prolonged and more troublesome. Moreover, severe allergic reacVOLUME 53, NO. 6, 2001
M Guelrud, I Herrera, H Essenfeld, et al.
Table 2. Yield of detecting SIM according to endoscopic pattern Pattern
SIM present
I II III IV
0/48 2/18 40/46 17/17
(0%) (11%) (87%) (100%)
Odds ratio
p
– 0.5 36 14
– 0.543 0.001 0.015
Statistical analyses were performed comparing results from pattern I (control) with patterns II, III, and IV. SIM, Specialized intestinal metaplasia.
tions have been reported with Lugol’s solution.11 Unlike Lugol’s solution, methylene blue selectively stains specialized columnar cell.4 The technique involves application of a mucolytic agent, followed by dye application and then washing of excess dye. Positive staining is defined as the presence of bluestained mucosa that persists despite vigorous water irrigation. In a prospective control study methylene blue revealed SIM in 18 of 26 patients with a 95% sensitivity for the biopsy finding of SIM.4 These results have been questioned in two recent studies. One study revealed SIM in 54% of focally stained mucosa and in 63% of random biopsy samples.5 A second study found SIM in 74% of methylene bluestained mucosa compared with 69% of random biopsy samples.6 Most important, in the latter study, the investigators found that if biopsy specimens were not obtained for unstained areas, SIM would have been missed in 6 of 21 patients. Besides adding significant time to the endoscopic procedure, methylene blue staining is somewhat subjective. It is influenced by the washing technique used. Inadequate lavage after staining may cause false-positive staining of gastric surface cells. Inadequate exposure of the BE to the mucolytic agent and methylene blue can cause SIM to appear pink, resulting in a falsepositive indication of gastric mucosa. Acetic acid has been used to improve detection of small remnant islands of Barrett’s epithelium after ablative therapy.9 In 11 of 21 (52%) patients in one study, acetic acid demonstrated small remnant islands of columnar epithelium not seen before acid instillation. However, the pattern of the columnar epithelium could not be demonstrated, making it impossible to differentiate gastric epithelium from BE. A better method of endoscopic diagnosis would be one that displays microscopic surface morphology during actual endoscopic observation. Magnification endoscopy has been used for several purposes.17 This is the first report of the use of enhanced magnification endoscopy, which combines magnification endoscopy and acetic acid instillation, in BE. Enhanced magnification endoscopy with acetic acid allows visualization of actual villi and crypt areas similar to those GASTROINTESTINAL ENDOSCOPY
563
M Guelrud, I Herrera, H Essenfeld, et al.
Enhanced magnification endoscopy in Barrett’s esophagus
Table 3. Statistical performance according to the endoscopic technique Technique SE SE+AA ME EME
Sensitivity
Specificity
PPV
NPV
Total accuracy
ROC
0 18.9 41.3 96.5
100 97.1 97.1 88.7
– 84.6 92.3 87.5
55 59.4 66.9 96.9
55 62 72 92.2
0.50 0.58 0.69 0.92
PPV, Positive predictive value; NPV, negative predictive value; ROC, receiver operating characteristic; SE, standard endoscopy; AA, acetic acid; ME, magnification endoscopy; EME, enhanced magnification endoscopy.
observed with a stereoscopic microscope. Our analysis of enhanced magnification endoscopy revealed that SIM was characterized by two patterns of surface features. These patterns could be clearly visualized by enhanced magnification endoscopy. Surface features characterized by the absence of round pits and the presence of fine villous or ridged (cerebriform) patterns are suggestive of SIM. The presence of circular dots or round pits is characteristic of gastric columnar epithelium. Our histologic examination was consistent with this hypothesis. Enhanced magnification endoscopy achieved a positive predictive value of 87.5% and a total accuracy of 92.2% in the detection of SIM for pattern IV and a specificity of 100% for the detection of fundic columnar epithelium for pattern I. This suggests that this technique might have the potential to achieve accuracies comparable to those of clinical histopathology. The detection rate of 11% of SIM in the reticular pattern II could be explained by the use of a regular forceps that may have sampled more mucosa than desired, including some metaplastic epithelium. The use of smaller biopsy forceps for more precise targeting might overcome this problem. In this regard, the villous pattern of the surface mucosa may be a clue to the presence of SIM in patients with BE or in patients with intestinal metaplasia of cardia. It may explain why the intestinal metaplasia of the cardia is not more common in patients with BE than in control subjects.18,19 It is also possible that if targeted biopsy specimens are obtained, the prevalence of intestinal metaplasia at the gastric cardia would be higher than previously reported. Moreover, our finding of small areas of intestinal metaplasia within normal gastric mucosa in the hiatal hernias might explain the conflicting results of the association between intestinal metaplasia of cardia and intestinal metaplasia elsewhere in the stomach observed in some studies in which biopsy samples were taken randomly.20-22 The potential advantage of enhanced magnification endoscopy with acetic acid instillation is that targeting biopsies can decrease the sampling error and the number of biopsy samples needed for diagnosis. The technique is simple, fast, not messy, and highly accurate. Given its high sensitivity, specificity, and predictive values, enhanced magnification endoscopy may 564
GASTROINTESTINAL ENDOSCOPY
be useful for a variety of clinical applications, such as supplementing visual assessment of BE discovered during upper endoscopy, targeting of biopsies when short segment BE or intestinal metaplasia of cardia is suspected, guiding ablative therapy, recognizing residual metaplastic mucosa after ablative therapy without obtaining biopsy specimens, and differentiating esophagitis and regenerative squamous epithelium from columnar epithelium. However, prospective studies with enhanced magnification endoscopy in these particular applications need to be performed before the technique can be recommended for routine use. Our study included data obtained from patients with known BE. Patterns were classified by visual appearance. Because the mucosal effect of the acetic acid lasts only 2 to 3 minutes, the esophagogastric junction is constantly moving (because of respiration, cardiac function, and secondary esophageal contractions caused by air insufflation), and the instrument has a short focal range, the endpoint of enhanced magnification endoscopy is somewhat subjective and is the most difficult part of the procedure to learn. Expertise can be achieved in approximately 10 procedures by experienced endoscopists. Moreover, and for the same reason, there is a chance of sampling error. Nonetheless there did not appear to be a strong correlation between misclassification by enhanced magnification endoscopy and SIM by histology. Enhanced magnification endoscopy does seem to be a useful aid to enhancing tissue diagnosis. However, the present study had significant limitations in that it was an open study, which introduces the potential for bias, it did not include a control group, and interobserver agreement, which could lead to problems with generalization, was not assessed. Thus, a randomized blinded study will be needed to confirm our results. In conclusion, simplicity, the ability to identify characteristic endoscopic patterns with outstanding clarity and resolution that correlate with histologic identification of SIM, makes enhanced magnification endoscopy an excellent method for the evaluation of patients with BE and possibly also for patients with intestinal metaplasia of the gastric cardia. VOLUME 53, NO. 6, 2001
Enhanced magnification endoscopy in Barrett’s esophagus
ACKNOWLEDGMENT We thank Lawrence J. Brandt, MD, for reviewing the manuscript. REFERENCES 1. Paull A, Trier JS, Dalton MD, Camp RC, Loeb P, Goyal RK. The histologic spectrum of Barrett’s esophagus. N Engl J Med 1976;295:476-80. 2. Spechler SJ, Goyal RK. The columnar-lined esophagus, intestinal metaplasia and Norman Barrett. Gastroenterology 1996;110:614-21. 3. Woolf GM, Riddell RH, Irvine EJ, Hunt RH. A study to examine agreement between endoscopy and histology for the diagnosis of columnar lined (Barrett’s) esophagus. Gastrointest Endosc 1989;35:541-4. 4. Canto MI, Setrakian S, Petras RE, Blades E, Chak A, Sivak MV Jr. Methylene blue selectively stains intestinal metaplasia in Barrett’s esophagus. Gastrointest Endosc 1996;44:1-7. 5. Jobson B, Goenka P, Manalo G, Thomas E. Methylene blue staining for intestinal metaplasia in Barrett’s esophagus—is it as good as we think? [abstract]. Gastrointest Endosc 1999;49:AB15. 6. Ray MB, Mayfield-Stokes S, Cecil B, Forouzandeh B, Davis A, Wilson MA, et al. Results of methylene blue-directed biopsy is similar to conventional biopsy for the diagnosis of intestinal metaplasia and dysplasia in Barrett’s esophagus [abstract]. Gastrointest Endosc 1999;49:AB22. 7. Cartier R. Practical colposcopy. Basel, Switzerland: Karger; 1978. 8. Cartier R. Role de la colposcopie dans le diagnostic et le traitement des dysplasies et des carcinomes intra-epitheliaux du col uterin. Bull Cancer 1979;66:447-54. 9. Guelrud M, Herrera I. Acetic acid improves identification of remnant islands of Barrett’s epithelium after endoscopic therapy. Gastrointest Endosc 1998;47:512-5. 10. Takemoto T, Sakaki N. Magnification endoscopy. In: Sivak Jr MV, editor. Gastroenterological endoscopy. Philadelphia: WB Saunders; 1999.
VOLUME 53, NO. 6, 2001
M Guelrud, I Herrera, H Essenfeld, et al.
11. Stevens PD, Lightdale CJ, Green PH, Siegel LM, GarciaCarrasquillo RJ, Rotterdam H. Combined magnification endoscopy with chromoendoscopy for the evaluation of Barrett’s esophagus. Gastrointest Endosc 1994;40:747-9. 12. Spechler SJ, Zeroogian JM, Antonioli DA, Wang HH, Goyal RK. Prevalence of metaplasia at the gastro-oesophageal junction. Lancet 1994;344:1533-6. 13. Weinstein WM, Ippoliti AF. The diagnosis of Barrett’s esophagus: goblets, goblets, goblets. Gastrointest Endosc 1996;44:91-4. 14. Kim SL, Waring JP, Spechler SJ, Sampliner RE, Doos WG, Krol WF, et al. Diagnostic inconsistencies in Barrett’s esophagus. Gastroenterology 1994;107:945-9. 15. Northmann BJ, Wright JR, Schuster MM. In vivo vital staining as an aid to identification of esophagogastric mucosal junction in man. Am J Dig Dis 1972;17:919-24. 16. Burbige EJ, Radigan JI. Characteristics of the columnar-cell lined (Barrett’s) esophagus. Gastrointest Endosc 1979;25: 133-6. 17. Tada M, Kawai K. Research with the endoscope: new techniques using magnification and chromoscopy. Clin Gastroenterol 1986;15:417-37. 18. Morales TG, Bhattacharyya A, Johnson C, Sampliner RE. Is Barrett’s esophagus associated with intestinal metaplasia of the gastric cardia? Am J Gastroenterol 1997;92:1818-22. 19. Weston AP, Krmpotich PT, Cherian R, Dixon A, Tapalovski M. Prospective evaluation of intestinal metaplasia and dysplasia within the cardia of patients with Barrett’s esophagus. Dig Dis Sci 1997;42:597-602. 20. Morales TG, Sampliner RE, Bhattacharyya A. Intestinal metaplasia of the gastric cardia. Am J Gastroenterol 1997;92: 414-8. 21. Hackelsberger A, Gunter T, Schultze V, Manes G, DominguezMuñoz JE, Roessner A, et al. Intestinal metaplasia at the gastro-oesophageal junction: Helicobacter pylori gastritis or gastro-oesophageal reflux disease. Gut 1998;43:17-21. 22. Goldblum JR, Vicari JJ, Falk GW, Rice TW, Peek RM, Easley K, et al. Inflammation and intestinal metaplasia of the gastric cardia: the role of gastroesophageal reflux and H pylori infection. Gastroenterology 1998;114:633-9.
GASTROINTESTINAL ENDOSCOPY
565
Background: Specialized intestinal metaplasia (SIM) in Barrett’s esophagus (BE) is not identifiable by standard endoscopy. Acetic acid instillation enhances the ability to detect columnar epithelium at the squamocolumnar union. Enhanced magnification endoscopy involves the combined use of magnification endoscopy with acetic acid. This study assessed the value of enhanced magnification endoscopy in detecting SIM in patients with BE. Methods: Patients undergoing endoscopic surveillance because of short segment BE without dysplasia underwent enhanced magnification endoscopy with 1.5% acetic acid instillation. Standard endoscopy was followed by magnification endoscopy and repeated after acetic acid spraying. Surface patterns were characterized prior to and after acetic acid spraying. The observed surface patterns were compared with histology results. Results: Forty-nine patients, 9 women and 39 men, with a mean age of 50.5 years were studied. One was excluded because of unclear definition of the surface pattern. Enhanced magnification endoscopy detected 4 different mucosal surface patterns: I, round pits; II, reticular; III, villous; and IV, ridged. A total of 129 areas were examined. Standard endoscopy identified an endoscopic pattern in 1.5% of the areas, standard endoscopy and acetic acid in 8.5%, magnification endoscopy alone in 38%, and enhanced magnification endoscopy in all 129 endoscopic areas. The yields for detecting SIM according to endoscopic patterns were as follows: pattern I, 0%; II, 11% (odds ratio 0.5, p = 0.54); III, 87% (odds ratio 36, p = 0.001); and IV, 100% (odds ratio 14, p = 0.015). Conclusions: Enhanced magnification endoscopy is an accurate method of predicting SIM in BE. The simplicity of the technique and its ability to identify characteristic endoscopic patterns with outstanding clarity and resolution that correlate with histologic identification of specialized intestinal metaplasia make enhanced magnification endoscopy an excellent method for the evaluation of patients with BE. (Gastrointest Endosc 2001;53:559-65.)
Barrett’s esophagus (BE) is an acquired condition in which the squamous epithelium of the distal esophagus undergoes metaplasia to specialized intestinal epithelium.1,2 Specialized intestinal metaplasia (SIM) in BE is not identifiable by standard endoscopy. The finding of any columnar-appearing mucosa in the distal esophagus on endoscopy raises the suspicion of BE, but does not make the diagnosis by itself.3 Endoscopic biopsies are necessary for the definitive diagnosis of BE. The drawback of the endoscopic biopsy is that sampling error influences sensitivity. In an effort to avoid this type of error, multiple biopsy specimens are often obtained. Received May 15, 2000. For revision August 15, 2000. Accepted January 5, 2001. From the Departments of Gastroenterology, Anatomic Pathology, and Biostatistics and Epidemiology, Policlinica Metropolitana, Caracas, Venezuela. Reprint requests: Moises Guelrud, MD, Policlínica Metropolitana, Urb. Caurimare, Calle A1, Caracas 1060, Venezuela. Copyright © 2001 by the American Society for Gastrointestinal Endoscopy 0016-5107/2001/$35.00 + 0 37/1/114059 doi:10.1067/mge.2001.114059 VOLUME 53, NO. 6, 2001
A technique that improves mucosal visualization would provide a more accurate endoscopic diagnosis of BE. To obtain a better visual recognition of BE, several endoscopic techniques have been used, alone or in combination, including chromoendoscopy, acetic acid instillation, and magnification endoscopy. Chromoendoscopy involves the use of dyes to enhance the accuracy of endoscopic examination. The two most frequently used dyes in the esophagus are Lugol’s solution and methylene blue. Lugol’s is an iodine-based absorptive dye with an affinity for the glycogen in nonkeratinized squamous epithelium. The normal esophageal mucosa turns a greenbrown color, leaving any columnar epithelium unstained. Thus, it is a negative stain for BE. By increasing the demarcation between brownish squamous mucosa and unstained, pink metaplastic columnar epithelium, Lugol’s staining can increase the accuracy of endoscopic diagnosis of BE to 91%.3 Methylene blue stains actively absorbing cells such as those of intestinal metaplasia and is not absorbed by normal squamous esophageal cells. Methylene blue staining has a 95% sensitivity and GASTROINTESTINAL ENDOSCOPY
559
M Guelrud, I Herrera, H Essenfeld, et al.
Enhanced magnification endoscopy in Barrett’s esophagus
A
B
C
D
Figure 1. Short segment Barrett’s esophagus. A, Standard endoscopic view. Short extensions of columnar-appearing mucosa with squamous islands within Barrett’s epithelium. Arrowhead indicates the field to be magnified. B, Magnification endoscopic view of the same field. A villiform appearance is imperfectly observed. C, Standard endoscopic view after acetic acid instillation. The border of the squamous-columnar junction is better visualized as well as the squamous islands within the Barrett’s epithelium. Arrowhead indicates the field to be magnified. D, Enhanced magnification endoscopic view of the same field. A fine villiform appearance is clearly evident.
97% specificity for the biopsy finding of SIM.4 In two recent studies, however, methylene blue staining with targeted biopsies failed to improve the diagnostic yield for BE.5,6 Moreover, the technique is messy and substantially lengthens the procedure time. Acetic acid produces reversible intracellular cytoplasmic protein denaturation.7 Currently, it is used during colposcopy as an aid to detect small lesions in the uterine cervical mucosa, a mucosa similar to that at the squamocolumnar junction.8 It has been shown that acetic acid instillation enhances the ability to detect small or indistinct remnant islands of columnar epithelium after endoscopic treatment of BE, and that this method is safe, rapid, clean, and inexpensive.9 Standard endoscopy is limited in terms of capability to evaluate columnar-appearing epithelium because the superficial mucosal layer is visualized at 560
GASTROINTESTINAL ENDOSCOPY
low magnification. Magnification endoscopy, which involves use of a special endoscope, has been used with and without dye spraying to better evaluate the columnar epithelium in BE.10,11 Enhanced magnification endoscopy involves the combined use of magnification endoscopy and acetic acid. Our goal was to determine the value of enhanced magnification endoscopy in detecting SIM in patients with BE. MATERIAL AND METHODS Patients A total of 49 patients with previously diagnosed short segment BE without dysplasia who were being followed in an endoscopic surveillance program were included in the study. Short segment BE was defined as a length of 3 cm or less of columnar-appearing mucosa proximal to the esophagogastric junction at endoscopy with histologic conVOLUME 53, NO. 6, 2001
Enhanced magnification endoscopy in Barrett’s esophagus
A
C
M Guelrud, I Herrera, H Essenfeld, et al.
B
D
Figure 2. Endoscopic views after acetic acid instillation; 4 different patterns of the mucosal surface were observed. A, Pattern I: round pits with a characteristic pattern of regular and orderly arranged circular dots. B, Pattern II: reticular pits that are circular or oval and are regular in shape and arrangement. C, Pattern III: villous with no pits present but a fine villiform appearance with regular shape and arrangement is evident. D, Pattern IV: ridged with no pits present but a thick villous convoluted shape with a cerebriform appearance with regular shape and arrangement is evident. firmation of SIM on biopsy.12 The study protocol was approved by the ethics committee of our institution. All participating patients gave written informed consent. Magnification gastroscope All procedures were performed with a magnification videoendoscope (GIF-200Z, Olympus America Inc., Melville, N.Y.), which has the capability of both standard videoendoscopy and an adjustable image magnification in a continuous range up to ×35. The instrument has a depth of field in the wide position of 7 to 100 mm and a narrow depth of field of 4 to 11 mm. The diameter of the endoscope is 10.5 mm with a working length of 103 cm and a 2.8 mm accessory channel. Tip angulation is slightly less in the upward direction than a standard gastroscope (180° vs 210°), but the downward and left-right deflection ranges are the same. Endoscopic procedure Endoscopic procedures were performed by a single gastroenterologist (MG) experienced in the technique of VOLUME 53, NO. 6, 2001
enhanced magnification endoscopy using acetic acid instillation. Patients first underwent standard endoscopy followed by magnification endoscopy. Then 10 to 15 mL of 1.5% acetic acid were sprayed onto the distal esophagus with a spray catheter (Olympus, PW-5V). Immediately afterward, the distal esophagus was irrigated with 3 to 5 mL of water by using the lens-cleaning water channel of the endoscope. There was no attempt to wash away the excess acetic acid. In the following 2 to 3 minutes, standard endoscopy and magnification endoscopy were repeated (Fig. 1). Acetic acid procedure time was recorded from the point at which acetic acid was first sprayed to the time at which biopsy specimens were obtained. All examinations were videotaped and photographed. Frequently the endoscopic pattern was diagnosed while the image was captured on the monitor to obtain a photograph. After acetic acid instillation, 4 different mucosal surface patterns were observed (Fig. 2). Pattern I: round pits with a characteristic pattern of regular and orderly GASTROINTESTINAL ENDOSCOPY
561
M Guelrud, I Herrera, H Essenfeld, et al.
Enhanced magnification endoscopy in Barrett’s esophagus
Table 1. Frequencies of endoscopic patterns observed with each endoscopic technique
Pattern I II III IV
Number of total areas 48 18 46 17
SE N areas (%) 0 0 1 1
SE+AA
ME
EME
N areas (%)
N areas (%)
N areas (%)
(0) (0) (2) (6)
0 0 6 5
(0) (0) (13) (29)
10 7 21 11
(21) (39) (46) (65)
48 18 46 17
(100) (100) (100) (100)
SE, Standard endoscopy; AA, acetic acid; ME, magnification endoscopy; EME, enhanced magnification endoscopy.
arranged circular dots. Pattern II: reticular. Pits were circular or oval and were regular in shape and arrangement. Pattern III: villous. No pits were present but there was a fine villiform appearance with regular shape and arrangement. Pattern IV: ridged. No pits were present but there was a thick villous convoluted shape with a cerebriform appearance with regular shape and arrangement. Biopsy protocol One targeted biopsy specimen was taken from each representative pattern with a standard biopsy forceps (FB24U, Olympus) under low magnification. Occasionally, more than one biopsy sample was obtained. In all cases, a single specimen was taken from a pattern I area located distal to the esophagogastric junction. Formalin-fixed samples were stained with H&E. Additional alcian blue staining at pH 2.5 was performed to detect SIM, which was defined by the presence of goblet cells. Biopsy specimens were considered positive for SIM even if they contained only minute, localized SIM. Two pathologists blinded to the method of obtaining the sample independently examined all specimens. Results of histologic examinations were compared with endoscopic findings. Statistical analysis Histologically confirmed SIM was used as the main outcome, and univariate logistical regression was used to calculate the odds ratios (OR) in enhanced magnification endoscopy, using pattern I as reference group. Sensitivity, specificity, positive predictive value, negative predictive value, and ROC (receiver operator characteristic) were calculated for the enhanced magnification technique to assess the yield of the model. A model with no predictive power has an area of 0.5; a perfect model has an area of 1. All values were calculated by using 0.05 significance criteria and 2 tails. STATA 5.0 (Stata Corp., College Station, Tex.) statistical software was used to perform the analysis.
RESULTS Patients Between August 1999 and November 1999, 49 patients with short segment Barrett’s esophagus were studied. One patient was excluded because of unclear definition of the surface pattern after acetic acid instillation. Of the 48 patients, 9 (19%) were women and 39 (81%) were men with a mean age of 50.5 years (range 24–76). Barrett’s epithelium was 562
GASTROINTESTINAL ENDOSCOPY
present in tongue-like or circumferential-type patterns. Twenty-one of the 48 patients were being treated long-term with lansoprazole 30 mg every day; 10 patients took this drug irregularly. The other 17 patients were not taking anti-secretory medication. No adverse clinical events were caused by the acetic acid instillation procedure. It was estimated that the technique added between 5 to 8 minutes to the length of the examination. Endoscopic patterns In the first 2 to 3 minutes after acetic acid spraying, there was a whitish discoloration of the esophageal mucosa, the BE, and the gastric mucosa. Thereafter the esophagus remained white and the columnar epithelium of both the BE and gastric epithelium became reddish. A total of 129 areas from 48 patients were examined. Table 1 shows the percentage of endoscopic patterns observed with each endoscopic technique. Standard endoscopy identified an endoscopic pattern in 1.5% (2 of 129) of the areas. When acetic acid was added, standard endoscopy increased the detection yield to 8.5% (11 of 129) areas. When magnification endoscopy alone was used, the detection yield increased to 49 of 129 areas (38%). In contrast, enhanced magnification endoscopy allowed recognition and classification of the 129 endoscopic patterns. Table 2 shows the yield of detecting SIM according to the endoscopic pattern observed with enhanced magnification endoscopy. In 48 biopsy samples obtained from the same number of endoscopic areas with pattern I located in the stomach (hernia sac) distal to the esophagogastric junction, all had fundic columnar epithelium without SIM. This endoscopic pattern served as a control group for the analysis. Frequently, small areas of pattern III or pattern IV were observed within a pattern I area. In order to have a homogeneous group of pattern I areas, these areas were avoided as sites for performing biopsies. Analysis showed correct identification of SIM in 11% (odds ratio 0.5, p = 0.54) of the biopsy specimens of pattern II, in 87% (odds radio 36, p = 0.001) of pattern III, and in 100% (odds ratio 14, p = 0.015) of pattern IV. VOLUME 53, NO. 6, 2001
Enhanced magnification endoscopy in Barrett’s esophagus
By using histology as the main outcome (Table 3), the sensitivity, specificity, positive predictive value, and negative predictive value of enhanced magnification endoscopy for patterns III and IV for detecting SIM were 96.5%, 88.7%, 87.5%, and 96.9%, respectively. The overall accuracy was 92.2% with an ROC of 0.92. DISCUSSION The enhanced magnification endoscopy technique revealed characteristic patterns in SIM and identified SIM with a high degree of accuracy. The fine surface structure of Barrett’s epithelium has a characteristic villous pattern, regular in shape and orderly in arrangement, that suggests a histologic diagnosis of intestinal metaplasia. Currently, the endoscopic diagnosis of BE is based entirely on gross changes in the mucosal appearance. Endoscopic examination usually can distinguish the pale, glossy squamous epithelium of the esophagus from the red, velvet-like columnar epithelium of BE. However, the endoscopic appearance of the fine surface structure of BE is too subtle to be detected with a standard endoscope. A normal mucosal appearance does not necessarily imply normal histology. Thus the different types of columnar epithelium can be distinguished only by histologic examination of tissue specimens.13 Because metaplasia is patchy, the use of random biopsies to assess the presence of SIM in the columnar epithelium is likely to suffer from sampling error. To avoid this type of error, multiple biopsy specimens are usually obtained. Even with this approach, BE was detected in follow-up examinations in only 80% of patients 6 weeks after the initial examination.14 Moreover, inaccurate biopsy targeting may worsen diagnostic accuracy. It can be difficult to precisely obtain biopsy samples given the tangential orientation of the wall in relation to the forceps. The use of regular forceps might sample more mucosa than desired, leading to heterogeneity of different types of columnar epithelium within a single sample. Careful endoscopic technique with targeted biopsies can prevent this problem. To improve the diagnosis of BE, chromoendoscopy, acetic acid instillation, and magnification have been used. Lugol’s solution has been used to define the squamocolumnar junction accurately,15 to identify areas of possible BE,16 and to improve the diagnosis of BE.3 It also has been used in combination with magnification endoscopy and indigo carmine instillation to diagnose the villiform appearance of BE.11 However, by using two different staining methods, the procedure is prolonged and more troublesome. Moreover, severe allergic reacVOLUME 53, NO. 6, 2001
M Guelrud, I Herrera, H Essenfeld, et al.
Table 2. Yield of detecting SIM according to endoscopic pattern Pattern
SIM present
I II III IV
0/48 2/18 40/46 17/17
(0%) (11%) (87%) (100%)
Odds ratio
p
– 0.5 36 14
– 0.543 0.001 0.015
Statistical analyses were performed comparing results from pattern I (control) with patterns II, III, and IV. SIM, Specialized intestinal metaplasia.
tions have been reported with Lugol’s solution.11 Unlike Lugol’s solution, methylene blue selectively stains specialized columnar cell.4 The technique involves application of a mucolytic agent, followed by dye application and then washing of excess dye. Positive staining is defined as the presence of bluestained mucosa that persists despite vigorous water irrigation. In a prospective control study methylene blue revealed SIM in 18 of 26 patients with a 95% sensitivity for the biopsy finding of SIM.4 These results have been questioned in two recent studies. One study revealed SIM in 54% of focally stained mucosa and in 63% of random biopsy samples.5 A second study found SIM in 74% of methylene bluestained mucosa compared with 69% of random biopsy samples.6 Most important, in the latter study, the investigators found that if biopsy specimens were not obtained for unstained areas, SIM would have been missed in 6 of 21 patients. Besides adding significant time to the endoscopic procedure, methylene blue staining is somewhat subjective. It is influenced by the washing technique used. Inadequate lavage after staining may cause false-positive staining of gastric surface cells. Inadequate exposure of the BE to the mucolytic agent and methylene blue can cause SIM to appear pink, resulting in a falsepositive indication of gastric mucosa. Acetic acid has been used to improve detection of small remnant islands of Barrett’s epithelium after ablative therapy.9 In 11 of 21 (52%) patients in one study, acetic acid demonstrated small remnant islands of columnar epithelium not seen before acid instillation. However, the pattern of the columnar epithelium could not be demonstrated, making it impossible to differentiate gastric epithelium from BE. A better method of endoscopic diagnosis would be one that displays microscopic surface morphology during actual endoscopic observation. Magnification endoscopy has been used for several purposes.17 This is the first report of the use of enhanced magnification endoscopy, which combines magnification endoscopy and acetic acid instillation, in BE. Enhanced magnification endoscopy with acetic acid allows visualization of actual villi and crypt areas similar to those GASTROINTESTINAL ENDOSCOPY
563
M Guelrud, I Herrera, H Essenfeld, et al.
Enhanced magnification endoscopy in Barrett’s esophagus
Table 3. Statistical performance according to the endoscopic technique Technique SE SE+AA ME EME
Sensitivity
Specificity
PPV
NPV
Total accuracy
ROC
0 18.9 41.3 96.5
100 97.1 97.1 88.7
– 84.6 92.3 87.5
55 59.4 66.9 96.9
55 62 72 92.2
0.50 0.58 0.69 0.92
PPV, Positive predictive value; NPV, negative predictive value; ROC, receiver operating characteristic; SE, standard endoscopy; AA, acetic acid; ME, magnification endoscopy; EME, enhanced magnification endoscopy.
observed with a stereoscopic microscope. Our analysis of enhanced magnification endoscopy revealed that SIM was characterized by two patterns of surface features. These patterns could be clearly visualized by enhanced magnification endoscopy. Surface features characterized by the absence of round pits and the presence of fine villous or ridged (cerebriform) patterns are suggestive of SIM. The presence of circular dots or round pits is characteristic of gastric columnar epithelium. Our histologic examination was consistent with this hypothesis. Enhanced magnification endoscopy achieved a positive predictive value of 87.5% and a total accuracy of 92.2% in the detection of SIM for pattern IV and a specificity of 100% for the detection of fundic columnar epithelium for pattern I. This suggests that this technique might have the potential to achieve accuracies comparable to those of clinical histopathology. The detection rate of 11% of SIM in the reticular pattern II could be explained by the use of a regular forceps that may have sampled more mucosa than desired, including some metaplastic epithelium. The use of smaller biopsy forceps for more precise targeting might overcome this problem. In this regard, the villous pattern of the surface mucosa may be a clue to the presence of SIM in patients with BE or in patients with intestinal metaplasia of cardia. It may explain why the intestinal metaplasia of the cardia is not more common in patients with BE than in control subjects.18,19 It is also possible that if targeted biopsy specimens are obtained, the prevalence of intestinal metaplasia at the gastric cardia would be higher than previously reported. Moreover, our finding of small areas of intestinal metaplasia within normal gastric mucosa in the hiatal hernias might explain the conflicting results of the association between intestinal metaplasia of cardia and intestinal metaplasia elsewhere in the stomach observed in some studies in which biopsy samples were taken randomly.20-22 The potential advantage of enhanced magnification endoscopy with acetic acid instillation is that targeting biopsies can decrease the sampling error and the number of biopsy samples needed for diagnosis. The technique is simple, fast, not messy, and highly accurate. Given its high sensitivity, specificity, and predictive values, enhanced magnification endoscopy may 564
GASTROINTESTINAL ENDOSCOPY
be useful for a variety of clinical applications, such as supplementing visual assessment of BE discovered during upper endoscopy, targeting of biopsies when short segment BE or intestinal metaplasia of cardia is suspected, guiding ablative therapy, recognizing residual metaplastic mucosa after ablative therapy without obtaining biopsy specimens, and differentiating esophagitis and regenerative squamous epithelium from columnar epithelium. However, prospective studies with enhanced magnification endoscopy in these particular applications need to be performed before the technique can be recommended for routine use. Our study included data obtained from patients with known BE. Patterns were classified by visual appearance. Because the mucosal effect of the acetic acid lasts only 2 to 3 minutes, the esophagogastric junction is constantly moving (because of respiration, cardiac function, and secondary esophageal contractions caused by air insufflation), and the instrument has a short focal range, the endpoint of enhanced magnification endoscopy is somewhat subjective and is the most difficult part of the procedure to learn. Expertise can be achieved in approximately 10 procedures by experienced endoscopists. Moreover, and for the same reason, there is a chance of sampling error. Nonetheless there did not appear to be a strong correlation between misclassification by enhanced magnification endoscopy and SIM by histology. Enhanced magnification endoscopy does seem to be a useful aid to enhancing tissue diagnosis. However, the present study had significant limitations in that it was an open study, which introduces the potential for bias, it did not include a control group, and interobserver agreement, which could lead to problems with generalization, was not assessed. Thus, a randomized blinded study will be needed to confirm our results. In conclusion, simplicity, the ability to identify characteristic endoscopic patterns with outstanding clarity and resolution that correlate with histologic identification of SIM, makes enhanced magnification endoscopy an excellent method for the evaluation of patients with BE and possibly also for patients with intestinal metaplasia of the gastric cardia. VOLUME 53, NO. 6, 2001
Enhanced magnification endoscopy in Barrett’s esophagus
ACKNOWLEDGMENT We thank Lawrence J. Brandt, MD, for reviewing the manuscript. REFERENCES 1. Paull A, Trier JS, Dalton MD, Camp RC, Loeb P, Goyal RK. The histologic spectrum of Barrett’s esophagus. N Engl J Med 1976;295:476-80. 2. Spechler SJ, Goyal RK. The columnar-lined esophagus, intestinal metaplasia and Norman Barrett. Gastroenterology 1996;110:614-21. 3. Woolf GM, Riddell RH, Irvine EJ, Hunt RH. A study to examine agreement between endoscopy and histology for the diagnosis of columnar lined (Barrett’s) esophagus. Gastrointest Endosc 1989;35:541-4. 4. Canto MI, Setrakian S, Petras RE, Blades E, Chak A, Sivak MV Jr. Methylene blue selectively stains intestinal metaplasia in Barrett’s esophagus. Gastrointest Endosc 1996;44:1-7. 5. Jobson B, Goenka P, Manalo G, Thomas E. Methylene blue staining for intestinal metaplasia in Barrett’s esophagus—is it as good as we think? [abstract]. Gastrointest Endosc 1999;49:AB15. 6. Ray MB, Mayfield-Stokes S, Cecil B, Forouzandeh B, Davis A, Wilson MA, et al. Results of methylene blue-directed biopsy is similar to conventional biopsy for the diagnosis of intestinal metaplasia and dysplasia in Barrett’s esophagus [abstract]. Gastrointest Endosc 1999;49:AB22. 7. Cartier R. Practical colposcopy. Basel, Switzerland: Karger; 1978. 8. Cartier R. Role de la colposcopie dans le diagnostic et le traitement des dysplasies et des carcinomes intra-epitheliaux du col uterin. Bull Cancer 1979;66:447-54. 9. Guelrud M, Herrera I. Acetic acid improves identification of remnant islands of Barrett’s epithelium after endoscopic therapy. Gastrointest Endosc 1998;47:512-5. 10. Takemoto T, Sakaki N. Magnification endoscopy. In: Sivak Jr MV, editor. Gastroenterological endoscopy. Philadelphia: WB Saunders; 1999.
VOLUME 53, NO. 6, 2001
M Guelrud, I Herrera, H Essenfeld, et al.
11. Stevens PD, Lightdale CJ, Green PH, Siegel LM, GarciaCarrasquillo RJ, Rotterdam H. Combined magnification endoscopy with chromoendoscopy for the evaluation of Barrett’s esophagus. Gastrointest Endosc 1994;40:747-9. 12. Spechler SJ, Zeroogian JM, Antonioli DA, Wang HH, Goyal RK. Prevalence of metaplasia at the gastro-oesophageal junction. Lancet 1994;344:1533-6. 13. Weinstein WM, Ippoliti AF. The diagnosis of Barrett’s esophagus: goblets, goblets, goblets. Gastrointest Endosc 1996;44:91-4. 14. Kim SL, Waring JP, Spechler SJ, Sampliner RE, Doos WG, Krol WF, et al. Diagnostic inconsistencies in Barrett’s esophagus. Gastroenterology 1994;107:945-9. 15. Northmann BJ, Wright JR, Schuster MM. In vivo vital staining as an aid to identification of esophagogastric mucosal junction in man. Am J Dig Dis 1972;17:919-24. 16. Burbige EJ, Radigan JI. Characteristics of the columnar-cell lined (Barrett’s) esophagus. Gastrointest Endosc 1979;25: 133-6. 17. Tada M, Kawai K. Research with the endoscope: new techniques using magnification and chromoscopy. Clin Gastroenterol 1986;15:417-37. 18. Morales TG, Bhattacharyya A, Johnson C, Sampliner RE. Is Barrett’s esophagus associated with intestinal metaplasia of the gastric cardia? Am J Gastroenterol 1997;92:1818-22. 19. Weston AP, Krmpotich PT, Cherian R, Dixon A, Tapalovski M. Prospective evaluation of intestinal metaplasia and dysplasia within the cardia of patients with Barrett’s esophagus. Dig Dis Sci 1997;42:597-602. 20. Morales TG, Sampliner RE, Bhattacharyya A. Intestinal metaplasia of the gastric cardia. Am J Gastroenterol 1997;92: 414-8. 21. Hackelsberger A, Gunter T, Schultze V, Manes G, DominguezMuñoz JE, Roessner A, et al. Intestinal metaplasia at the gastro-oesophageal junction: Helicobacter pylori gastritis or gastro-oesophageal reflux disease. Gut 1998;43:17-21. 22. Goldblum JR, Vicari JJ, Falk GW, Rice TW, Peek RM, Easley K, et al. Inflammation and intestinal metaplasia of the gastric cardia: the role of gastroesophageal reflux and H pylori infection. Gastroenterology 1998;114:633-9.
GASTROINTESTINAL ENDOSCOPY
565