| Abstract|| |
It's still controversial whether certain macroscopic endoscopic features can be used to diagnose Helicobacter pylori (HP) related gastritis. The literature dealing with this subject is confusing, because of the lack of precise terminology, no large control trials, major discrepancies in interpretations of macroscopic changes and poor correlation of the macroscopic appearance and histological finding of gastritis. We conducted a prospective study of 208 dyspeptic patients, who underwent upper gastrointestinal endoscopies from February 1997 to June 1997. Only those patients who had either normal looking gastric mucosa or macroscopic gastritis were included in the study. Endoscopically normal looking mucosa was seen in 67 patients (65.6%), erythematous gastritis in 51 (74%), mosaic appearance in 18 (88%), erosive gastritis in 14 (85%), nodular gastritis in 17 (94%), atrophic gastritis in 12 (75%), and fundal rugae hypertrophies in 5 (80%). We suggest that the antral nodularity, raised erosions, mosaic appearance and mixed findings, are the reliable indicators of the underlying HP induced gastritis. However, these endoscopic findings are very specific, though not sensitive, for HP gastritis.
|How to cite this article:|
Khan MQ, Alhomsi Z, Al-Momen S, Ahmad M. Endoscopic features of Helicobacter pylori induced gastritis. Saudi J Gastroenterol 1999;5:9-14
There is a great deal of ignorance about the normal macroscopic appearance of the gastric mucosal lining. In all probability many of the appearance which endoscopists interpret as "Normal" are presumably not normal , . Endoscopic findings such as erythema are frequently labelled as gastritis despite a long recognized lack of evidence supporting a correlation between endoscopic features and histologic gastritis  .
The endoscopic appearances of the stomach in acute infectious phases of H. pylori-induced inflammation are poorly known. An occasional patient has been described in the past  with a strikingly red, friable and slightly erosive gastric mucosa, covered with abundant fibrinopurulent exudate in the antrum. Rocha et al  described seven patients with H. pylori induced acute gastritis who presented endoscopic picture simulating gastric carcinoma or lymphoma. The endoscopic examination revealed a very similar appearance in all patients, varying slightly only in the intensity, including edema and multiple ulceration with irregular edges surrounded by red and friable mucosa in the antrum. We believe that there are certain subtle changes in gastric mucosa which can be used to predict the presence of HP and the purpose of this study is to delineate this.
In Saudi Arabia 74% cases of non ulcer dyspepsia were found to be H. pylori positive  and 68% of gastroduodenitis showed the histological evidence of H. pylori  . There are some reports on various endoscopic findings and the presence of H. pylori in Saudi Arabia ,,, . In this study, we prospectively evaluated the antral biopsies from patients presenting with mild to severe symptoms of dyspepsia, in an attempt to determine the role of endoscopic appearance and the possibility of yet another diagnostic approach for H. pylori induced gastritis.
| Patients and Methods|| |
A total of 208 dyspeptic adult patients, who underwent diagnostic upper gastrointestinal endoscopy from February 97 to June 97 at Dammam Central Hospital, were considered to be included in this study. The patients suffered from dyspepsia if they had epigastric pain, heartburn, nausea, vomiting, abdominal distention, flatulence, erection, fullness after meals, regurgitation and early satiety, regardless of the severity of dyspeptic symptoms. According to our protocol, the inclusion criteria were endoscopically normal looking gastric mucosa or macroscopic gastritis. The exclusion criteria were, the patients with history of H. pylori infection or received antibiotics for H. pylori during last one year, or used NSAID for the past two months. After endoscopy the patients of gastric ulcers, duodenal ulcers, duodenitis, reflux esophagitis were also excluded from the study, because, in the present study, the interest was to find out the H. pylori induced macroscopic features of endoscopic gastritis. However, endoscopically normal looking gastric mucosa was included in the study for comparison as control group.
Endoscopy was performed with an Olympus GIFQ20 endoscope after local anesthesia of the oropharynx, and occasionally after intravenous injection of diazepam. A full endoscopic evaluation of the esophagus, stomach, and duodenum was performed in all subjects by a single endoscopist according to a set protocol. Each endoscopy was recorded into on VHS videotapes, and reviewed by another senior endoscopist independently for the consensus opinion. Before review of these tapes, the three endoscopist together observed numerous endoscopies to choose specific gastic features that they could identify regularly. They then agreed on definitions and appearances of these features to develop consensuses.
| Diagnosis|| |
We used histologic examination for diagnosis of H. pylori in this study because this is generally considered to be the "Gold Standard" for diagnosis. Although, some degree of patchiness has been noted in the distribution of H. pylori in the stomach on histological examination. Genta & Graham  also reported that fewer than 9% of biopsy specimens from any one site in the stomach are negative in patients with documented H. pylori infection. In patients with abnormal mucosal appearance, four to five antral biopsies were taken by the standard 7 FG seized biopsy forceps from the suspicious area of antral gastritis, but in normal looking gastric mucosa, biopsies were taken 2 to 4 cm away from the pylorus. The biopsies were fixed in neutral buffered formalin, routinely processed and stained with eosin and hematoxylin. Identification of H. pylori by light microscopy is facilitated with modified Giemsa's stain. The biopsies were examined by three different pathologists for a consensus and the presence of H. pylori was reconfiiuned.
Normal / abnormal mucosa
The patients were divided into two groups, according to the macroscopic appearance of normal looking and abnormal looking mucosa. The gastric mucosal appearance is called "normal" when the color is an even shade of pink, and there are uniform smoothness and luster throughout the mucosa. The antrum usually appears flat, or shows only slightly elevated prepyloric folds with adequate but not excessive insufflation of air. Less commonly one or two roofing folds can be seen, forming arches over the pyloric channel or a few more prominent antro-pyloric folds are present. The folds or rugal pattern in the corpus-fundus area is regular and even, not exceeding 5 mm in cross-diameter. The "abnormal" looking endoscopic features were chosen according to modified Sydney system of classification, which include erythema/exudation, erosion (raised or flat), mosaic pattern or cobble stoning, hypertrophic rugae, atrophic (thinning of the mucosa accompanied by visibility of ramifying vessels) and nodular appearance [Table - 1].
Asssociation of H. pylori test results with endoscopically normal and abnormal looking gastric mucosa was calculated by Chi Square (X 2 ) tests. The comparison of a normal group with an abnormal group in general, and normal group with specific endoscopic abnormalities like erythematous, nodular, mosaic, atrophic and erosive mucosa in particular, was also correlated for the statistical significance.
| Results|| |
Out of 208 subjects who were included in this study, 127 (61%) were women and 81 (39%) men. Of this 181 (87%) patients were Saudis and 27 (13%) were nonSaudis. Their age ranged from 14 to 82, with a mean of 39 ± 11 (mean ± SD). In total 159 (76%) were positive for H. pylori and 49 (24%) negative.
The most common dyspeptic symptom was epigastric pain, followed by heart burn, flatulence, nausea, eruction, fullness after meals, regurgitation, early satiety and vomiting respectively.
Endoscopically abnormal looking gastric mucosa was seen in 141 patients (58%), H. pylori was positive in 115 (82.1%) and negatives in 26 (18.4%). Endoscopically normal looking gastric mucosa was seen in 67 (27.5%), H. pylon was positive in 44 (65.6%), and negatives in 23 (34.3%). Single endoscopic finding was seen in 117 patients (82.9%) and mixed in 24 (17.1%) Erythematous gastritis was seen in 51 patients (74%), Cobblestone and/or mosaic appearance in 18 (HP 88%), erosive gastritis in 14 (85%), nodular gastritis in 17 (94%), atrophic gastritis in 12 (75%), fundal rugae hypertrophy in five (80%) and mixed findings in 24 (HP 83.5%) as shown in [Table - 2].
The diagnosic value (sensitivity, specificity and positive predictive value) of macroscopically abnormal gastric mucosa is shown in [Table - 3]. Subjects with nodular, mosaic or erosive pathology showed specificity of 96%, 92% and 92% respectively as compared to sensitivity which was 28%, 27% and 22% respectively.
Mixed findings were found in 24 patients (17.1 %). In two abnormal findings the prevalence of H. pylori was 80%, but with three or four mixed findings, there were 100 positive results of H. pylori [Table - 4].
| Discussion|| |
In many patients with chronic H. pylori infection, the endoscopic abnormalities of erythematous and exudative gastritis consisted of patches of erythema alternating with paler areas, some loss of luster, slight unevenness of the mucosal lining especially of the antral area, and tiny punctate whitish spots of exudates  . We found very poor agreement on the finding of erythema, the feature most commonly labelled as gastritis. We observed that the erythematous gastritis is the commonest single endoscopic finding (43%) but H. pylori was present in only 74% of cases, which was insensitive and statistically insignificant (p<0.2) as predictors of H. pylori infection, when compared with normal mucosa (66% HP+ve).
Antral nodularity has been reported to be a common and specific finding in pediatric patients with H. pylori infection and may be associated with microscopic evidence of lymphoid nodular hyperplasia, previously considered being peculiar to H. pylori infection in children ,,,,, . Antral nodularity has recently been suggested as a potential endoscopic feature of H. pylori infection in adults as well  . We therefore included nodularity in our study, and observed that the most reliable endoscopic indicator of H. pylori infection is nodular gastritis, as the specificity was 96% and a high positive predictive value reached up to 95%. Laine L, et al  observed that antral nodularity had a high specificity for H. pylori (96%) with positive predictive value of 90%. He also suggested that, a combination of both antral nodularity and prominent body area gastricae, although not common, gave a specificity and positive predictive value of 100%. Sbeih, et al  described that all of this 25 patients with endoscopic antral nodularity had H. pylori in their antral biopsy specimens (100%). Recently, Grellier L, et al  suggested that the nodularity of gastric antrum is a specific marker of mucosal colonization by H. pylori with the sensitivity of 74.1 % and specificity of 100%. He advised that in the presence of antral nodularity H. pylori colonization may be assumed, thus avoiding the need for an additional biopsy. We don't agree with this information because, in our study, nodularity has a very low sensitivity (28%), so it would not be acceptable to be used as a screening test. Sensitivities for the standard diagnostic test for H. pylori (serelogy, urea breath test, rapid gastric biopsy urease test and histological examination) range from approximately 80-100%  . However, antral nodularity, showed specificity of 96% for H. pylori infection certainly matches the specificity of any of the diagnostic test used to diagnose H. pylori.
Stolte &Edit  concluded that chronic erosion of the antral mucosa represent a sequela of H. pylori gastritis, and that these H. pylori-induced chronic erosion should, in future would be differentiated from other erosions. We demonstrated that the antral erosion showed the specificity of 92% and sensitivity of 22% with positive predictive value of 86% to H. pylori infection. We also conclude that the raised erosions were more precisely associated with H. pylori infection than flat erosions.
Atrophic gastritis is diagnosed  when the vascular pattern becomes visible in the nonoverdistended stomach with some pearly whitish mucosal discolouration and fold atrophy of varied degree. Sakaki  demonstrated the utility of categorising endoscopic findings related to H. pylori infection was found to be closely related to atrophic changes. In our study atrophic gastritis was an uncommon finding (10%), but for H. pylori positive predictive value was 75% and specificity of 89%. The difference was not statistically significant when compared with normal looking mucosa.
Giant gastric folds or fundal rugae hypertrophy is present when folds do not flatten on insufflation of air and the thickness of the folds is approximately 515 mm. Hyper rugosity is associated with H. pylori  infection, when the fold thickness is more than 5 mm. Mond et al  indicated that enlarged gastric folds on the upper GI barium series of a symptomatic adult are very suggestive of H. pylori gastritis. In our study the number of cases of hypertrophic gastric folds is not enough to come out with statistically significant inference.
Cobblestone/mosaic appearance has not been described as a separate entity of gastritis in Sydney classification  , but it has been considered as a specific feature of H. pylori induced gastritis  . In our study, it showed the specificity (p<0.05) values when compared with normal looking gastric mucosa.
Mixed findings  have been described in H. pylori induced gastritis with significant predictive values. It looks obvious that if one finding could give rises some higher statistically significant values, then more than two findings would definitely reach an almost 100% predictive values  of H. pylori infection.
Bah A, et al  , and others  recently concluded that it is not possible to diagnose H. pylori gastritis on the bases of the endoscopic appearance alone. The diagnosis should be based on other criteria, such as a rapid urease test or a histological examination of gastric biopsies or both. The number of cases in this study was very small, and the conclusion should not be accepted as such, until further large control trial published.
We observed that none of the endoscopic features assessed, had a good sensitivity for the diagnosis of H. pylori infection, so none of these features would be acceptable for use as screening test, because 65.5% of H. pylori were present in normal looking gastric mucosa. However, the H. pylori infection certainly matches the specificity of any of the diagnostic tests used to diagnose H. pylori. We conclude that at least four endoscopic features are associted with H. pylori related gastritis, these are nodularity, erosion, mosaic appearance and mixture of these findings. When we tested their sensitivity, specificity and positive predictive value against normal mucosa, the results were statistically significant.
| References|| |
|1.||Fung WP, Papadimitriou JM and Matz LR. Endoscopic, histological and ultrastructural correlation in chronic gastritis. Amer J Gastroent 1979;71:269-79. |
|2.||Taor RE, Fox B, Ware J and Johnson AG. Gastritis: gastroscopic and microscopic. Endosc 1975;7:209-15. |
|3.||Sauerbruch T, Schreiber MA, Schussler P and Permanetter W. Endoscopy in the diagnosis of gastritis. Diagnostic value of endoscopic criteria in relation to histological diagnosis. Endosc 1984;16:101-4. |
|4.||Salmeron M, Desplaces N, Lavergne A and Houdart R. Campylobacter-like organism and acute purulent gastritis. Lancet 1986;ii:975-6. |
|5.||Roclia GA, Queiroz DM, Mendes EN, et al. Helicobacter pylori acute gastritis: Histological, endoscopical, clinincal and therapeutic features. Am J Gastroenterol 1991;86:1592-5. |
|6.||Darwish AA. Association of Helicobacter pylori and non ulcer dyspepsia. Annals of Saudi Medicine 1992;12:83-7. [PUBMED] |
|7.||Satti MB, Twum-Danso K, Al-Freihi HM, et al. Helicobacter pylori associated upper gastrointestinal disease in Saudi Arabia. A pathologic evaluation of 298 endoscopic biopsies from 201 consecutive patients. Am J of Gastroenterol 1990;85:527-34. |
|8.||Al-Freihi HM, Al-Quorain A, Al-Gindan Y, et al. Compylobacter pylori in Saudi patients undergoing upper gastrointestinal endoscopy. Hepato-gastro-enterology 1989;36:516-8. |
|9.||Al-Moagel MA, Evans DG, Abdulghani ME, et al. Prevalence of Helicobacter (formerly known as Compylobacter) pylori infection in Saudi Arabia and comparison of those with and without upper gastrointestinal symptoms. Am J Gastroenterol 1990;85:527-34. |
|10.||Sbeih F, Abdullah A, Sullivan S, et al. Antral nodularity, gastric hyperplasia, and H. pylori in adult. J Clin Gastroenterol 1996;22:227-30. |
|11.||Mohamed AE, Al-Karawi MA, Al-Jumah AA, et al. Helicobacter pylori: Prevalence in 352 consecutive patients with dyspepsia. Ann Saudi Med 1994;14:134-5. [PUBMED] |
|12.||Genta RM, Graham DY. Comparison of biopsy sites for the histopathologic diagnosis of H. pylori: A topographic study of H. pylori density and distribution. Gastrointst Endoscopy 1994;40:342-5. |
|13.||Tytgat GNJ, The Sydney system: Endoscopic division. Endoscopic appearances in gastritis/duodentitis. J Gastroenterol Hepatol 1991;6:223-4. |
|14.||Drumm B, Sherman P, Cutz E, et al. Association of Compylobacter pylori on the gastric mucosa with antral gastritis in children. N Engl J Med 1987;316:1557-61. |
|15.||Czinn SJ, Dahms BB, Jacobs GH, et al. Campylobacter like organism in association with symptomatic gastritis in children. J Pediatr 1986;109:80-3. [PUBMED] |
|16.||Oderda G, Lerro P, Poli E, et al. Childhood nodular gastritis and Campylobacter pylori. Endosc 1988;20:86. |
|17.||Cadranil S, Goossens H, De Boeck M, et al. Campylobacter pyloridis in children. Lancet 1986;1:735-6. |
|18.||Bujanover Y, Konikoff F, Baratz M. Nodular gastritis and Helicobacter pylori. J Pediatr Gastroenterol Nutr 1990;11:41-4. [PUBMED] |
|19.||Hassal E, Dimmick JE. Unique features of Helicobacter pylori disease in children Dig Dis Sci 1991;36:417-23. |
|20.||Labenz J, Gyenes E, Ruhl GS, et al. Is Helicobacter pylori gastritis a macroscopic finding ? Dtsch Med Wochenschr 1993;118:176-80. |
|21.||Lame L, Cohen H, Sloane R, et al. Interobserver agreement and predictive value of endoscopic findings for H. pylori and gastritis in normal volunteers. Gastrointest Endosc 1995;42:420-3. |
|22.||Grellier L, Tanner P, Grainger SL. Anti-al nodularity: Macroscopic marker for Helicobacter pylori gastritis. Gut I993;34(suppl):S35. |
|23.||Brown KE, Peura DA. Diagnosis of Helicobacter pylori infection. Gastroenterol. Clin North Am 1993;22:105-15. |
|24.||Stolte M and Eidt S. Chronic erosions of the antral mucosa: A sequelae of H. pylori-induced gastritis. Z-Gastroenterol 1992;30:846-50. |
|25.||Sakaki N. Endoscopic diagnosis of gastritis in relation to Helicobacter pylori infection and subjective symptoms. J Clin Gastroenterol 1995;21:suppl:S135-9. |
|26.||Herz R, Lombardi E, Wipping F, et al. Helicobacter pylori associated hypertrophic gastritis-imitation of Menetrier's disease. Fortschr-Med 1992;110:37-40. [PUBMED] |
|27.||Mond DJ, Pochaczevsky R, Vemace F, et al. Can the radiologist recognize Helicobacter pylori gastritis ? J Clin Gastroenterol 1995;20:199-202. |
|28.||Ho YY, Daskalopoulos G. Can you diagnose Helicobacter pylori by macroscopic appearance on endoscopy ? (Phase I trial) Gut 1995;Supp1:4C38. |
|29.||Knox TA, Mueller JD. Diagnosis of Helicobacter pylori gastritis: Comparision of CLO test, histology, and endoscopic findings. Gastroenterol 1991;100:A99. |
|30.||Bah A, Armstrong D, Vouillamoz D, et al. Endoscopic features of f(elicobacterpylori gastritis. Endosc 1995;27:593-6. |
|31.||Cutler A, AlKhatibe T, Schubert T. Correlation of antral mucosal appearance with histology and Helicobacter pylori. Gastrointes Endosc 1992;38:285-6. |
Mohammed Qaseem Khan
Consultant Gastroenterologist, Division of Gastroenterology, P.O. Box 4517, Dammam 31412
Source of Support: None, Conflict of Interest: None
[Table - 1], [Table - 2], [Table - 3], [Table - 4]