Helicobacter Pylori Is Not Associated with the Manifestations of Gastroesophageal Reflux Disease

PAPER Helicobacter pylori Is Not Associated With the Manifestations of Gastroesophageal Reflux Disease

Stefan O¨ berg, MD; Jeffrey H. Peters, MD; John J. Nigro, MD; Jo¨rg Theisen, MD; Jeffrey A. Hagen, MD; Steven R. DeMeester, MD; Cedric G. Bremner, MD; Tom R. DeMeester, MD

Hypothesis: Helicobacter pylori is not associated with gus and GEJ) were correlated with the presence of H gastroesophageal reflux disease and its complications, in- pylori. cluding adenocarcinoma of the esophagus and the gastroesophageal junction (GEJ). Results: Helicobacter pylori was found on the biopsy specimens of the gastric antrum in 14.0% (32/229) of the pa- Design: Retrospective analysis. tients with benign disease. It was not related to the features of gastroesophageal reflux disease, including abnormal Setting: University tertiary referral center. esophageal acid exposure, erosive esophagitis, or Barrett esophagus. The presence of inflamed cardiac mucosa at the Patients: Two hundred twenty-nine patients with GEJ or carditis was inversely related to H pylori infection symptoms suggestive of foregut disease underwent and strongly associated with increased esophageal acid ex- esophageal manometry, 24-hour pH monitoring, and posure. There was no association between the presence of upper gastrointestinal tract endoscopy, with biopsy intestinal metaplasia and H pylori infection. Helicobacter py- specimens obtained from the gastric antrum, the GEJ, lori was found in 22 (19.3%) of the 114 patients with esoph- and the distal esophagus. In these and in an additional ageal adenocarcinoma, which was not different from the 114 patients with adenocarcinoma of the esophagus prevalence of H pylori in patients with benign disease. and the GEJ, the presence of H pylori was determined by Giemsa stain. The presence of gastroesophageal Conclusion: Helicobacter pylori plays no role in the patho- reflux disease, defined by abnormal esophageal acid genesis of gastroesophageal reflux disease or its compli- exposure, and its manifestations (carditis, erosive cations. esophagitis, intestinal metaplasia limited to the GEJ, Barrett esophagus, and adenocarcinoma of the esopha- Arch Surg. 1999;134:722-726

HE ISOLATION of Helicobac- tis, Barrett esophagus, and adenocarci- ter pylori from gastric mu- noma of the esophagus and the gastro- cosa by Marshall and War- esophageal junction (GEJ), have increased ren1 in 1983 dramatically dramatically during the same period.2,5,6 altered our understanding The role of H pylori in the pathogenesis of the pathophysiological characteristics of GERD and its complications has not T 7,8 of acid peptic disease. Helicobacter pylori been fully evaluated. Reports to date have is accepted as an important factor in the been conflicting, with some authors ar- development of duodenal ulcer and dis- guing an important role for H pylori in tal gastric cancer. It is also true that, in GERD and others arguing that it does not most developed countries, the preva- play a role. Recent data suggest that H py- lence of these disorders has dropped dra- lori may actually protect against the de- matically during the past several development of erosive esophagitis,9,10 Barrett cades.1-4 It is likely that the decreasing esophagus, and esophageal adenocarci- From the Departments of incidence of these disorders is the result noma.11 Furthermore, it has been sug- Surgery, Lund University of improvements in sanitation and pub- gested that eradication of H pylori may pre- Hospital, Lund, Sweden ¨ lic health, which has resulted in a lower cipitate reflux esophagitis in patients with (Dr Oberg); and University of 12-14 Southern California, Los prevalence of H pylori infection in the gen- duodenal ulcers. Angeles (Drs Peters, Nigro, eral population. To evaluate the role of H pylori in the Theisen, Hagen, S. R. On the contrary, the prevalence of pathogenesis of GERD, its presence in a DeMeester, Bremner, and gastroesophageal reflux disease (GERD) large group of patients representing the full T. R. DeMeester). and its complications, erosive esophagi- spectrum of GERD was studied.

ARCH SURG/ VOL 134, JULY 1999 722

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 PATIENTS AND METHODS hematoxylin-eosin and analyzed for the type and the condi- tion of the epithelium. The biopsy specimens were evaluated for the presence of H pylori infection using a Giemsa stain. PATIENT POPULATION Gastric fundic mucosa was characterized by glands that contain parietal and chief cells but are devoid of mucous cells Between April 1993 and June 1998, 229 consecutive pa- except those lining the surface and foveolar region. Glands tients (122 men and 107 women; median age, 51 years; age composed entirely of mucous cells without any parietal or range, 16-85 years) with symptoms of foregut disease and chief cells characterized cardiac mucosa. Inflammation of car- no history of gastric or esophageal surgery were exam- diac mucosa or carditis was identified by the presence of eo- ined. Patients with a named motility disorder and those in sinophil or plasma cell infiltration of the lamina propria and whom the histological evaluation of the biopsy specimens hyperplasia of the mucous cells in the foveolar region of the did not include a staining specific for H pylori (Giemsa) cardiac mucosa. Specialized intestinal metaplasia in cardiac were excluded. The foregut symptoms consisted of heart- mucosa was identified by the presence of a columnar epithe- burn, regurgitation, dysphagia, chest pain, epigastric pain, lium with a villiform surface, mucous glands, and well- or symptoms suggestive of aspiration, such as recurrent defined goblet cells. The presence of goblet cells was con- pneumonia, wheezing, and cough. All patients underwent firmed by positive staining with Alcian blue at pH 2.5. upper endoscopy with biopsy, standard esophageal manometry to evaluate the lower esophageal sphincter, and 24-HOUR pH MONITORING 24-hour esophageal pH monitoring to quantify esophageal exposure to gastric juice. A glass pH probe (Ingold, Urdorf, Switzerland) was posi- An additional 114 patients (96 men and 18 women; tioned 5 cm above the manometrically defined upper bor- median age, 68 years; age range, 28-82 years) with adeno- der of the lower esophageal sphincter. Patients were in- carcinoma of the esophagus or GEJ and with biopsy speci- structed to follow their daily routine as closely as possible mens obtained from the gastric antrum were studied. and to record in a diary all food ingested, symptoms expe- rienced, and time spent in the upright and supine posi- ENDOSCOPIC DATA tions. The diet was restricted to food with a pH exceeding 5. Medications known to affect gastrointestinal tract mo- All patients underwent a systematic endoscopic examina- tility or acid secretion were discontinued 3 days before tion of the duodenum, pylorus, stomach, and esophagus. testing, except for proton pump inhibitors, which were The GEJ was defined as the place where the gastric rugal discontinued at least 2 weeks earlier. folds ended and the tubular esophagus began. A columnar- Analysis was performed with a commercially avail- lined esophagus was identified when the squamocolum- able software program (Synectics Medtronics Inc, Minne- nar junction or any part of its circumference extended above apolis, Minn). A reflux episode was defined as any de- the GEJ. In patients with an irregular squamocolumnar junc- crease in pH below 4.0. The percentage of time the tion, biopsy specimens were obtained from the tongues of esophageal mucosa was exposed to a pH below 4.0 for the the glandular mucosa extending into the esophagus. In pa- total monitored period was recorded. Based on the data from tients whose squamocolumnar junction was above the GEJ, 50 asymptomatic volunteers, patients with an esophageal 4 quadrant biopsy specimens were obtained every 2 cm of pH below 4.0 for more than 4.4% of the recording time were the columnar-lined segment. Multiple biopsy specimens classified as having abnormal esophageal acid exposure.15 were also obtained immediately below the squamocolumnar junction and from the gastric antrum in all patients. STATISTICAL ANALYSIS Patients were identified as having Barrett esophagus by the presence of intestinal metaplasia in a columnar- Data are reported as medians and interquartile ranges, un- lined esophagus. less otherwise stated. The Fisher exact test was used to compare proportions between individual groups. Compari- HISTOLOGICAL FEATURES sons of proportions between more than 2 groups were done using the ␹2 test. Continuous data were compared using The biopsy specimens were fixed in 10% buffered formalin, the Kruskal-Wallis test, and comparisons between indi- embedded in paraffin, sectioned, and mounted on slides by vidual groups were performed using the Mann-Whitney U means of standard technique. Slides were stained with test. PϽ.05 was considered significant.

RESULTS Table 2 shows the relationship between H pylori infection and the features of GERD in the total study popu- The overall prevalence of H pylori in patients with foregut lation. There was no difference in the prevalence of ab- symptoms and benign disease was 14.0% (32/229). normal esophageal acid exposure, erosive esophagitis, and Table 1 shows demographic data and the distribution Barrett esophagus in patients with and without H pylori of H pylori infection in these patients. In addition to an- infection. The prevalence of erosive esophagitis in the sub- tral infection, 24 patients also had evidence of H pylori set of patients with abnormal esophageal acid exposure on biopsy specimens just below the squamocolumnar was not different when stratified according to the pres- junction, indicating a generalized infection. When found ence or absence of H pylori infection (25.0% vs 31.1%; just below the squamocolumnar junction, H pylori in- P = .54). Similarly, the prevalence of Barrett esophagus fection was always present in the antrum. in patients with abnormal esophageal acid exposure was

ARCH SURG/ VOL 134, JULY 1999 723

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 Table 1. Demographic Data of the Patients Without Table 4. Prevalence of Helicobacter pylori in Patients Adenocarcinoma Grouped According to the Presence With and Without Inflamed Fundic Epithelium and Distribution of Helicobacter pylori Infection at the Gastroesophageal Junction and No Evidence of Barrett Esophagus H pylori Sex, Median Age Infection Status No. (%) of Patients M/F (Range), y Fundic Mucosa

None 197 (86.0) 105:92 52 (16-85) Normal Inflamed Antral only 8 (3.5) 6:2 46 (25-69) Variable (n = 141) (n = 41) P Generalized 24 (10.5) 11:13 51 (36-74) Median age (range), y 51 (16-85) 48 (27-80) .83 Esophageal pH Ͻ4.0* 3.8 (1.5-7.6) 3.6 (1.0-7.8) .77 Abnormal acid exposure† 73 (51.8) 22 (53.7) .86 H pylori infection† 4 (2.8) 27 (65.9) .001 Table 2. Relationship Between Helicobacter pylori Infection and the Features of Gastroesophageal Reflux Disease *Values are given as median (interquartile range) percentage of time. †Values are given as number (percentage) of patients. Features of H pylori Infection Gastroesophageal Twenty-two (19.3%) of the 114 patients with ad- Reflux Disease No (n = 197) Yes (n = 32) P enocarcinoma of the esophagus and GEJ had H pylori in- Esophageal pH Ͻ4.0* 5.5 (2.3-10.4) 4.4 (2.3-9.2) .47 fection. There was no difference in the prevalence of H Abnormal acid exposure† 110 (55.8) 16 (50.0) .57 pylori between patients with benign (32 [14.0%] of 229) Erosive esophagitis† 61 (30.9) 8 (25.0) .54 and malignant disease regardless of the location or be- Barrett esophagus† 35 (17.7) 5 (15.6) 1.00 tween patients with tumors of the esophagus (5 [13.5%] *Values are given as median (interquartile range) percentage of time. of 37) and the GEJ (17 [22.1%] of 77). †Values are given as number (percentage) of patients. COMMENT

Table 3. Prevalence of Helicobacter pylori in Patients With This study demonstrates that H pylori plays no role in and Without Carditis and No Evidence of Barrett Esophagus the pathogenesis of GERD or its complications. We found no difference in the median time of esophageal Carditis acid exposure or the prevalence of abnormal esophageal acid exposure, erosive esophagitis, and Barrett Variable No (n = 73) Yes (n = 115) P esophagus in patients with and without H pylori. Median age (range), y 50 (16-80) 51 (23-85) .13 In the present study, the prevalence of H pylori was Esophageal pH Ͻ4.0* 2.9 (0.5-6.8) 4.9 (2.9-9.5) .001 Abnormal acid exposure† 28 (38.4) 64 (55.7) .02 similar in patients with and without erosive esophagitis. H pylori infection† 16 (21.9) 11 (9.6) .03 Furthermore, there was no difference in the prevalence of esophagitis in patients with pH-positive GERD when *Values are given as median (interquartile range) percentage of time. grouped according to the presence or absence of H py- †Values are given as number (percentage) of patients. lori infection. Although our data suggest that H pylori plays no role in the development of erosive esophagitis, there not different in patients with and without H pylori in- are studies suggesting that it may have a protective ef- fection (15.6% vs 17.8%; P Ϸ 1.00). In the presence of fect. Labenz et al14 have shown that patients with duo- Barrett esophagus, H pylori was never found in an un- denal ulcer and successful eradication of H pylori devel- derlying mucosa that was intestinalized, although it was oped esophagitis significantly more frequently during the seen in biopsy specimens of cardiac mucosa from the lower ensuing 3 years than those who remained positive for H esophagus in 2 patients with Barrett esophagus. Five of pylori. The investigators suggested that the increased fre- the 40 patients with Barrett esophagus had low-grade dys- quency of esophagitis may result from increased gastric plasia, none of whom had histological evidence of H py- acid secretion following eradication therapy. Others have lori infection. suggested, however, that gastric acid secretion may de- Inflamed cardiac mucosa at the GEJ or carditis was crease following H pylori eradication16 and have ques- found in 61.2% (115/188) of patients with a normal GEJ. tioned the association between eradication therapy and The presence of carditis was inversely related to H py- GERD with and without erosive esophagitis.17-19 lori infection and significantly associated with greater Recent investigations have focused on the role of sub- esophageal acid exposure (Table 3). In contrast, when populations of H pylori, including cagA+ strains, in the fundic mucosa was found on biopsy specimens of the GEJ, development of GERD and its complications. Vicari et inflammation was strongly associated with the presence al12 demonstrated that in patients with H pylori infec- of H pylori but not with higher esophageal acid expo- tion the prevalence of cagA+ strains progressively de- sure (Table 4). Eleven (5.8%) of the 188 patients with creased with the severity of GERD, including Barrett an endoscopically normal GEJ had cardiac mucosa har- esophagus and esophageal adenocarcinoma. Chow et al11 boring intestinal metaplasia found at the GEJ. One (9.1%) confirmed an inverse relationship between the presence of these patients had evidence of H pylori infection that of cagA+ and adenocarcinoma of the esophagus and the was not different from those without intestinal metapla- GEJ. Both groups postulated that cagA+ strains may pro- sia (26 [14.6%] of 178, P = .52). tect from the development of adenocarcinoma by induc-

ARCH SURG/ VOL 134, JULY 1999 724

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 ing more severe mucosal inflammation and atrophic gas- 3. Parkin DM, Laara E, Muir CS. Estimates of the worldwide frequency of sixteen tritis, thereby decreasing acid reflux.11 Present data major cancers in 1980. Int J Cancer. 1988;41:184-197. 4. Blaser MJ. Helicobacter pylori infection and adenocarcinoma of the distal stom- regarding gastric acid secretion are conflicting, how- ach. Eur J Gastroenterol Hepatol. 1993;5(suppl 1):S99-S102. ever, and further studies are required to test if this hy- 5. Blot WJ, Devesa SS, Kneller RW, Fraumeni JF Jr. Rising incidence of adenocar- pothesis is true.16,20-22 cinoma of the esophagus and gastric cardia. JAMA. 1991;265:1287-1289. We have shown that carditis represents one of the 6. Pera M, Cameron AJ, Trastek VF, Carpenter HA, Zinsmeister AR. Increasing in- earliest histological changes of GERD,23 while oth- cidence of adenocarcinoma of the esophagus and esophagogastric junction. Gas- 24-28 troenterology. 1993;104:510-513. ers have suggested that carditis and intestinal meta- 7. Werdmuller BF, Loffeld RJ. Helicobacter pylori infection has no role in the patho- plasia of the cardia are caused by H pylori. We found no genesis of reflux esophagitis. Dig Dis Sci. 1997;42:103-105. association between the presence of cardiac mucosa or 8. Csendes A, Smok G, Cerda G, Burdiles P, Mazza D, Csendes P. Prevalence of carditis, with or without intestinal metaplasia to H py- Helicobacter pylori infection in 190 control subjects and in 236 patients with gas- lori infection. In fact, carditis was inversely related to the troesophageal reflux, erosive esophagitis or Barrett’s esophagus. Dis Esopha- gus. 1997;10:38-42. presence of H pylori, and patients with carditis had sig- 9. Varanasi RV, Fantry GT, Wilson KT. Protective role of Helicobacter pylori infec- nificantly greater esophageal acid exposure than pa- tion in gastroesophageal reflux disease [abstract]. Gastroenterology. 1998;114: tients without carditis. Inflammation in fundic epithe- A322. lium found on biopsy specimens obtained from the area 10. Subei IM, Salama M, Elderiny S, O’Brien MJ. Does Helicobacter pylori play a role just below the squamocolumnar junction, however, was in gastroesophageal reflux disease (GERD) [abstract]? Gastrointest Endosc. 1998; 47:AB76. strongly associated with H pylori infection. The contro- 11. Chow WH, Blaser MJ, Blot WJ, et al. An inverse relation between cagA+ strains versy about the role of H pylori in the pathogenesis of of Helicobacter pylori infection and risk of esophageal and gastric cardia adeno- carditis exists in part because most researchers26,28-30 do carcinoma. Cancer Res. 1998;58:588-590. not relate the location of the biopsy specimens to the un- 12. Vicari JJ, Peek RM, Falk GW, et al. The seroprevalence of cagA-positive Helico- derlying type of mucosa in which inflammation and in- bacter pylori strains in the spectrum of gastroesophageal reflux disease. Gas- troenterology. 1998;115:50-57. testinal metaplasia occurs. Carditis in these reports is de- 13. Weston AP, Badr AS, Topalosvki M, Cherian R, Dixon A. Prospective evaluation fined by the location of the biopsy specimens, ie, the of the association of gastric H pylori infection with Barrett’s dysplasia and Bar- anatomic gastric cardia, which infrequently harbors car- rett’s adenocarcinoma [abstract]. Gastroenterology. 1998;114:A703. diac mucosa.31 Inflammation and intestinal metaplasia 14. Labenz J, Blum AL, Bayerdorffer E, Meining A, Stolte M, Borsch G. Curing Heli- arising in fundic or antral mucosa have been associated cobacter pylori infection in patients with duodenal ulcer may provoke reflux esopha- 26-28 gitis. Gastroenterology. 1997;112:1442-1447. with H pylori infection. In contrast, intestinal meta- 15. Jamieson JR, Stein HJ, DeMeester TR, et al. Ambulatory 24-h esophageal pH plasia in cardiac mucosa is not associated with the pres- monitoring: normal values, optimal thresholds, specificity, sensitivity, and re- ence of H pylori or other gastric pathological features; producibility. Am J Gastroenterol. 1992;87:1102-1111. rather, it is associated with gastroesophageal re- 16. El-Omar EM, Penman ID, Ardill JE, Chittajallu RS, Howie C, McColl KE. Helico- 23,32,33 bacter pylori infection and abnormalities of acid secretion in patients with duo- flux. denal ulcer disease. Gastroenterology. 1995;109:681-691. The incidence of adenocarcinoma of the esophagus 17. Carbone F, Neri M, Laterza F, Ricciuti M, Sajterova Z, Cuccurullo F. Twenty-four and GEJ has risen more rapidly during the past decade than hour esophageal pH-metry in patients with non-ulcer dyspepsia is unchanged any other solid tumor.5,6,34 Most of these cancers arise in after H pylori eradication [asbtract]. Gastroenterology. 1998;114:A84. patients with Barrett esophagus. The reason for its pro- 18. Talley NJ, Janssens J, Lauritsen K, et al. No increase of reflux symptoms or esophagitis in patients with non-ulcer dyspepsia 12 months after Helicobacter pylori gression to cancer in a few patients is unknown. Helico- eradication: a randomized double blind placebo-controlled trial [abstract]. Gas- bacter pylori is likely important in the pathogenesis of troenterology. 1998;114:A306. chronic atrophic gastritis and cancer of the gastric an- 19. O’Connor HJ, McGee C, Mehana N, Cunnane K. Prevalence of gastroesophageal trum and body.1,35-38 The result of the present study con- reflux disease (GERD) in H pylori–positive peptic ulcer disease and the impact firms the results of other reports7,8 and finds no associa- of eradication therapy [asbtract]. Gastroenterology. 1998;114:A244. 20. Peterson WL, Barnett CC, Evans DJJ, et al. Acid secretion and serum gastrin in tion with the presence of H pylori, indicating that it has normal subjects and patients with duodenal ulcer: the role of Helicobacter py- no role in the pathogenesis of adenocarcinoma of the lori. Am J Gastroenterol. 1993;88:2038-2043. esophagus and GEJ. 21. Levi S, Beardshall K, Haddad G, Playford R, Ghosh P, Calam J. Campylobacter Helicobacter pylori has no role in the pathogenesis pylori and duodenal ulcers: the gastrin link. Lancet. 1989;1:1167-1168. 22. Cover TL, Glupczynski Y, Lage AP, et al. Serologic detection of infection with of GERD or its manifestations. Whether specific strains cagA+ Helicobacter pylori strains. J Clin Microbiol. 1995;33:1496-1500. of H pylori protect against the complications of GERD 23. Spechler SJ, Wang HH, Chen YY, Zeroogian JM, Antonioli DA, Goyal RK. GERD needs further investigation. versus H pylori infections as potential causes of inflammation in the gastric cardia [abstract]. Gastroenterology. 1997;112:A297. Presented at the 106th Scientific Session of the Western Sur- 24. O¨ berg S, Peters JH, DeMeester TR, et al. Inflammation and specialized intestinal metaplasia of cardiac mucosa is a manifestation of gastroesophageal reflux dis- gical Association, Indianapolis, Ind, November 17, 1998. ease. Ann Surg. 1997;226:522-530. Reprints: Jeffrey H. Peters, MD, Department of Sur- 25. Spechler SJ, Zeroogian JM, Antonioli DA, Wang HH, Goyal RK. Prevalence of gery, University of Southern California, School of Medi- metaplasia at the gastro-oesophageal junction. Lancet. 1994;344:1533-1536. cine, 1510 San Pablo St, Suite 514, Los Angeles, CA 90033- 26. Morales TG, Sampliner RE, Bhattacharyya A. Intestinal metaplasia of the gastric cardia. Am J Gastroenterol. 1997;92:414-418. 4612 (e-mail: [email protected]). 27. Spechler SJ, Wang HH, Chen YY, Zeroogian JM, Antonioli DA, Goyal RK. Inflam- mation of the cardia and H pylori infections are not risk factors for intestinal meta- REFERENCES plasia at the esophagogastric junction. Gastroenterology. 1997;112:A297. 28. Goldblum JR, Vicari JJ, Falk GW, et al. Inflammation and intestinal metaplasia of the gastric cardia: the role of gastroesophageal reflux and H pylori infection. 1. Marshall BJ, Warren JR. Unidentified curved bacilli in the stomach of patients Gastroenterology. 1998;114:633-639. with gastritis and peptic ulceration. Lancet. 1984;1:1311-1315. 29. Weston AP, Krmpotich PT, Cherian R, Dixon A, Topalovski M. Prospective evalu- 2. El-Serag HB, Sonnenberg A. Opposing time trends of peptic and reflux [ab- ation of intestinal metaplasia and dysplasia within the cardia of patients with Bar- stract]. Gastroenterology. 1997;112:A113. rett’s esophagus. Dig Dis Sci. 1997;42:597-602.

ARCH SURG/ VOL 134, JULY 1999 725

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 30. Morales TG, Bhattacharyya A, Johnson C, Sampliner RE. Is Barrett’s esophagus ulcer disease, inflammatory bowel disease, mucosal- associated with intestinal metaplasia of the gastric cardia? Am J Gastroenterol. associated lymphoma of the stomach, and probably other en- 1997;92:1818-1822. tities yet to be discovered. This paper confirms that gastro- 31. Jain R, Aquino D, Harford WV, Lee E, Spechler SJ. Cardiac epithelium is found esophageal reflux disease, a common surgical disease with infrequently in the gastric cardia [abstract]. Gastroenterology. 1998;114:A160. 32. Bowrey DJ, Clark GWB, Stockdale E, Williams GT, Carey PD. Histological abnor- effective surgical treatment, will not likely be removed from malities at the gastro-esophageal junction: correlation with acid and bile reflux our therapeutic arena by antibiotics. The authors’ study was [abstract]. Gastroenterology. 1998;114:A79. well designed and thoroughly carried out. I believe the results 33. Van Vaerenbergh W, Tack J, Geboes K, et al. Microscopic Barrett’s esophagus: speak for themselves. Did any of the patients with benign or relationship to acid and bile reflux in man [abstract]. Gastroenterology. 1998; malignant disease harboring H pylori infection have active pep- 114:A322. tic ulcer disease? Since 16% of your total patient population 34. Haggitt RC. Adenocarcinoma in Barrett’s esophagus: a new epidemic? Hum Pathol. were H pylori positive, should surgeons be routinely screening 1992;23:475-476. for this infection and treating it preoperatively? Might this not 35. Price AB, Levi J, Dolby JM, et al. Campylobacter pyloridis in peptic ulcer dis- reduce stress ulcerations and postoperative GI [gastrointesti- ease: microbiology, pathology, and scanning electron microscopy. Gut. 1985; 26:1183-1188. nal tract] bleeding? 36. Sipponen P. Helicobacter pylori infection: a common worldwide environmental Dr Peters: Twenty years ago at this meeting, we probably risk factor for gastric cancer? Endoscopy. 1992;24:424-427. would have had 6 or 7 papers on peptic ulcer disease, rather than 37. De KE, Buset M, Fernandes E, Deltenre M. Helicobacter pylori: the link with gas- the 6 or 7 on reflux that we heard today. I wonder if that is not tric cancer. Eur J Cancer Prev. 1994;3:247-257. because of this bug. The title here says that there is no associa- 38. Correa P. Human gastric carcinogenesis: a multistep and multifactorial process— tion with gastroesophageal reflux disease. As our data suggest, First American Cancer Society Award Lecture on Cancer Epidemiology and Pre- it is not involved in the pathogenesis. It may be, however, that vention. Cancer Res. 1992;52:6735-6740. there is an inverse relationship between Helicobacter and reflux disease. I suspect this will become clear in the years to come. DISCUSSION To answer the questions, there was not any peptic ulcer disease in the population of patients that were included in this James R. DeBord, MD, Peoria, Ill: These authors have pre- study. It is actually fairly uncommon for us to see concomi- sented a defining paper that clearly and scientifically confirms tant reflux disease and ulcer disease these days. the absence of any role of Helicobacter pylori in the pathogen- Second, we do perform routine biopsies, and if H pylori is esis of reflux disease or adenocarcinoma of the esophagus. Bac- identified, it is my practice to treat it. Routine treatment of H terial infections have plagued surgeons for centuries. How- pylori is a controversial topic that is much debated. Largely be- ever, bacterial vectors for surgical disease is a relatively new cause of its association with gastric cancer, many feel that it is concept that has impacted surgeons in the treatment of peptic reasonable to do so, and we do.

IN OTHER AMA JOURNALS

ARCHIVES OF INTERNAL MEDICINE Quality of Survival After Cardiopulmonary Resuscitation Rien de Vos, RN; Hanneke C. J. M. de Haes, MS, PhD; Rudolph W. Koster, MD, PhD; Rob J. de Haan, RN, PhD Background: Outcome of cardiopulmonary resuscitation (CPR) can be poor, in terms of life expectancy and quality of life. Objectives: To determine the impact of patient characteristics before, during, and after CPR on these outcomes, and to compare results of the quality-of-life assessment with published studies. Methods: In a cohort study, we assessed by formal instruments the quality of life, cognitive functioning, depression, and level of dependence of survivors after in-hospital CPR. Follow-up was at least 3 months after discharge from the hospital (tertiary care center). Results: Of 827 resuscitated patients, 12% (n = 101) survived to follow-up. Of the survivors, 89% participated in the study. Most survivors were independent in daily life (75%), 17% were cognitively impaired, and 16% had depressive symptoms. Multivariate regression analysis showed that quality of life and cognitive function were determined by 2 factors known before CPR—the reason for admission and age. Factors during and after resuscitation, such as prolonged cardiac arrest and coma, did not significantly determine the quality of life or cognitive functioning of survivors. The quality of life of our CPR survivors was worse compared with a reference group of elderly individuals, but better than that of a reference group of patients with stroke. The quality of life did not importantly differ between the compared studies of CPR survivors. Conclusions: Cardiopulmonary resuscitation is frequently unsuccessful, but if survival is achieved, a relatively good quality of life can be expected. Quality of life after CPR is mostly determined by factors known before CPR. These findings may be helpful in informing patients about the outcomes of CPR. (1999;159:249-254) Reprints: Rien de Vos, RN, Academic Medical Center, University of Amsterdam, Division Operation Center H1-212, Meibergdreef 9, PO Box 22700, 1100 DE Amsterdam, the Netherlands (e-mail: [email protected]).

ARCH SURG/ VOL 134, JULY 1999 726