JAMA & ARCHIVES
Arch Fam Med
SEARCH
GO TO ADVANCED SEARCH
HOME  PAST ISSUES  TOPIC COLLECTIONS  CME  PHYSICIAN JOBS  CONTACT US  HELP
Institution: STANFORD Univ Med Center  | My Account | E-mail Alerts | Access Rights | Sign In
  Vol. 9 No. 4, April 2000 TABLE OF CONTENTS
  Archives
  •  Online Features
  Brief Report
 This Article
 •Abstract
 •PDF
 •Send to a friend
 • Save in My Folder
 •Save to citation manager
 •Permissions
 Citing Articles
 •Citing articles on Web of Science (13)
 •Contact me when this article is cited
 Related Content
 •Related article
 •Similar articles in this journal

Prescription of Proton Pump Inhibitors Before Endoscopy

A Potential Cause of Missed Diagnosis of Early Gastric Cancers

John Wayman, MB, FRCS; Nick Hayes, MB, FRCS; Simon A. Raimes, MD, FRCS; S. Michael Griffin, MD, FRCS

Arch Fam Med. 2000;9:385-388.

ABSTRACT

Background  Early gastric cancer is frequently seen with nonspecific dyspeptic symptoms and subtle endoscopic features. Treatment at this stage of the disease produces a high chance of cure. If the diagnosis is missed at this early stage, then the prognosis may be much poorer depending on the subsequent delay in reaching a diagnosis.

Objectives  To report the healing effect of proton pump inhibitors on early gastric cancer.

Methods  This article reports a case series of 7 patients with ulcerated early gastric cancers indistinguishable as malignant gastric ulcers at endoscopy who were inadvertently prescribed a short course of a proton pump inhibitor prior to a second confirmatory endoscopy. The cases studied were patients with dyspeptic symptoms referred from primary care physicians for upper gastrointestinal endoscopy.

Results  In each case the patient became asymptomatic, the endoscopic signs seen at the first endoscopy had resolved, and the lesions could not be recognized even by an experienced endoscopist. If the proton pump inhibitors had been prescribed by the referring physician before the first endoscopy, the diagnosis probably would have been missed. These cases demonstrate the potentially serious masking effect of prescribing a short course of these drugs before making an endoscopic diagnosis. Even though the patient has been referred for endoscopy, the endoscopist may fail to identify the lesion and thus miss the diagnosis.

Conclusions  Primary care physicians must resist the pressures to prescribe proton pump inhibitors before endoscopy, particularly in patients older than 45 years, if the diagnostic yield of gastric cancer in the early curable stages is to be maximized.



INTRODUCTION
 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Report of cases
 •Comment
 •Author information
 •References

THE INCIDENCE of gastric cancer varies worldwide and has changed significantly with time. In the 1930s it accounted for 38% of all cancer deaths in the United States1 and is the eighth most deadly cancer with approximately 24 000 new cases and 14 000 deaths per year.2 In the United Kingdom a similar temporal trend has been observed, and gastric cancer is the fourth most common cause of death accounting for approximately 10 000 deaths per year.3 In Japan the disease remains approximately 4 times as common as these western countries.4

Improvement in the survival of patients with gastric cancer relies largely on earlier diagnosis. Despite improved endoscopic services, there are still significant potentially avoidable delays in diagnosis.5-6 Many patients with dyspepsia take histamine type 2-receptor antagonists prior to referral for endoscopy. Up to 2% of all patients older than 65 years in North America take regular proton pump inhibitors (PPIs),7 and almost one fifth of the patients with gastric cancer will have received at least 1 course of a PPI prior to diagnosis.8

We report a case series of 7 patients with ulcerating early gastric cancer (EGC) who were inadvertently prescribed a PPI after their first endoscopy and who underwent a second endoscopy within 1 to 2 months to take further biopsy specimens to confirm the initial diagnosis. These cases illustrate the effect of a short course of a PPI on the symptoms and endoscopic appearances of EGC.


PATIENTS AND METHODS
 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Report of cases
 •Comment
 •Author information
 •References

An Open Access Endoscopy service was established in 1994 to accept referrals from primary care physicians covering a population of approximately 250 000. The only patient exclusion criterion was being younger than 35 years; these patients were seen in the clinic by the surgical gastroenterologist (S.M.G.) who decided on the appropriateness of further investigation. During the period from August 1994 to July 1997, 4018 patients were assessed with the following results: 620 (15.4%) had a normal endoscopy; cancer was diagnosed in 150 (3.7%), including 43 patients with esophageal cancer, 93 with gastric cancer, and 14 with other types of malignant neoplasms. Three thousand two hundred forty-eight (80.8%) had benign disease diagnosed, including 232 patients who apparently had benign gastric ulcers.

To improve the detection of occult malignancy occurring in gastric ulcers, a protocol was developed. At initial endoscopy extensive biopsy of the edge and base of any gastric ulcer was performed. Patients underwent a second endoscopy to ensure resolution after receiving 1 course of a PPI and to exclude any precipitating factor. If initial biopsy specimens of an apparently benign or indeterminate ulcer produced evidence of malignancy, the patients underwent endoscopy by the surgeon who would be undertaking their operation. The position and size of the ulcer was recorded. Complete macroscopic resolution and absence of neoplasia or dysplasia in all biopsy specimens was taken as evidence of a benign lesion. The patients in whom malignancy was excluded by resolution and histological analysis were discharged from the hospital (n=212); patients in whom histological examination and failure to resolve confirmed underlying malignancy (n=13) were treated. Patients in whom histological examination of the first endoscopic biopsy specimens confirmed the presence of malignancy or dysplasia but in whom macroscopic resolution had occurred at the second endoscopy form this case series (n=7).


REPORT OF CASES
 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Report of cases
 •Comment
 •Author information
 •References

CASE 1

A 69-year-old woman was initially seen by her primary care physician with a 1-month history of epigastric discomfort; she was otherwise asymptomatic. Endoscopic examination 3 weeks after referral revealed a 5-cm diameter ulcer on the greater curve of the stomach that appeared to have an epithelial lining over the crater. Multiple biopsy specimens were obtained and the patient was prescribed omeprazole sodium, 20 mg twice daily, pending histological analysis. Histological examination was suggestive, but not diagnostic, of malignancy. The patient was asymptomatic 4 weeks later when endoscopy showed no evidence of an ulcer. Multiple biopsy specimens of the previously ulcerated area confirmed the occult presence of intramucosal adenocarcinoma. The patient underwent a subtotal gastrectomy. Subsequent histological evidence confirmed lymph node–negative intramucosal EGC.

CASE 2

A 41-year-old man was initially seen by his primary care physician with a 1-month history of dyspepsia. Endoscopy revealed a 1-cm, apparently benign, ulcer in the gastric fundus from which biopsy specimens were obtained. The patient was prescribed a course of omeprazole sodium, 20 mg twice daily. The histological examination suggested that this was a signet cell carcinoma. At the second endoscopy 4 weeks later, the patient was still receiving omeprazole therapy. By then the patient was asymptomatic and the ulcer had completely healed. Random biopsy specimens in the same area reconfirmed evidence of signet cell carcinoma. The patient underwent a total gastrectomy and was found to have lymph node–positive EGC.

CASE 3

A 67-year-old woman was initially seen by her primary care physician with a 6-month history of heartburn, nausea, waterbrash (ie, vagally mediated excessive salivation that results from esophageal acidification during reflux), and weight loss. An apparently benign 5-mm diameter prepyloric gastric ulcer, associated with a Helicobacter pylori infection, was diagnosed at endoscopy. A 2-week course of eradication therapy (a combination of omeprazole, amoxicillin hydrochloride, and metronidazole hydrochloride) was commenced with prompt resolution of symptoms. Histological examination of biopsy specimens revealed the presence of adenocarcinoma. A second endoscopy, performed 6 weeks after the first, with the patient taking no medication, revealed no ulcer. Random biopsy specimens in the area revealed inflammatory changes only. Despite this equivocation, a subtotal gastrectomy was performed. Subsequent histological examination of the resection specimen revealed lymph node–negative EGC.

CASE 4

A 63-year-old man was initially seen by his primary care physician with a 4-month history of epigastric pain and heartburn. At the initial endoscopy a 1-cm diameter ulcer was seen near the cardioesophageal junction. Biopsy specimens of this area were obtained and the patient was prescribed lansoprazole, 15 mg twice daily. Histological examination of the biopsy specimens revealed severe dysplasia. The patient underwent a second endoscopy 4 weeks later while receiving lansoprazole therapy. At this time the patient was asymptomatic and the ulcer had healed without trace of disease. Multiple random biopsy specimens in the area of the former ulceration confirmed the presence of intramucosal adenocarcinoma. The patient underwent total gastrectomy. Histological examination showed lymph node–positive EGC.

CASE 5

A 63-year-old woman was initially seen by her primary care physician with a 4-month history of epigastric pain, acid reflux, and constipation. Her symptoms had not responded to a simple over-the-counter antacid. Endoscopy revealed a 1-cm ulcer on the incisura of the lesser curve of the stomach. A biopsy specimen of the ulcer edge was obtained and she was prescribed omeprazole sodium, 20 mg twice daily, with symptomatic resolution within 1 week. A second endoscopy was performed 4 weeks later, while the patient was still receiving omeprazole therapy, after the biopsy specimens confirmed the presence of a poorly differentiated adenocarcinoma. At this reexamination the ulcer had healed and subsequent biopsy specimens were negative for malignancy. A subtotal gastrectomy was performed on the basis of the initial biopsy. Histological examination confirmed the presence of a healed malignant ulcer overlying a submucosal EGC.

CASE 6

A 57-year-old man was initially seen by his primary care physician with a 3-month history of epigastric pain and vomiting exacerbated by eating. Endoscopy revealed a 1.5-cm diameter gastric ulcer on the lesser curve of the stomach. A biopsy specimen of the ulcer edge was obtained and the patient was prescribed omeprazole, 40 mg/d. His symptoms rapidly and completely resolved. The biopsy specimens showed evidence of moderate dysplasia. A second endoscopy performed after a 12-week treatment with omeprazole showed that the ulcer had healed, leaving only very slight mucosal irregularity. Repeated biopsy specimens obtained from the previously ulcerated area confirmed the presence of adenocarcinoma. The patient underwent a radical total gastrectomy. Histological examination confirmed the presence of an intramucosal EGC.

CASE 7

A 55-year-old woman was initially seen by her primary care physician with a 3-month history of epigastric pain and the sensation of food sticking in her lower esophagus. Endoscopy revealed the presence of a single 1-cm ulcer at the cardioesophageal junction. Omeprazole sodium, 20 mg twice daily, was prescribed pending the results of histological examination. Despite histological evidence of adenocarcinoma in the biopsy specimens, no ulcer was visible at the second endoscopy 4 weeks later while the patient was receiving omeprazole therapy. Extensive repeated biopsy specimens obtained from the previously ulcerated area confirmed the presence of malignancy. A total gastrectomy was performed and, although no macroscopic lesion could be seen by the pathologist in the resected specimen, intramucosal cancer was found on sectioning of the cardioesophageal junctional area.


COMMENT
 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Report of cases
 •Comment
 •Author information
 •References

The prognosis of patients with gastric cancer depends almost entirely on the stage at which the disease is diagnosed. The results of surgery for EGC where malignant cells are confined to the mucosa or submucosa of the stomach are excellent, with a 5-year survival rate of more than 98% compared with 93% for stage I, 69% for stage II, and 28% for stage III gastric cancer.3 However, in the United Kingdom, less than 20% of gastric cancers are diagnosed as EGCs and more than 75% are diagnosed as stage III or IV.9 In the United States, findings are similar with an incidence of stage I disease in 1991 of 19.4% compared with stage III and IV of 65%.10 The best way of increasing the proportion of EGCs is by screening the asymptomatic at-risk population, as happens in Japan. In that country more than half of all gastric cancers are diagnosed at this early curable stage.4 Screening in the West is infeasible or not cost effective and so early diagnosis has to rely on referral of the symptomatic at-risk population for gastroscopy. It has been shown that some patients with EGC do experience symptoms and in most these are nonspecific dyspeptic symptoms indistinguishable from those of benign disease.11 The definition of this at-risk group is thus any patient older than 45 years (younger, if there are additional risk factors) who develops new dyspeptic symptoms. Diagnosis depends on the patient reporting his or her symptoms and the primary care physician then making the referral for endoscopy without delay. In our case series, approximately 1.4% of the population aged older than 35 years were referred for endoscopy during a 3-year period by their primary care physician. In a survey of 1644 residents of Olmsted County (Minnesota), Talley et al12 found a dyspepsia prevalence of 22%. There clearly are a large number of patients with dyspepsia who have not sought medical advice or have not been referred for endoscopy.

The diagnosis of EGC requires the endoscopist to recognize often quite subtle lesions in the gastric mucosa. Most frequently there is an ulcerated lesion that may be indistinguishable from a benign ulcer. Biopsy specimens of all gastric ulcers must be obtained at several points to allow a confident histological diagnosis of the nature of the ulcer. In this case series we describe the prescription of a short course of a PPI after diagnosis of a gastric ulcer of indeterminate or apparently benign nature has led to complete or nearly complete resolution of the endoscopic signs. In each case we believe that even an experienced endoscopist may not have identified the abnormal area at the second endoscopy. These changes have occurred within 4 weeks of treatment in most cases and just a 2-week course in 1. If we move 1 step back down the diagnostic path and the PPI were to be prescribed before the first endoscopy, then we feel certain that the diagnosis of an EGC would have been missed in some, if not all of these cases.

When cimetidine hydrochloride was first introduced in the mid 1970s, there were concerns that inappropriate prescription without an endoscopic diagnosis could delay diagnosis of gastric cancer.13 At that time it was unknown whether H2-receptor antagonists could change the endoscopic appearance or even heal ulcers. There was and still is pressure from patients for treatment of their dyspeptic symptoms before endoscopy and especially where waiting lists have developed for endoscopic services. Even in our dedicated service, the median time from referral to procedure was 64 days (range, 1-258 days). The increased potential of PPIs to heal ulcers and relieve dyspeptic symptoms may not have been fully appreciated. Patients are being prescribed this group of drugs between referral for endoscopy and the actual procedure. Recent studies show that up to one third of the patients with gastric cancer have received PPIs before the diagnosis has been made.14 In our unit, 20% of the patients with gastric cancer have been shown to have taken PPIs for a median of 12 weeks (range, 4-200 weeks) prior to the diagnosis being established.8 It has to be appreciated that the manufacturers of all these drugs specifically warn against prescription in patients who might have malignant ulcers, and by definition this includes any patient older than 45 years with new or changed dyspeptic symptoms. This article provides the first hard evidence to demonstrate the potential diagnostic confusion if a patient with gastric cancer is prescribed a PPI before the initial endoscopy.

No data show how long the diagnosis of gastric cancer is delayed if a PPI has been inappropriately prescribed before the initial gastroscopy. Given that no ulcer will have been seen at endoscopy, there is a fear that repeated courses may be prescribed when symptoms return. It is known that the progression from early to the advanced stages of gastric cancer takes years rather than months. We have treated several cases of advanced gastric cancer in which prolonged control of symptoms by continuous PPI treatment has delayed diagnosis for longer than 1 year. We should add a warning against the repeat prescription of PPIs in patients with nonulcer dyspepsia if they had received these drugs prior to their gastroscopy, in case an early lesion has been missed.

The cases we have reported provide a model for the potential effects of the inappropriate prescription of PPIs on the diagnosis of EGC. The proportion of patients with EGC who are receiving these drugs before the initial endoscopy is unknown. While the vast majority of primary care physicians have heeded the advice to request an endoscopy in patients older than 45 years who have new dyspeptic symptoms, it is likely that most will not have appreciated the effects on diagnostic yield resulting from the prescription of even a short course of PPI before the initial endoscopy.15 The pressures to prescribe a potent antisecretory drug must be addressed and not least by the provision of a rapid access endoscopy service. We urge primary care physicians to withstand the pressures to prescribe PPIs before the initial endoscopy in the at-risk group if gastric cancer really is to become a curable disease.


AUTHOR INFORMATION
 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Report of cases
 •Comment
 •Author information
 •References

Accepted for publication November 16, 1999.

Dr Wayman is a research fellow with the Northern Oesophago-Gastric Cancer Unit, The Royal Victoria Infirmary, Newcastle upon Tyne, England.

Corresponding author: S. Michael Griffin, MD, FRCS, Consultant Surgical Gastroenterologist, The Northern Oesophago-Gastric Cancer Unit, The Royal Victoria Infirmary, Queen Victoria Road, Newcastle upon Tyne NE1 4LP, England.

From the Northern Oesophago-Gastric Cancer Unit, The Royal Victoria Infirmary, Queen Victoria Infirmary, Newcastle upon Tyne, England.


REFERENCES
 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Report of cases
 •Comment
 •Author information
 •References

1. Miller BA, ed, Ries LAG, ed, Hankey BF, ed. Cancer Statistics Review: 1973-1990. Bethesda, Md: National Cancer Institute; 1993. NIH publication 93-2789.
2. American Cancer Society. Cancer Facts and Figures 1994. Atlanta, Ga: American Cancer Society; 1994.
3. Office of Population Censuses and Surveys. Mortality Statistics by Cause. London, England: Her Majesty's Stationery Office; 1991.
4. Kaneko E, Nakmura T, Umeda N, et al. Outcome of gastric carcinoma detected by mass survey in Japan. Gut. 1977;18:626-631. FREE FULL TEXT
5. Martin IG, Young S, Sue-Ling HM, Johnston D. Delays in the diagnosis of oesophago-gastric cancer: a consecutive case series. BMJ. 1997;314:467-471. FREE FULL TEXT
6. Wile AG, Hourani L, Schell MJ. The contributions of patient factors, physician delay, and tumor biology to the outcome of gastric cancer. Am Surg. 1993;59:850-854. ISI | PUBMED
7. McBride JE, Pater JL, Dorland JL, Lam YM. Extent and variation of omeprazole prescribing in an elderly population of Ontario. Ann Pharmacother. 1997;31:411-416. ABSTRACT
8. Wayman J, Hayes N, Biggin CS, Karat D, Raimes S, Griffin SM. Proton pump inhibitors delay the diagnosis of gastric cancer [abstract]. Br J Surg. 1997;84(suppl 1): 62.
9. Sue-Ling HM, Johnston D, Martin IG, et al. Gastric cancer: a curable disease in Britain. BMJ. 1993;307:591-596.
10. Lawrence W, Menck HR, Steele GD, Winchester DP. The National Cancer Data Base Report on Gastric Cancer. Cancer. 1995;75:1734-1744. FULL TEXT | ISI | PUBMED
11. Hundahl SA, Menck HR, Mansour EG, Winchester DP. The National Cancer Data Base Report on Gastric Carcinoma. Cancer. 1997;80:2333-2341. FULL TEXT | ISI | PUBMED
12. Talley NJ, Zinsmeister AR, Schleck CD, Melton LJ. Smoking, alcohol, and analgesics in dyspepsia and among dyspepsia subgroups: lack of an association in a community. Gut. 1994;35:619-624. FREE FULL TEXT
13. Taylor RH, Lovell D, Menzies-Gow N, La Brooy SJ, Misiewicz JJ. Misleading response of malignant ulcers to cimetidine. Lancet. 1978;1:686-688. ISI | PUBMED
14. Harris AW, Beveridge I, Dove-Edwin I, Keen J, Silk DBA, Misiewicz JJ. Audit of the use of anti-secretory drugs [abstract]. Gut. 1997;40(suppl 1):A59.
15. Mitchell RMS, Tham TCK, Collins JSA, Watson RGP. Differences in diagnostic yield of upper gastrointestinal endoscopy in dyspeptic patients on proton pump inhibitors and H2-receptor antagonists [abstract]. Gut. 1997;40(suppl 1):A60.

RELATED ARTICLE

The Archives of Family Medicine Continuing Medical Education Program
Arch Fam Med. 2000;9(4):375-377.
FULL TEXT  





HOME | CURRENT ISSUE | PAST ISSUES | TOPIC COLLECTIONS | CME | PHYSICIAN JOBS | HELP
CONDITIONS OF USE | PRIVACY POLICY | CONTACT US | SITE MAP
 
© 2000 American Medical Association. All Rights Reserved.