Post-cimetidine Surveillance for up to Ten Years: Incidence of Carcinoma of the Stomach and Oesophagus

Transcription

1 Quarterly Journal of Medicine, New Series 7, No., pp. 9, January 99 Postcimetidine Surveillance for up to Ten Years: Incidence of Carcinoma of the Stomach and Oesophagus DG COLINJONES, MJS LANGMAN*, DH LAWSONf, RFA LOGANt KR PATERSON and MP VESSEY** From the Queen Alexandra Hospital, Portsmouth, ^University of Birmingham, ^Glasgow Royal Infirmary and University of Strathclyde, %University of Nottingham, ^Glasgow Royal Infirmary and ** University of Oxford Accepted July 9 99 SUMMARY The longterm effects of cimetidine on the occurrence of gastric and oesophageal cancer were assessed in a prospective cohort study of 99 patients who had been prescribed cimetidine. They were first identified between 97 and 9, and cancer registrations and deaths were identified among them over a period of up to years. One hundred and eleven cancers were identified after the start of cimetidine treatment: 7 were adenocarcinomas of the stomach, 7 were carcinomas of the cardia and/or oesophagus ( adenocarcinomas, five unknown histology) and the remaining tumours were squamous cell cancers of the oesophagus. Only six patients presented with early gastric cancers. Over a period of eight years the ratio of observed to expected (O/E) gastric cancer deaths has fallen from.7 (p <.) to. (NS). The O/E ratio of oesophageal cancer deaths also fell over the first six years of study, from. (p <.) to. (NS) but it hasrisenslightly in years 7 and to.7 (p <.). Thesefindingsdo not suggest that there is an increased risk of developing oesophageal or gastric cancer from cimetidine treatment, and are generally consistent with cimetidine being used inadvertently to treat the early symptoms of gastric and oesophageal cancer. The slight rise in oesophageal cancer deaths in years 7 and was unexpected and will be the subject of further observation. Downloaded from by guest on April, INTRODUCTION In 97 we began a postmarketing surveillance study to detect possible adverse effects of cimetidine treatment. Information was collected in four centres (Glasgow, Nottingham, Oxford and Portsmouth) on a group of 99 patients who had been prescribed cimetidine []. In 9, preliminary results were reported [] on the incidence of gastric cancer in the study group. We found cases of gastric cancer diagnosed after the patient had received cimetidine, but following careful review of the records we concluded that the cancer was present before the patient took cimetidine in almost all cases. We have continued to follow our original group of patients and now have data on gastric and oesophageal cancer detected after up to years of surveillance. Oxford University Press 99

2 DG ColinJones and others METHOD The records of patients from the original group of 99 identified between 97 and 9 who took cimetidine and who continued in the study after the first year have been 'flagged' at the NHS Central Registries at Southport and Edinburgh. We are thus automatically notified of any patient who dies or who is registered as having cancer. Cancer registrations are complete up to 9 and deaths up to 9. Hospital, general practitioner and study records have been examined for each notified case or death to confirm the diagnosis and obtain details for review and analysis. The use of H receptor antagonists in the study group was assessed in 9 and again in 97 by a questionnaire sent to general practitioners caring for a per cent randomly selected sample of the original 99 patients. Data were collected from the patient's doctor and information about previous use of cimetidine and other H receptor antagonists was analysed in month periods. To put our data into the context of the changing incidence of gastric and oesophageal cancer, the death rates per million from these causes in England and Wales (ICD codes and ) were obtained from the Office of Population, Censuses and Surveys for the years 9 to 9 and analysed by age groups. Expected deaths from gastric and oesophageal cancer were calculated from death rates for England and Wales for each year, taking into account age (using year groupings) and sex. Tumour sites as specified on death certificates were used for comparisons between observed and expected rates. RESULTS Numbers of Cases One hundred and fortyone cases of gastric and oesophageal cancer have been notified (including reported previously []); however, cancers were diagnosed before treatment with cimetidine was commenced and are not discussed further. Of the cancers diagnosed after cimetidine treatment, 7 were adenocarcinoma of the fundus, body, or antrum of the stomach, while 7 patients had tumours of the cardia and/or oesophagus, of which were adenocarcinoma and were of unknown histology (Table ). The remaining were squamous cell tumours of the oesophagus. Table shows the distribution of the intervals between treatment with cimetidine and diagnosis: in the great majority of patients, this interval was less than four years. Downloaded from by guest on April, TABLE. Numbers of cases of stomach and oesophageal carcinoma diagnosed postcimetidine treatment by centre Category Glasgow Nottingham Oxford Portsmouth Total () (9) () () (99) Adenocarcinoma of stomach (fundus, body, 7 antrum) Adenocarcinoma of cardia and/or oesophagus 7 Squamous cell carcinoma of oesophagus Cell type not known carcinoma of cardia and/or oesophagus Total

3 Postcimetidine Surveillance for up to Ten Years TABLE. Stomach and oesophageal carcinoma in relation to interval between treatment and diagnosis Category Time between Ifirst exposure to cimetidine anci data of diagnosis (months) 7^ Total Adenocarcinoma of stomach Adenocarcinoma of cardia and/or oesophagus Squamous cell carcinoma of oesophagus Cell type not known carcinoma of cardia and/or oesophagus Total 9 7 Observed and Expected Deaths Of the patients with gastric or oesophageal cancer diagnosed after receiving cimetidine, six are still alive (five gastric cancer and one oesophageal). Careful review of all hospital and general practitioner records suggested that in patients the underlying cause of death was at variance with that given on the death certificate. These two factors should be borne in mind when examining the observed to expected ratios (which are based entirely on the death certificate diagnoses as shown in Table for the first years of the study []). Nonetheless, the ratios give a good indication of overall trends. The ratios have declined to around unity for gastric cancer, but after a fall in oesophageal cancer during years, there has been a rise in years 7 and. Downloaded from by guest on April, TABLE. Observed and expected deaths cancer of stomach and oesophagus* Survey year Cancer of stomach: Cases Expected. Ratio.7***..**..7..* Cancer of oesophagus: Cases 7 Expected.. Ratio.** *.. 7*. * /?<. ***/7<. Expected numbers calculated from death rates for England and Wales for the appropriate year, allowing for sex and for age (year groups). Underlying cause of death as defined by Office of Population, Censuses and Surveys.

4 DG ColinJones and others Cancer of the stomach mortality England and Wales a.. IO ^ D^ a c a. A. D A» o o A A 7 D D 7 ~" D "~ a o O_ A IOO Cancer of the oesophagus mortality England and Wales o o. A A 7 D D 7 ^ *^ O _ c a a *"" a a o L Year I I I I I I I I I I I Year FIG.. Mortality from cancer of the stomach and oesophagus in England and Wales, 99. National Mortality Rates Mortality rates from cancer of the oesophagus in England and Wales (which practically equate with incidence rates) show a steady increase during the period under review, especially in the younger age groups (Fig. ). By contrast, death rates from cancer of the stomach have fallen markedly in recent years, especially in younger individuals (Fig. ). H r receptor Antagonist Usage Details of cimetidine consumption were available in of the 9 patients with tumours which were gastric or possibly gastric in origin. The amount taken varied from g to g (median g). Courses of cimetidine for the 9 patients were given within a variable time span ranging from. to months (median. months) after starting treatment before the diagnosis of carcinoma was made (Table ). One patient was also known to have received ranitidine prior to the diagnosis of carcinoma. Of 9 questionnaires sent out in 97 as a per cent sample of survivors, 9 (7. per cent) were returned. Of these 9 patients, (7 per cent) did not take an H receptor antagonist in the year under review, ( per cent) had taken cimetidine in at least one quarter, ( per cent) had taken ranitidine and ( per cent) had taken another H receptor antagonist (some patients took more than one H receptor antagonist). Usage differed between the four centres during 97 in Glasgow per cent had received an H receptor antagonist, whilst in Portsmouth only 7 per cent were so treated. Downloaded from by guest on April, Location of Tumour and Previous Dyspeptic History The sites of the tumours were unremarkable (Table ). In some instances it was impossible to

5 Postcimetidine Surveillance for up to Ten Years 7 TABLE. Tumour staging of patients with gastric and oesophageal tumours diagnosed after cimetidine treatment Site Stage 'Early' II III/ IV Not known Total Stomach: fundus, body, antrum Adenocarcinoma of cardia and/or oesophagus Carcinoma of cardia and/or oesophagus: cell type not known () () 7 Total 9 9 ( ) = Number of surviving patients determine whether the tumour was oesophageal or gastric in origin and we have, therefore, included all cases of adenocarcinoma of the oesophagus and/or cardia in a category of their own ( patients). There were five further cases of tumour in the oesophagus and/or cardia region where the cell type was not known. Of the 7 tumours known definitely to be gastric in origin, seven were in the fundus, in the body, nine in the body/antrum, in the antrum, six involved the whole stomach, and in three patients the exact site within the stomach was uncertain. Thirtythree of the 9 patients were known to have had no previous history of gastric disorders (and in one further patient no information was available). Of the remaining, duodenal ulcer had previously been recorded in patients (plus one further patient in whom a duodenal ulcer was diagnosed concurrently with the diagnosis of carcinoma), gastric ulcer in 7 patients, duodenal and gastric ulcer in seven patients, and hiatus hernia (only) in eight patients. The remaining three patients had mucosal inflammation. Of the tumours known to have involved the antrum, 7 patients had a previous history of duodenal ulcer. In of the patients with a previous history of gastric complaints, the length of history was known; it ranged from three months to years (median. years) prior to the diagnosis of cancer. Downloaded from by guest on April, Early Gastric Cancer Six patients (. per cent of 9) had early gastric cancer []; four have been reported previously [] (Table ). In the two further cases, one (male, years) had received g of cimetidine over months for a duodenal ulcer (perforated in 9). This patient had a barium meal 9 months prior to the diagnosis of carcinoma, when an hiatal hernia and duodenal deformity were found. The other patient (male, years) had received at least g of cimetidine over 7 months for a gastric ulcer documented radiologically. In both patients a benignlooking gastric ulcer had been found at endoscopy but the biopsies revealed malignancy. Both were mucosal cancers with negative lymph nodes. Ninetyone of the 9 patients had advanced disease at the time of diagnosis, of whom nine had stage II tumours (invading the muscularis propria without full penetration or lymph node involvement). In one patient no details concerning staging were available (Table ).

6 DG ColinJones and others DISCUSSION Concern has been expressed that the widespread use of drugs which reduce gastric acid output might predispose to gastric cancer []. Although we found no cause for concern when we reviewed our experience over four years of nearly patients who had received cimetidine [], we considered it important to continue with the review as any carcinogenic effect might have a prolonged induction. It can be difficult to determine the exact origin of an adenocarcinoma which lies at the cardia, so we have separated the cancers into adeno, squamous and unknown cell types; the adenocarcinomas have again been subdivided into those that are clearly gastric in origin and those that could be either gastric or oesophageal. Our data cover up to years' surveillance and we find it reassuring that the mortality ratio for gastric cancer diminished during the observation period. This supports our original suggestion [] that the cases of gastric cancer previously reported were already present when cimetidine was prescribed [7]. We believe the identification of cancer using death certificates and cancer registration to be complete for up to seven years of surveillance. This belief is supported by the results of searching computerized histopathology records twice at intervals in one centre (Portsmouth) and once in another (Nottingham). We failed to identify any cases not already notified from death certificates and cancer registration. The rising incidence of oesophageal cancer in England and Wales is generally regarded to be a consequence of trends over the years in tobacco and alcohol consumption [] (Fig. ). The population in our study received cimetidine for dyspeptic symptoms, so many would have had reflux oesophagitis, as it is now the most common endoscopic diagnosis. The prevalence of Barrett oesophagus in our cohort is unknown, but this condition, in which there is a spread of columnar epithelium proximally, is associated with an increased risk of malignancy [9]. The population under study is selected by a perceived need for cimetidine treatment and so is not typical of the general population. Furthermore, the symptoms of benign and malignant disease overlap [], investigations are not infallible [], and thus oesophageal as well as gastric cancer cases may have received treatment with antisecretory agents. Such a bias should diminish with duration of observation, although it could last for several years since early gastric cancer has been shown to be confined to the mucosa for more than six years in some cases [, ], sometimes with a long history of dyspepsia [7]. In our patients who developed gastric or oesophageal adenocarcinoma, of 9 ( per cent) were known to have had a prior history of gastroduodenal disease. This makes accurate diagnosis of cancer on the basis of the history difficult and also, incidentally, confirms that a gastric cancer can develop with a past history of duodenal ulcer []. Our survey of per cent of the cohort (over a period of one year) shows that about onethird are continuing to take an H receptor antagonist either intermittently or continuously. It has been suggested that by reducing intragastric acidity with drugs such as H receptor antagonists, bacteria will proliferate and convert dietary nitrates and nitrites to potentially carcinogenic nitrosamines [, ]. This has been disputed [] and there is no support for this suggestion from a UK survey which shows no correlation between levels of salivary nitrate and nitrite and the incidence of gastric cancer [7]. Furthermore, there has been a fall in the overall incidence of gastric cancer since cimetidine became available in 977 (Fig. ), despite continued widespread prescribing. The fall in the incidence of gastric cancer in our group for years after exposure to cimetidine is close to that predicted if there were no association between cimetidine and gastric cancer. Thus, for the short and medium term after exposure, ourfindingssupport the view that cimetidine does not induce gastric cancer. Deaths from oesophageal cancer initially fell but in years 7 and have shown a small rise, which clearly needs further monitoring to assess its significance. Downloaded from by guest on April,

7 ACKNOWLEDGEMENTS Postcimetidine Surveillance for up to Ten Years 9 We thank Mrs M Edmond, Mr T Lucas, Mrs M Rillie, and Mrs S Wood for their work in conducting the study; Mr J Beresford and Mrs P de Rivaz for help with computing; Dr AC Flind for his help; and the many general practitioners, receptionists, and pharmacists for their enthusiastic cooperation in this project. We are indebted to the Cancer Registries and Office of Population and Censuses and Surveys, and Family Practitioner Committees for their continued assistance. The study was supported by a grant from Smith Kline Beecham Pharmaceuticals. REFERENCES. ColinJones DG, Langman MJS, Lawson DH, Vessey MP. Review: postmarketing surveillance of the safety of cimetidine the problems of data interpretation. Aliment Pharmacol Ther 97; : 77.. ColinJones DG, Langman MJS, Lawson DH, Vessey MP. Cimetidine and gastric cancer: preliminary report from postmarketing surveillance study. Br Med J (9); :.. ColinJones DG, Langman MJS, Lawson DH, Vessey MP. Postmarketing surveillance of the safety of cimetidine: month mortality report. Br Med J 9; : 77.. Tasaka S. Statistical study of early gastric cancer collected throughout Japan. Gastroenterol. Endoscopy (Tokyo) 9; :.. Elder JB, Ganguli PC, Gillespie IE. Cimetidine and gastric cancer. Lancet 979;.. ColinJones DG, Langman MJS, Lawson DH, Vessey MP. Postmarketing surveillance of the safety of cimetidine: Mortality during second, third, and fourth years of follow up. Br Med J 9; 9:. 7. Green PHR, O'Toole KM, Weinberg LM, Goldfarb JP. Early gastric cancer. Gastroenterology 9; : 7.. Chilvers C, Fraser P, Beral, V. Alcohol and oesophageal cancer: an assessment of the evidence from routinely collected data. J Epidemiol Commun Health 979; : Spechler SJ, Goyal RK. Barrett's oesophagus. New Engl J Med 9; : 7.. O'Brien MJ, Burakoff R, Robbins EA, Golding RM, Zamcheck N, Gottlieb LS. Early gastric cancer: clinicopathologic study. Am J Med 9; 7: 9.. Deakin M, ColinJones DG, Vessey MP. Routine practice in the diagnosis of adenocarcinoma of the stomach: A survey of tumours diagnosed in the Portsmouth and Oxford Health Districts Postgrad Med J 9; : 7.. Tsukuma H, Mishima P, Oshima A. Prospective study of early gastric cancer. Int J Cancer 9; :.. Guanrei Y, Songliang Q, He H, Guizen F. Natural history of early oesophageal squamous carcinoma and early adenocarcinoma of the gastric cardia in the People's Republic of China. Endoscopy 9; : 99.. Ellis DJ, Kingston RD, Brookes VS, Waterhouse JAH. Gastric carcinoma and previous peptic ulceration. Br J Surg 979; : 79.. Reed PI, Haines K, Smith PLR, House FR, Walters CL. Gastric juice Nnitrosamines in health and gastroduodenal disease. Lancet 9;.. Hall CN, Darkin D, Brimblecome, R, Cook AJ, Kirkham JS, Northfield TC. Evaluation of the nitrosamine hypothesis of gastric carcinogenesis in precancerous conditions. Gut 9; 7: Coggon D, Acheson ED. The geography of cancer of the stomach. Br Med Bull 9; :. Downloaded from by guest on April,