H2-receptor antagonists may increase the risk of cardio-oesophageal adenocarcinoma: a case-control study

Adenocarcinoma of the lower oesophagus and the gastric cardia has shown a dramatic worldwide increase in incidence over the last 25 years, but the cause is unknown. A large number of drugs have been introduced over this period of time, and it has been suggested that drugs that relax the lower oesophageal sphincter (DRLOS) might be causative, and on the other hand that non-steroidal anti-inflammatory drugs (NSAIDS) may be protective. H2-receptor antagonists (H2RAs) may allow achlorhydric reflux to continue without symptoms, and it is postulated that such asymptomatic reflux is uncontrolled by the usual conservative measures and may lead to increased oesophageal damage. H2RAs were first marketed in 1970 and might be the cause of the observed increase of cardio-oesophageal adenocarcinoma (COA). In a case-control study, the records of 56 subjects who died of COA in the period 1 January 1990 to 31 December 1992 were compared with those of 56 age-/sex-matched controls who died of myocardial infarction. They were 28 females and 84 males, mean age 69.8 years. The NHS records containing the lifetime prescription history of each subject were retrieved from the health authority. Each prescription was recorded, omitting drugs taken in the two years before diagnosis. Analysis was performed using conditional logistic regression. Other variables, including the use of antacids, steroids, smoking and alcohol, were also examined. Subjects dying of COA were more likely to have consumed H2RAs (relative risk (RR) 7.50, 95% CI 1.33-42.09, P < 0.02). On the other hand, they were less likely to have consumed NSAIDs (RR 0.16, 95% CI 0.03-0.93, P < 0.04) or DRLOS (RR 0.14, 95% CI 0.02-1.0, P = 0.05). This study supports a protective effect from NSAIDs against COA, but the similar effect of DRLOS is related to the increased use of cardiac drugs in the control group. H2RAs appear to have a harmful effect, which may be related to the worldwide increase in COA. However, the trend may have been apparent before cimetidine was widely available, and it is possible that the cause is multi-factorial.