Subject, methods, and results

Through the inpatient registry, which in 1983 covered 85% of the Swedish population, we identified 67 812 patients admitted to hospital between 1965 and 1983 for peptic ulcer disease but who did not have a vagotomy. Through the same registry we also identified 7198 patients who had a vagotomy between 1971 and 1979. Through linkage with the Swedish death and emigration registry as well as the Swedish cancer registry, all new cases of lung cancer in the two cohorts were identified until the end of 1989. Expected numbers of new cases were estimated from age specific and period specific population rates.

After we excluded the first year after vagotomy, the ratio of observed to expected cases up to the end of the follow up was 2.20 (95% confidence interval=1.82 to 2.63), with an increase in the ratio from 1.86 (one to five years after operation) to 2.52 (10 years or more after operation). Among the patients with peptic ulcer disease who had not had a vagotomy the ratio of observed to expected cases was 1.56 (1.49 to 3.67), with a slight decrease after the first five years (1).

Comment

We found that, although patients with peptic ulcer disease had an increased risk of lung cancer, the excess risk was substantially lower five years after inclusion in the study among those who had not had a vagotomy than among those who had. This suggests that those who had not had a vagotomy might have reduced their level of smoking, as the risk of lung cancer has a clear dose-response pattern with regard to tobacco use.4 This reduction in risk may be the result of persisting symptoms and antismoking counselling. In the patients who had had a vagotomy the risk of lung cancer increased, suggesting a continuous or increased exposure to tobacco after the operation. Other underlying biological mechanisms, however, cannot be ruled out (for example, changes in diet as the result of surgery).

The strengths of the study are its population based setting, complete follow up, and the reliable classification of outcomes through the Swedish registries. Limitations, however, include diagnostic misclassification of peptic ulcer disease—more likely among patients who did not have a vagotomy than among those who did—especially during the pre-endoscopic era. Moreover, patients who had a vagotomy probably had a more severe disease. However, the ratios of observed to expected cases of lung cancer were similar (P=0.70) during the first five years after inclusion in the study, indicating that the percentage of smokers and the magnitude of tobacco consumption were similar. The difference in risk of lung cancer was most evident 10 years or more after inclusion; this result is similar to results from other studies5 and consistent with a latency period of at least five to 10 years before changes in smoking habits would affect the risk of lung cancer.

As increasingly effective long term drug treatments are developed, less attention might be being paid to the role of smoking in peptic ulcer disease. Extrapolation of our results in Sweden implies that an excess of 50 to 100 lung cancers per 10 000 patients with peptic ulcer disease would occur in a 15 year period in cured patients. If this excess risk had been ascribed directly to the operation or to curative drug treatment these treatment methods would have been ultimately abandoned. Physicians should therefore be urged to include antismoking counselling when treating patients with peptic ulcer disease.