This issue of Gut presents a paper from Osaka, Japan1 reporting the long term outcome of 71 patients diagnosed with early gastric cancer (EGC) before 1988, but who were not immediately operated on because of age, infirmity, or refusal (see page 618). Thirty six EGCs (51%) progressed to advanced cancer. After exclusion for incomplete follow up, a Kaplan-Meier plot showed the likelihood of progression over five years to be 63%. Thirty eight of the 71 patients were never operated on. Of these, 23 (61%) died from gastric cancer. The Kaplan-Meier plot predicted a five year survival in this group ranging from 63 to 68% depending on the allocation of patients in whom the outcome was unknown.

This is an important paper because few studies have followed the natural history of early gastric cancer and it will be difficult to repeat in the future because of the development of endoscopic mucosal resection (EMR) which enables elderly and infirm patients with EGC to be effectively treated without recourse to major gastric surgery.

The conclusion drawn by the authors is that EGC is one step in an inevitable progression to advanced cancer and eventual death. The data, however, do not entirely support this view. Only 51% of the inception cohort were actually shown to have developed advanced cancer after a follow up period of more than 10 years. In those not operated on and where the true natural history can be deduced, only 23 (61%) died from gastric cancer and the five year survival in this group was over 60%. Furthermore, these values have to be interpreted with the knowledge that EGC can be reliably identified only on resected surgical or autopsy specimens. The authors accept that the accuracy of an EGC diagnosis at endoscopy is around 80%, thus 20% of the inception cohort were likely to have harboured advanced stage cancer when recruited into the study. The implication of these data is that a significant minority of elderly or infirm patients diagnosed with EGC will, if left alone, die from something other than their gastric cancer. Parallels can be drawn with prostatic cancer, a condition that may remain dormant for a long period. What is less often appreciated is that certain melanomas behave in a similar manner.2 In situ cancer of the cervix often fails to progress3 and as many as 70% of breast neoplasms never develop into a clinically relevant disease.4 Premalignant colonic polyps have a long gestation period before showing invasive tendencies so it would be surprising if all EGCs were invasive from inception. It follows that the more effective a screening policy is in detecting very early lesions the less likely it is that the lesion has the potential to cause death.

The proportion of patients in Japan diagnosed with EGC is rising as a percentage of the gastric cancer load. This reflects the remarkable skill and considerable experience of Japanese endoscopists and also the investment that has been made in improving endoscopic technology. The proportion of EGC being detected at an even earlier stage is also rising with so called “gastritis like” cancer representing more than 50% of EGC in some units.5 This means that most cancers are identified before becoming ulcerated and implies that a greater proportion are at the intramucosal as opposed to the submucosal stage of development. These earlier cancers carry a better prognosis6 so it is possible that a higher proportion will take longer to progress and may be less clinically relevant in elderly patients.

What lessons can be drawn from this paper? Firstly, there can be no doubt that most early gastric cancers if not treated will eventually develop into advanced cancer and ultimately kill the patient. Secondly, a significant proportion of elderly patients with early gastric cancer will die of other diseases before their cancer becomes a clinical problem. There may be subtypes of EGC that do not progress or progress so slowly that even over a prolonged period they do not become invasive.

From a research perspective it is important to identify those factors that cause EGC to progress from a mucosal lesion to a submucosal lesion, local invasion, and metastasis. This information may enable us to develop strategies to prevent what is, at present, considered an inevitable process. From a practical clinical perspective, the data suggest that radical surgery is not necessarily the correct approach in all cases of endoscopically diagnosed EGC. We must follow the lead given by our Japanese colleagues in developing the technique of EMR which in suitable cases enables EGC to be removed in its entirety thus reducing the morbidity and mortality associated with radical surgery7 the treatment used in most of these cases at present.