Adenocarcinoma of the Oesophagus and Oesophagogastric Junction: Analysis of Incidence and Risk Factors

Discussion

In our study, with a follow-up of nearly 30 years, the age-adjusted incidence of EAC for both sexes was found significantly increased in Finland. Our results support previous reports with shorter follow-ups conducted in our country, that have found that the incidence of EAC in Finland is increasing in the long term.

Previously, the incidence of EAC was found to have increased among Finnish males during 1977-1995, but no increase was found among the female population (8). This, together with our results, supports the finding that the incidence of EAC among male individuals is increasing during long-term follow-up. Contrary to the findings by Sihvo et al. (7), a significant increase was also found in our study in the incidence of EAC among the female population. The increase in EAC among women is now seen for the first time in our country. The reasons for the increase are unknown. In accordance with our material from the Pirkanmaa Hospital District, the incidence of EGJAC was higher during the entire study time in the whole country. However, the incidence of EGJAC in Finland decreased significantly from the first to the last period, which is in accordance with some earlier reports (1).

According to the literature, male predominance has been found for EAC and EGJAC, being highest in the age group of 50-59 (8). We also encountered a male predominance, but the male-to-female ratio in our study was smaller than in previous ones. In addition, we found that females developed EAC or EGJAC at a significantly later age than males, suggesting that oestrogens might protect females. However, hormone replacement therapy was not found to be associated with EAC (16), which contradicts the previous speculations on the preventive role of oestrogens in EAC or EGJAC. Exposure to some of the established risk factors for EAC seems to be more common in men than in women, and perhaps the difference in the effect of these factors can better explain the difference in the incidence of EAC and EGJAC between men and women (17, 18). Our results further support this speculation, since we found more risk factors for EAC – including smoking, GERD, BE and previous antireflux surgery – in men.

According to multivariate analysis, we found two independent risk factors, BE for EAC and cholecystectomy for EGJAC. The association of BE with EAC is not a new discovery. However, the association of previous cholecystectomy with EGJAC has not, to our knowledge, been reported before. The incidence of cholecystectomy among the EGJAC patients in our study was four and, among the EAC patients, 1.5-times higher than among the non-affected Finnish population of the same age. In fact, among the female EGJAC patients, the incidence of cholecystectomy was more than 10-times higher. The number of patients who had undergone cholecystectomy might be even higher, since we analyzed this aspect in only 158 patients. Previously, a risk of distal gastric cancer among male individuals has been detected for up to 10 years after cholecystectomy, but no clear association between cholecystectomy and EGJAC was found (19). Instead, cholecystectomy was associated with a moderate risk (20) of EAC. The risk was speculated to be caused by the toxic effect of refluxed duodenal juice (20). Previously, we found that cholecystectomy had been performed for 28% of the patients who were diagnosed to have EAC after antireflux surgery (14). The cholecystectomy preceded the diagnosis of EAC by an average of 17.5 years. Although we did not report the coincidence of fundoplication failure with cholecystectomy in that study, most of the patients who developed EAC late after fundoplication had a functional failure (14). This further supports the speculation that the risk of cholecystectomy relates to the toxic effect of duodenal juice.

For 34% of EAC and EGJAC patients for whom the stage of disease was known, the cancer was in the early stage (Stage I/IIA), while among BE patients in whom the cancer was discovered during BE surveillance, the cancer was in early state in 73% of the patients. This suggests that BE surveillance may be beneficial. However, the BE was known before the cancer diagnosis in only 11.7% and 2.6% of EAC and EGJAC patients, respectively, while only 4.3% of the cancer cases were detected in BE surveillance. In the study by Voutilainen et al. (2005), two EAC patients (18.2%) had BE, and one of them was identified during BE surveillance and could be operated on. In the literature, the annual risk of EAC among BE patients has been found to be 0.12%, which is much lower than the assumed risk of 0.5% that has been the basis for current surveillance guidelines in BE (22). It seems that patients detected during BE surveillance are in an earlier stage of disease than others but that the cancer risk in BE is lower than previously assumed. Therefore, we should find other risk factors that will allow us to pinpoint those BE patients who benefit from surveillance and, on the other hand, to identify those who may have BE but remain undetected. In addition, the clinicians should consider the patients' symptoms more seriously, since the mean duration of cardinal symptoms (most commonly dysphagia, heartburn, regurgitation, weight loss) among our patients was as long as 18.9 weeks.

Our study has certain limitations. The most important one is the retrospective nature of the study. It is impossible to recover all risk factors comprehensively from documents. In addition, tumour classification in the region of the EGJ is sometimes difficult to define retrospectively. Finally, the classification of these tumours has changed at least three times over the course of the study period. However, tumour classification was carried out using as many investigation methods as possible, and all classifications were made by the first author who specialises in oesophagogastric surgery. In addition, all classifications were based on the 7th edition of the AJCC cancer staging manual (13). The strength of our study is that we could analyze the risk factors of EAC and EGJAC much more profoundly than has been done in previous studies, which are mostly population-based. In addition, to our knowledge, this is the first study from Finland in which the risk factors of EAC and EGJAC have been evaluated. Finally, our EAC and EGJAC incidence rates in Finland were based on data from the Finnish Cancer Register, which has been regarded as reliable.

In conclusion, in accordance to the previous report from our country, a significant increase in EAC among Finnish male individuals was found over the course of nearly 30 years, indicating that the increase in EAC among Finnish men is existent in the long term. However, a significant increase in EAC was also found in the female population, which is now documented for the first time in Finland. We found a previously performed cholecystectomy to be associated with EGJAC, which is, to the best of our knowledge, reported for the first time herein.