[Clinical and pathological prognostic factors for cancers of the esophagogastric junction].

INTRODUCTION: Adenocarcinoma of the esophagogastric junction (AEG) is a particular tumour entity because two substantially different surgical procedures are required according to the location. There is no difference in long-term prognosis between the tumour types in spite of the different surgical procedures. We were interested to evaluate the clinical and pathological prognostic factors of the AEGs which were operated in our department.

PATIENTS AND METHODS: 108 patients were operated for AEG between 1.1.2000 and 1.4.2006 in our institution. 32 (29.6 %) patients with distal esophageal cancer (type I according to Siewert) underwent a transthoracic esophagectomy with gastric pull-up and two-field lymphadenectomy. 57 (52.8 %) patients with type II and 19 (17.6 %) patients with type III cancers received an extended gastrectomy with D2 lymphadenectomy. The retrospective analysis was focused on clinical and pathological parameters. Possible differences between the tumour types were also evaluated. Median follow-up was 11.4 months (range: 1-57 months).

RESULTS: Follow-up data were complete for 107 patients. A median survival of 17.4 +/- 3.25 months and a cumulative survival of 30 % were independent of the tumour location and the surgical procedure. Overall hospital mortality was 3.7 %. The univariate analysis showed that survival was significantly associated with the T category, lymph node status, lymphangio- and angioinvasion and tumour grading. In the multivariate analysis, only lymph node status was identified as an independent prognosis factor for survival. Where-as the R status was not a prognostic factor per se, how-ever, patients with an R0 situation without lymphangio- and angioinvasion had a significantly better survival compared to all other patients (p = 0.001). An increased angioinvasion rate was observed in type III tumours (52.6 %) in comparison to type I (21.9 %) and type II (21.1 %) tumours.

CONCLUSION: The prognostic factors of our patients determined substantially the prognosis of the patients. Patients with lymph- or haemangioinvasion should regarded as high-risk patients independent of the R status. Close oncological follow-up including potential adjuvant treatment in these patients is recommended.