The aim of the study was to analyze epidemiologic parameters, treatment-related data and prognostic factors in the management of gastric cancer patients of a university surgical center under conditions of routine clinical care before the onset of the era of multimodal therapies. By analyzing our data in relation with multi-center quality assurance trials [German Gastric Cancer Study - GGCS (1992) and East German Gastric Cancer Study - EGGCS (2004)] we aimed at providing an instrument of internal quality control at our institution as well as a base for comparison with future analyses taking into account the implementation of evolving (multimodal) therapies and their influence on treatment results.
Material and methods: Retrospective analysis of prospectively gathered data of gastric cancer patients treated at a single institution during a defined 10-year time period with multivariate analysis of risk factors for early postoperative outcome.
Results: From 04/01/1993 through 03/31/2003, a total of 328 gastric cancer patients were treated. In comparison with the EGGCS cohort there was a larger proportion of patients with locally advanced and proximally located tumors. 272 patients (82.9%) underwent surgery with curative intent; in 88.4% of these an R0 resection was achieved (EGGCS/GGCS: 82.5%/71.5%). 68.2% of patients underwent preoperative endoluminal ultrasound (EUS) (EGGCS: 27.4%); the proportion of patients undergoing EUS increased over the study period. Diagnostic accuracy of EUS for T stage was 50.6% (EGGCS: 42.6%). 77.2% of operated patients with curative intent underwent gastrectomy (EGGCS/GGCS: 79.8%/71.1%). Anastomotic leaks at the esophagojejunostomy occurred slightly more frequently (8.8%) than in the EGGCS (5.9%) and GGCS (7.2%); however, postoperative morbidity (36.1%) and early postoperative mortality (5.3%) were not increased compared to the multi-center quality assurance study results (EGGCS morbidity, 45%); EGGCS/GGCS mortality, 8%/8.9%). D2 lymphadenectomy was performed in 72.6% of cases (EGGCS: 70.9%). Multivariate analysis revealed splenectomy as an independent risk factor for postoperative morbidity and ASA status 3 or 4 as an independent risk factor for early postoperative mortality. The rate of splenectomies performed during gastric cancer surgery decreased substantially during the study period.
Conclusions: Preoperative diagnostics were able to accurately predict resectability in almost 90% of patients which is substantially more than the corresponding results of both the EGGCS and the GGCS. In the future, more wide-spread use of EUS will play an increasing role as stage-dependent differentiation of therapeutic concepts gains acceptance. However, diagnostic accuracy of EUS needs to be improved. Our early postoperative outcome data demonstrate that the quality standard of gastric cancer care established by the EGGCS is being fulfilled at our institution in spite of distinct characteristics placing our patients at higher surgical risk. Besides being a valuable instrument of internal quality control, our study provides a good base for comparison with ongoing analyses on future developments in gastric cancer therapy.