Factors influencing noncompliance and contamination in a randomized trial of 'Western' (r1) versus 'Japanese' (r2) type surgery in gastric cancer

Background. A randomized trial was undertaken comparing the Western R1 resection with limited N1-level lymphadenectomy and the Japanese R2 resection with extended lymphadenectomy, including the N2-level for curative resection of gastric cancer patients. After 389 patients were entered in the trial, protocol deviations were observed that reduced the intended distinction between the two types of lymphadenectomy: noncompliance, i.e., no substantiation of lymphadenectomy by nodal yields at indicated stations, and contamination, i.e., extension of lymphadenectomy outside the allocated level of nodal clearance. Methods. To identify factors underlying these protocol deviations, the authors analyzed the influence of six patient-, tumor, and treatment-related characteristics on the magnitude of deviations per patient, and on the incidence of deviations per lymph node station. Results. Protocol deviations were influenced by the following station-specific factors: (1) the number of nodes per station; (2) the clarity of anatomical station definition; (3) the location of stations; (4) local conventions on the type of gastrectomy; and (5) technical features to allow complete en bloc dissection. Furthermore, nonspecific factors such as inadequate retrieval of nodes, incomplete dissection, and careful selection of clinically overt metastases outside the allocated level of nodal clearance were randomly distributed over stations, and they, too, contributed to the deviations. Conclusions. Based on the findings, the authors took additional steps to preserve the distinction between limited and extended lymphadenectomy and to improve the accuracy of nodal staging. These factors should be considered when standardization of both surgicopathologic trials and clinical protocols for the treatment of gastric cancer is pursued. Cancer 1994; 73:1544-51. Key words: randomized controlled trial, stomach neoplasms, quality control, lymph node excision, neoplasm staging.