Determination of the optimal cutoff percentage of residual tumors to define the pathologic response rate (pathRR) of gastric cancer (GC) treated with preoperative therapy (JCOG1004-A).

Abstract

4104 Background: PathRR is a common endpoint used to assess the efficacy of preoperative therapy for GC. PathRR is estimated based on the percentage of the residual tumor area in the preexisting tumor bed. Various cut-off definitions that have used for past studies (e.g. 10%, 33%, 40%, 50%, 67%) often impair the comparability of pathRRs between studies. Methods: Individual patient data from four JCOG trials evaluating preoperative chemotherapy were used (JCOG0001, irinotecan+cisplatin, n=55; JCOG0002, S-1, n=55; JCOG0210, S-1+cisplatin, n=50; JCOG0405, S-1+cisplatin, n=53). Pathological specimens were evaluated from 173 out of 188 patients (92%) who underwent surgery. Residual and preexisting tumor areas were traced on a virtual microscopic slide by one pathologist and another confirmed these areas. The hazard ratio (HR) in overall survival was calculated for each cut-off percentage by stratified Cox regression analysis including the study as a stratification factor, and concordance probability estimates (CPE) were also calculated. Results: The numbers of patients with 0-10%/11-33%/34-50%/51-66%/67-100% residual tumors were 43/33/27/23/47, respectively. Overall, HR was the largest in the 10% cut-off and CPE was the largest in 33%. When patients with R1/R2 resections were excluded, both HR and CPE were the largest in the 10% cut-off. In subgroup analyses, almost all cut-offs predicted survival well regardless of the histologic type (intestinal/diffuse), and no cut-off predicted survival for type 4 (linitis plastica type) tumors. Conclusions: PathRR is not recommended for clinical trials including type 4 tumors. The 10% cut-off is recommended for non-type 4 tumors, though 33% is also applicable. [Table: see text]