Background: Response of metastatic sites on computed tomography (CT) has been used to assess response in metastatic gastric carcinoma (MGC); however, the role of endoscopy to evaluate the response of the primary gastric lesion is unclear. We undertook a prospective study to compare prognostic values of endoscopy-based response criteria with those of CT-based response criteria in MGC patients treated with 5-fluorouracil (5-FU) and cisplatin.
Methods: MGC patients naïve to chemotherapy were treated with 5-FU (1000 mg/m(2), days 1-5) and cisplatin (60 mg/m(2), day 1) (FP) every 21 days. Response was assessed by CT [World Health Organization (WHO) criteria/Response Evaluation Criteria In Solid Tumors (RECIST)] every three cycles and/or endoscopy after the third cycle.
Results: With a median follow-up of 26.2 months, 103 patients were assessed by CT and endoscopy. There was good concordance between the WHO and RECIST criteria (kappa = 0.91; P = 0.0001), but poor agreement between CT and endoscopic assessments (kappa = 0.17; P = 0.01). On multivariate analysis, both CT (WHO/RECIST) (hazard ratio, 5.20; P < 0.0001) and endoscopic response (hazard ratio, 2.78; P < 0.0001) were significantly associated with survival. The combination of CT and endoscopy defined patients into four groups with distinct prognoses according to chemotherapy response: responders on both tests, responders on CT alone, responders on endoscopy alone, and nonresponders on both tests (hazard ratio, 1.00 versus 2.87 versus 5.35 versus 12.3; P < 0.0001).
Conclusions: The use of endoscopy together with CT provides a better prognostic tool than CT alone in MGC patients treated with FP.