Multivariate prognostic factor analysis in locally advanced and metastatic esophago-gastric cancer--pooled analysis from three multicenter, randomized, controlled trials using individual patient data.

PURPOSE To identify baseline prognostic factors and assess whether pretreatment quality of life (QoL) predicts survival in patients with locally advanced or metastatic esophago-gastric cancer. PATIENTS AND METHODS Between 1992 and 2001, 1,080 patients were enrolled into three randomized, controlled trials assessing fluorouracil-based combination chemotherapy. All patients were required to complete the European Organization for Research and Treatment of Cancer core QoL questionnaire before random assignment. RESULTS Of the 1080 patients randomly assigned, 979 (91%) died. Four independent poor prognostic factors were identified by multivariate analysis: performance status >or= 2 (hazard ratio [HR], 1.58; 99% CI, 1.25 to 1.98), liver metastases (HR, 1.41; 99%CI, 1.14 to 1.74), peritoneal metastases (HR, 1.33; 99%CI, 1.01 to 1.74) and alkaline phosphatase >or= 100 U/L (HR, 1.41; 99% CI, 1.14 to 1.76). A prognostic index was constructed dividing patients into good (no risk factor), moderate (one or two risk factors) or poor (three or four risk factors) risk groups. One-year survival for good, moderate, and poor risk groups were 48.5%, 25.7%, and 11%, respectively, and the survival differences among these groups were highly significant (P <.00001). Compared with the good risk group, the moderate risk group had nearly twice the risk of death, and the poor risk group had 3.5-fold increased risk of death. Pretreatment physical (P =.003), role functioning (P <.001), and global QoL (P <.001) predicted survival. CONCLUSION Four poor prognostic factors were identified and a simple prognostic index was devised. Information from this analysis can be used to aid clinical decision-making, help individual patient risk stratification, and serve as benchmark for the planning for future phase III trials.