Analysis of esophagogastric cancer patients enrolled in National Cancer Institute (NCI) Cancer Therapy Evaluation Program (CTEP)-sponsored phase I trials.

Abstract

12 Background: In phase I trials, an important entry criterion is life expectancy predicted to be more than 90 days, and this is notoriously difficult to predict. The Royal Marsden Hospital (RMH) prognosis score was developed to help project patient outcomes. At present, there have been no systemic analyses of clinical benefits and prognoses for esophagogastric cancer patients who were registered in phase 1 trials. Methods: All non-pediatric phase I oncology trials sponsored by NCI-CTEP, that began between 2001 and 2013 were considered in this review. The RMH score was determined by 3 variables: LDH level, albumin level, and number of metastatic sites of disease. A log-rank test was used to assess the differences of treatment duration and overall survival (OS) between the groups with different RMH scores. Multivariate Cox proportional hazards model was also used to select the covariates related to the prognosis. Results: Of 4,722 patients with solid tumors, 115 (2.4%) patients were eligible for our analysis. Median age was 59 (range: 30-85), and 73% of patients were men. 54 (47%) patients with esophagus cancer and 61 (53%) with stomach cancer were analyzed. 86 (75%) patients had good (0-1) RMH score and 29 (25%) had poor (2-3) RMH score. Although response rates for good and poor RMH patients were not significantly different (8% vs. 3%; 1-side Fisher's exact test p=0.35), clinical benefit rate (complete response +partial response +stable disease) were significantly different (49% vs. 17%; 1-side Fisher's exact test p=0.002). The median treatment duration and OS for good and poor RMH patients were significantly different, respectively (treatment duration: 86 days and 29 days, Log-rank test P=0.016; OS: 332 days and 63 days, Log-rank test P<0.001). In the multivariate analysis that included the RMH score, age (≥60), ECOG PS (≥2) and the RMH score (2-3) were significant predictors of poor survival. Conclusions: The RMH score is a strong tool to predict the prognosis of esophago-gastric cancer patients who will participate in a phase I trial. The median OS of 63 days for patients with poor RMH score suggests that these patients should not be included in phase 1 clinical trials.