Survival of patients in a phase 1 clinic: The M. D. Anderson Cancer Center experience

BACKGROUND: Patients with advanced malignancies for whom standard therapy is ineffective may participate in phase 1 trials. To gain a better understanding of the clinical features that could influence benefit versus risk, the authors of this report assessed prognostic factors and survival for patients who were referred to a phase 1 clinic focused primarily on targeted agents, METHODS: The medical records of 200 sequential patients who presented to the Phase 1 Clinic at The University of Texas M. D. Anderson Cancer Center were reviewed, and their characteristics and survival were analyzed. RESULTS: The median patient age was 58 years (range, 12-85 years), and 57% of patients were men. The median number of prior therapies was 4. Of 200 patients, 182 were treated on at least 1 phase 1 clinical trial. The median follow-up of surviving patients was 21 months, and the median overall survival was 9 months (95% confidence interval [Cl], 7.4-10.8). In univariate analysis, the factors that predicted shorter survival were primary tumor in the gastrointestinal tract; a history of thrombosis, liver metastases, and elevated levels of serum lactate dehydrogenase; platelet count; carbohydrate antigen 9 (Ca19-9) and Ca-125 levels; aspartate aminotransferase levels, and alkaline phosphatase levels (P < .05 for each). In multivariate analysis, independent factors that predicted shorter survival were a history of thromboembolism (hazard ratio [HR], 2.38; 95% Cl, 1.29-4.39; P = .005), platelets ≥440 × 10 ⁹/L (HR, 1.72; 95% Cl, 1.12-2.65; P = .014), and the presence of liver metastases (HR, 1.51; 95% Cl, 1.09-2.09; P = .013), CONCLUSIONS: Patients who were referred to phase 1 studies had a short median survival (9 months). Patients with thrombocytosis, liver metastases, and a history of thromboembolism had worse outcomes. A prognostic score is proposed.