D Olmos, RD Baird, TA Yap, C Massard, L Pope, SK Sandhu, G Attard, J Dukes, D Papadatos-Pastos, P Grainger, SB Kaye, JS de Bono
Clin Cancer Res
BACKGROUND: High circulating tumor cell (CTC) counts are associated with poor prognosis in several cancers. Enrollment of patients on phase I oncology trials requires a careful assessment of the potential risks and benefits. Many patients enrolled on such trials using established eligibility criteria have a short life expectancy and are less likely to benefit from trial participation. We hypothesized that the incorporation of CTC counts might improve patient selection for phase I trials. METHODS: This retrospective analysis evaluated patients who had baseline CTCs enumerated prior to their starting on a phase I trial. CTCs were enumerated using the CellSearch System. RESULTS: Between January 2006 and December 2009 a total of 128 patients enrolled in phase I trials had CTC counts evaluated. Higher CTC counts as a continuous variable independently correlated with risk of death in this patient population (P = 0.006). A multivariate point-based risk model was generated using CTCs as a dichotomous variable (≥3 or <3), and incorporated other established prognostic factors, including albumin <35 g/L, lactate dehydrogenase greater than upper limit of normal, and >2 metastatic sites. Comparison of receiver operating characteristic curves demonstrated that the addition of baseline CTC counts improved the performance of the prospectively validated Royal Marsden Hospital phase I prognostic score, which now identifies three risk groups (P < 0.0001): good prognosis [score 0-1, median overall survival (OS) 63.7 weeks], intermediate prognosis (score 2-3, median OS 37.3 weeks), and poor prognosis (score 4, median OS 13.4 weeks). CONCLUSION: CTC enumeration improved the performance of a validated prognostic score to help select patients for phase I oncology trials.