Purpose: Standardized enumeration of CEC counts is required to minimize variability and allow cross-studies comparisons. The purpose of this paper is to identify CEC threshold proposal, by CellSearch system, for determining response to bevacizumab-based chemotherapy in metastatic colorectal cancer.
Methods: From July 2007 to June 2008, 33 patients treated with FOLFOX4 plus bevacizumab were enrolled in a prospective study. From January 2007 to June 2007, before bevacizumab was approved by the government in Japan, 31 patients treated with FOLFOX4 as a control were enrolled. CECs of whole blood at the baseline, day 4, 2 weeks after initiation of chemotherapy were isolated and counted using CellSearch system.
Results: There was no correlation between CEC levels and the outcome in the FOLFOX4. In the bevacizumab-based chemotherapy, CEC levels at the baseline were significantly associated with the outcome. Patients with 65 or more CECs at the baseline had a shorter median PFS and OS, than the median PFS and OS of less than 65 CECs at the baseline in the bevacizumab-based chemotherapy (P = 0.003, P = 0.027, respectively). By univariate and multivariate Cox proportional-hazards regression, CEC levels (cut-off; 65) at the baseline indicated the strongest predictor for the outcome to bevacizumab-based chemotherapy.
Conclusion: A threshold of lower than 65 CECs, by the CellSearch System, at the baseline was a significant predictor of the outcome for colorectal cancer patients treated with bevacizumab-based chemotherapy.