Combination of circulating tumor cell enumeration and tumor marker detection in predicting prognosis and treatment effect in metastatic castration-resistant prostate cancer

Abstract

Although circulating tumor cell (CTC) enumeration in peripheral blood has already been validated as a reliable biomarker in predicting prognosis in metastatic castration-resistant prostate cancer (mCRPC), patients with favorable CTC counts (CTC < 5/7.5 ml) still experience various survival times. Assays that can reduce patients’ risks are urgently needed. In this study, we set up a real-time quantitative polymerase chain reaction (RT-qPCR) method to detect epithelial-mesenchymal transition (EMT) and stem cell gene expression status in peripheral blood to validate whether they could complement CTC enumeration. From January 2013 to June 2014 we collected peripheral blood from 70 mCRPC patients and enumerated CTC in these blood samples using CellSearch system. At the same time, stem cell-related genes (ABCG2, PROM1 and PSCA) and EMT-related genes (TWIST1 and vimentin) were detected in these peripheral blood samples using an RT-qPCR assay. Patient overall survival (OS) and treatment methods were recorded in the follow-up. For patients who received first-line chemotherapy, docetaxel plus prednisone, PSA progression-free survival (PSA-PFS) and PSA response rate were recorded. At the time of analysis, 35 patients had died of prostate cancer with a median follow-up of 16.0 months. Unfavorable CTC enumerations (CTC ≥5/7.5 ml) were predictive of shorter OS (p = 0.01). Also, positive stem cell gene expression indicated poor prognosis in mCRPC patients (p = 0.01). However, EMT gene expression status failed to show any prognostic value in OS (p = 0.78). A multivariate analysis indicated that serum albumin (p = 0.04), ECOG performance status (p < 0.01), CTC enumeration (p = 0.02) and stem cell gene expression status (p = 0.01) were independent prognostic factors for OS. For the 40 patients categorized into the favorable CTC enumeration group, positive stem cell gene expression also suggested poor prognosis (p < 0.01). A combined prognostic model consisting of stem cell gene expression and CTC enumeration increased the concordance probability estimated value from 0.716 to 0.889 in comparison with CTC enumeration alone. For patients who received docetaxel plus prednisone as first-line chemotherapy, positive stem cell gene expression suggested a poor PSA-PFS (p = 0.01) and a low PSA response rate (p = 0.008). However, CTC enumeration and EMT gene expression status did not affect PSA-PFS or PSA response rates. As a result, detection of peripheral blood stem cell gene expression could complement CTC enumeration in predicting OS and docetaxel-based treatment effects in mCRPC patients.