Abstract
BACKGROUND: Intermediate or surrogate endpoints for survival can shorten time lines for drug approval. We aimed to assess circulating tumour cell (CTC) count as a prognostic factor for survival in patients with progressive, metastatic, castration-resistant prostate cancer receiving first-line chemotherapy. METHODS: We identified patients with progressive metastatic castration-resistant prostate cancer starting first-line chemotherapy in the IMMC38 trial. CTCs were isolated by immunomagnetic capture from blood samples at baseline and after treatment. Baseline variables, including CTC count, titre of prostate-specific antigen (PSA), and concentration of lactate dehydrogenase (LDH), and post-treatment variables (change in CTCs and PSA) were tested for association with survival with Cox proportional hazards models. Concordance probability estimates were used to gauge discriminatory strength of the informative factors in identifying patients at low-risk and high-risk of survival. FINDINGS: Variables associated with high risk of death were high LDH concentration (hazard ratio 6.44, 95% CI 4.24-9.79), high CTC count (1.58, 1.41-1.77), and high PSA titre (1.26, 1.10-1.45), low albumin (0.10, 0.03-0.39), and low haemoglobin (0.72, 0.64-0.81) at baseline. At 4 weeks, 8 weeks, and 12 weeks after treatment, changes in CTC number were strongly associated with risk, whereas changes in PSA titre were weakly or not associated (p>0.04). The most predictive factors for survival were LDH concentration and CTC counts (concordance probability estimate 0.72-0.75). INTERPRETATION: CTC number, analysed as a continuous variable, can be used to monitor disease status and might be useful as an intermediate endpoint of survival in clinical trials. Prospective recording of CTC number as an intermediate endpoint of survival in randomised clinical trials is warranted.