Detection of circulating tumor cells in different stages of prostate cancer

Abstract

Purpose: To explore circulating tumor cell (CTCs) counts in different stages of prostate cancer (PC) in association with tumor burden, metastatic pattern and conventional serum biomarkers. Overall survival (OS) analyses were conducted with respect to optimized CTC cutoff levels.

Methods: Circulating tumor cell counts were assessed in healthy controls (n = 15) as well as in locally advanced high risk (LAPC, n = 20), metastatic castration resistant (mCRPC, n = 40) and taxane-refractory (mTRPC, n = 15) PC patients. CTCs were detected using the CellSearch System.

Results: In metastatic PC (mPC), CTC counts were significantly increased compared to LAPC (p < 0.001). In LAPC, CTCs were at control level (p = 0.66). Patients with both bone and visceral lesions revealed the highest median CTC count (p = 0.004), whereas patients with sole soft tissue metastases displayed CTC counts comparable to controls (p = 0.16). No correlation was observed between CTC counts and osseous tumor burden assessed by bone lesion count (p = 0.54) or bone scan index (p = 0.81). CTC counts revealed a positive correlation with alkaline phosphatase (p < 0.001) and lactate dehydrogenase (p < 0.001) as well as a negative association with hemoglobin (p = 0.004) and PSA-doubling time (p = 0.01). Kaplan-Meier analyses demonstrated a cohort adjusted cutoff level of 3 CTCs with a shorter OS in case of ≥3 CTCs compared to <3 CTCs (p = 0.001), a cutoff level applicable in mCRPC (p = 0.003) but not in mTRPC patients (p = 0.054).

Conclusions: Circulating tumor cell counts are applicable as a prognostic molecular marker, especially in mCRPC patients harboring bone metastases with or without visceral metastases. For clinical practice, mPC patients with elevated CTC counts in combination with short PSA-DT, high alkaline phosphatase and lactate dehydrogenase levels as well as low hemoglobin levels are at high risk of disease progression and limited OS.

Fig. 1

Boxplot diagram showing CTC findings in venous blood a in controls (n = 15), locally advanced (LAPC; n = 20), metastatic castration refractory (mCRPC; n = 40) and metastatic taxane-refractory (mTRPC; n = 15) patients (Kruskal–Wallis test on linear trend: p < 0.001) b in association with metastatic pattern comparing non-metastatic PC patients (n = 20) with mPC patients harboring only a soft tissue involvement (LN ± visceral metastases; n = 7), only bone metastases (n = 23) or visceral and bone metastases (n = 25), (Kruskal–Wallis test on linear trend: p = 0.004); ο outliers, Asterisks extreme outliers, extreme outliers for mCRPC, bone, bone and visceral metastases not shown (n = 3)

Fig. 2

Kaplan–Meier plot for survival in metastatic PC patients in dependency of CTC levels a comparing patients with <5 CTCs (n = 23) versus ≥5 CTCs (n = 32); b comparing patients with <3 CTCs (n = 20) versus ≥3 CTCs (n = 35)

Fig. 3

Receiver operator characteristic curve demonstrating the significance and specificity for the cutoff values of ≥3 CTCs and ≥5 CTCs predictive for survival in metastatic prostate cancer

Fig. 4

Kaplan–Meier plot for survival in dependency of CTC levels comparing patients with <3 CTCs versus ≥3 CTCs a in mCRPC patients (n = 14 with <3 CTCs; n = 26 with ≥3 CTCs); b in mTRPC patients (n = 6 with <3 CTCs; n = 9 with ≥3 CTCs)