Accepted: 13 February 2017/Published online: 28 February 2017© Springer-Verlag Berlin Heidelberg 2017 performed with EpCAM (epithelial cell adhesion molecule) based approaches [12]. This might be even more interesting since more insights are gained into biology of non-EpCAM positive tumor cells—cells which might have shifted to a mesenchymal rather than an epithelial phenotype and thus, might escape an EpCAM enrichment strategy [13]. Moreover, focussing on AR-V7 mRNA from CTCs origin only presumably excludes significance of AR-V7 transcripts originated from extracellular vesicles, which might exhibit clinical biomarker validity as well [14]. Another advantage of the whole blood RNA extraction method is the missing necessity of special processing, presumably allowing higher reproducibility among different laboratories. However, results of these studies are highly diverse. Therefore, we would like to sound a note of caution. In the study of Liu et al. the authors separated the blood sample by using CD45-antibody coupled magnetic beads to discriminate between leukocytes which do not express AR-V7 (CD45+) as well as the remaining cell population in which AR-V7 positive tumor cells are supposed to be present (CD45−). However, the authors detected AR-V7 in CD45+ samples and explained this phenomenon by cross contamination due to non-specific binding of the CD45 antibody to tumor cells, aberrant expression of CD45 by tumor cells or leukocyte/CTC clusters. In the study by Todenhöfer et al. the authors used blood samples from pre-treated CRPC patients (n= 64) as well as healthy donors of similar sex and age comparable to prostate cancer patients, ie men≥ 50 years of age which had serum PSA value< 1 ng/ml as well as no clinical signs of prostate cancer [10]. The authors performed RNA extraction followed by cDNA synthesis and gene-specific pre-amplification. Subsequently, they used quantitative real time RT-PCR (qPCR) to detect AR-V7 as well as other prostate cancer marker genes, eg