Background: Detection of t-SCNC in mCRPC patients relies primarily on histopathologic evaluation (Histo) of a metastatic tumor biopsy (bx), likely leading to underdiagnosis. The clinical features of t-SCNC and the diagnostic utility of serum NE markers (neuron-specific enolase (NSE) and chromogranin (CGA)) were evaluated. Methods: Eligible patients (pts) underwent a metastatic bx at one of 5 centers. Histo was performed by 3 independent pathologists (JH, GT, LT). NE markers were evaluated in a central lab (lower limit = 1 ng/mL). Kruskal-Wallis and chi-square test were used to compare continuous and categorical variables, respectively. Receiver-operative-curve (ROC) analysis of serum NE markers was undertaken. Results: 160 consecutive pts with available Histo and NE markers were included. t-SCNC was found in 27 pts (17%). Detection of t-SCNC was observed in all bx sites, including liver (14%), lymph node (19%) and bone (14%). Clinical features are shown in the Table. By ROC analysis, if both serum NSE was > 6.05 ng/mL and chromogranin was > 3.1 ng/mL, the sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) for the detection of t-SCNC were 95%, 50%, 98%, and 22%, respectively. Conclusions: Many of the classic features of de novo SCNC, including low PSA levels, do not reliably distinguish t-SCNC. In contrast, serum NE markers have diagnostic utility with high sensitivity and NPV, but low specificity and PPV. Heterogeneous NE differentiation may partially account for these findings. Clinical trial information: NCT02432001