Endocrine Abstracts

Background: The identification and characterization of CTCs as part of a minimally invasive “liquid biopsy” has potential as a real-time biomarker in cancer patients. Using blood samples from NET patients, we evaluate the epitope-independent Parsortix, which enriches CTCs based on size and rigidity, and compare to the EpCAM-dependent CellSearch platform.

Methods: About 10 patients with histologically confirmed metastatic NET had synchronous 7.5 ml blood samples collected for processing in the CellSearch and Parsortix. CellSave samples were processed in the CellSearch for identification and enumeration of CTCs as previously described. EDTA samples were run in the Parsortix and an automated protocol for fixing and staining within the separation cassette was applied. Cells were stained for cytokeratin, CD45, EpCAM and DAPI and immunofluorescence imaging used to identify CTCs based on morphological and staining characteristics outlined in pre-defined criteria. Image acquisition and identification of CTCs were performed separately by 2 independent operators. The images acquired by the observer with the higher number of CTCs were then reviewed independently by the second observer in order to establish a final CTC count.

Results: CTCs could be detected in all 10 NET patients using Parsortix (range 6–93). Parallel samples processed by CellSearch revealed only 70% (7/10) positive for CTCs (range 0–85). In the 3 patients where CellSearch was unable to identify any CTCs, Parsortix identified counts of 6, 29 and 13. CTCs detected by Parsortix were heterogeneous with regards to EpCAM status (high; 22%, low; 70%, absent; 8%). Regression analysis demonstrated a correlation coefficient (R2) of 0.24 between the 2 operators, indicating only weak correlation. Despite this, significantly more (P=0.032) CTCs were identified using Parsortix compared to CellSearch (mean 31 vs 13 respectively) and there was only 1 patient in whom higher CTC counts were identified using CellSearch.

Conclusion: The epitope-independent Parsortix can be used to increase the yield of CTCs captured from NET patients when directly compared to CellSearch. In view of the inter-observer variability seen, Parsortix does not appear to be the ideal platform for enumeration but may be of value in increasing the yield of CTCs available for downstream molecular and functional analysis.