Circulating tumor cells in metastatic lung cancer enriched by EpCAM expression and physical characteristics

Introduction: Circulating tumor cells (CTC) measured with the CellSearch system in patients with metastatic carcinomas are associated with poor survival. The frequency of CTC detected by the CellSearch system in non-small cell lung cancer (NSCLC) patients is relatively low, raising the question whether some CTC are not detected by the CellSearch system. To investigate this, additional antibodies were added to broaden the coverage of cytokeratins and leukocytes. Additionally, a device was designed that collects the sample material of the individual samples that are discarded by CellSearch. This collected waste is filtered for CTC isolation based on physical characteristics and the CTC are stained with a cocktail of antibodies. Methods: A device was designed that uses optical sensing to detect the presence of blood in the waste tube of the CellTracks Autoprep. It collects the waste of individual samples in a 50 mL conical tube. After collection the blood is passed with 100 mbar pressure through a 8x8 mm2 microfabricated silicon microsieve containing 300,000 pores of 5 µm in diameter (VyCAP, Deventer, The Netherlands). The performance was tested using four pre-stained cell lines: Colo320 (size 11 µm, ∼1,616 EpCAM antigens), SW480 (size 11 µm, ∼63,233 EpCAM antigens), T24 (size 16 µm, ∼2,167 EpCAM antigens) and SKBR3 (size 16 µm, ∼445,000 EpCAM antigens). Cells are spiked in 7.5 mL of blood collected in CellSave tubes from healthy volunteers. Spiked blood samples from healthy donors and patients with NCSLC and small cell lung cancer (SCLC) (enrollment is ongoing) were processed on the CellSearch and filtration system between 24 and 96 hours of collection. The cells on the microsieves were stained with a nucleic acid dye, antibodies recognizing leukocytes and all cytokeratins. Additional antibodies were added to the CellSearch test to cover all cytokeratins and broaden the coverage of leukocytes. Results: The recovery percentage of the CellSearch system for the different cell lines used was: 2% of COLO320, 91% of SW480, 2% of T24 and 87% of SKBR3. Additional recovery on the microsieves after filtration of the CellSearch waste was: 18% of COLO320, 6% of SW480, 59% of T24 and 2% of SKBR3. The combined recovery accounts for 20% of COLO320, 97% of SW480, 61% of T24 cells and 89% of SKBR3. In patients with NSCLC and SCLC either no CTC were detected at all, or in various proportions in the CellSearch cartridge, on the microsieves after filtration of the CellSearch waste or with the additional antibodies that were added. Conclusions: We combined the CellSearch system with a device for collecting and filtering the CellSearch waste. On cell lines this demonstrated that a low EpCAM expression results in the presence of CTC in the waste that would not be detected by the CellSearch system. In both NSCLC and SCLC additional CTC can be detected but it still remains to be determined whether the CTC not detected by the original CellSearch approach are also of clinical relevance