CellSearch circulating tumour cell count

What it measures. CellSearch pulls circulating tumour cells out of 7.5 millilitres of blood with magnetic beads coated in an antibody to EpCAM, stains them for cytokeratin and a nuclear dye, excludes white cells with CD45, and counts the cells that pass. In metastatic breast cancer five or more cells, and in metastatic prostate and colorectal cancer five or three, mark an unfavourable group.

Evidence. The 2004 New England Journal of Medicine study in metastatic breast cancer showed that women with five or more cells before starting a new treatment had far shorter progression-free and overall survival, and the FDA cleared the system the same year, then for prostate and colorectal cancer in 2008. The count also tracks response: a fall below the threshold after one cycle is a good sign. The disappointment was SWOG S0500, which switched chemotherapy early in women whose count stayed high and found no survival gain, so the count is prognostic rather than a guide to which drug to use. In hormone receptor positive breast cancer the French STIC CTC trial used the count to choose between endocrine therapy and chemotherapy and found the strategy non-inferior to the physician's choice.

Who should have it and what changes. Guidelines do not recommend routine CellSearch counts because no decision has been shown to improve outcomes from them, so use is mainly in trials and in centres studying circulating cells, where the captured cells can also be stained for HER2, androgen receptor variants or PD-L1. The test is available through Menarini Silicon Biosystems and reference laboratories at a cost similar to a specialised blood test. Circulating tumour DNA has taken over most of the clinical questions the count was meant to answer.