NEW YORK – A team of researchers at the University of Southern California has used a commercial blood test to count circulating tumor cells (CTCs) and predict how prostate cancer patients will respond to certain therapies and their survival outcomes.
The researchers described their findings in a study published Monday in JAMA Network Open. Using Menarini Silicon Biosystems' CellSearch test, the team counted CTCs in 503 blood samples from men who were started on systemic hormonal therapy for metastatic hormone-sensitive prostate cancer and found that elevated CTC counts were associated with worse overall survival, progression-free survival, and treatment response. The researchers used samples from a National Cancer Institute-sponsored Phase III clinical trial that investigated the use of androgen deprivation therapy combined with either Takeda's CYP17A1 inhibitor orteronel or AstraZeneca's Casodex (bicalutamide).
Menarini's US Food and Drug Administration-approved test, acquired from Janssen Diagnostics in 2016, uses antibodies against a protein commonly expressed on the surface of CTCs, the epithelial cell adhesion molecule (EPCAM), said Amir Goldkorn, a professor of medicine at USC and the JAMA paper's lead author. The antibodies are bound to magnetic beads that flow through the instrument and pull the CTCs out of the blood and enrich the cells.
The instrument images and stains the enriched cells with three stains that determine whether the cell is nucleated, whether the cell has cytokeratin, and whether it has the CD45 cell surface marker. Cytokeratin is a common protein found in epithelial cells, while CD45 is found in white blood cells, so a CTC will be nucleated, CD45-negative, and CK-positive, Goldkorn said.
The test's operators then look at those images to determine whether a cell is a CTC and get a final count of the number of CTCs in the blood. The counts were categorized as zero, one to four, or more than five CTCs, thresholds that were determined in prior studies, Goldkorn said. Zero CTCs indicates a patient is likely to do well on treatment and has a good prognosis, one to four CTCs indicate a medium likelihood of responding well to treatment, and five or more CTCs indicate a worse prognosis and lower likelihood of responding to treatment.
He noted that his team and other groups have investigated other cutoff points that also work "to some degree" because CTC counts are a "continuous variable," meaning the higher the count, the worse the prognosis.
While there are "a whole lot of different mousetraps to try to enrich and count CTCs," he noted, the researchers chose CellSearch because it is "by far the most well validated instrument … for enriching and counting circulating tumor cells in multiple cancers" and the researchers had previous experience using it. CellSearch has been approved by the FDA for use in metastatic prostate, breast, and colorectal cancers.
Multiple other companies have developed CTC capture and detection methods, including Epic Sciences, which offers its AR-V7 assay. Meantime, CellMax Life, a California-based firm, uses a microfluidic chip to isolate CTCs, and Angle has developed multiple CTC assays using its Parsortix cell sorting platform.
The research team wanted to see if CTCs could help them determine how aggressive a patient's cancer was, which could potentially help guide treatment decisions, Goldkorn said. In their study, the researchers found that the baseline CTC count was "very prognostic of whether or not the men were going to respond [to the treatments in the study] …, how long they would respond, and how long they would live," even if those patients went on to receive multiple other therapies, he added.
Median overall survival for patients with five or more CTCs was 27.9 months, while survival for patients with one to four CTCs was 56.2 months. Median overall survival was not reached for patients with zero CTCs after 78 months. In addition, patients with more than five CTCs had a significantly higher hazard of death and disease progression and a lower likelihood of prostate-specific antigen complete response compared to patients with zero CTCs, the researchers wrote in the study. The survival figures were for all patients in the study across both treatment arms, Goldkorn noted.
The assay could be used to "give us a sense of how someone would do right when they come in initially" for treatment, he said. Goldkorn cautioned that the CTC count wasn't a companion diagnostic and wouldn't be used to determine exactly which drugs a patient should take, but rather would provide a "general baseline sense" of whether a patient would respond well to treatment and survive long term.
It's "helpful because we're constantly developing new drugs, new combinations … and these can be more toxic, more aggressive," Goldkorn said. "Not every man needs that."
Patients with high CTC counts could be directed toward new clinical trials with more aggressive treatments, and the test could be used to determine eligibility for those trials, he added.
However, in the future it could potentially be used as a companion diagnostic if trials were designed to test multiple drugs to determine whether high CTC counts indicated a response to one therapy or another, Goldkorn said.
In addition, Goldkorn's team is working on ways to combine CTC counts with other liquid biopsy readouts — such as DNA sequencing of CTCs or mutational profiling for cell-free DNA — along with other readouts like CT scans to create "highly multiplexed signatures" that could be more predictive of survival or treatment response. "It's kind of like looking at [the cancer] from six different angles at the same time and creating a biomarker from that."