CTCs Serve as Prognostic Marker for Outcomes in Metastatic Hormone-Sensitive Prostate Cancer

Circulating tumor cell (CTC) count is a beneficial biomarker in patients with metastatic castration-sensitive prostate cancer (mCSPC), according to Amir Goldkorn, MD, who added that the use of CTCs could be of critical importance in future research.

For example, in the randomized, phase 3 SWOG S1216 trial, 1313 patients with newly diagnosed mCSPC were randomized to receive androgen deprivation therapy (ADT) with orteronel (TAK-700) or bicalutamide (Casodex). The primary end point was overall survival (OS); secondary end points were 7-month prostate-specific antigen (PSA), 2-year progression-free survival (PFS) rate, safety, and toxicity correlatives.

Liquid biopsy studies were integrated into SWOG S1216 as correlative end points of the trial. 

In the multi-parametric analysis of the study, investigators sought to determine if CTC count could be a valuable biomarker in mCSPC. CTCs were counted by the FDA-cleared CellSearch platform at baseline and disease progression and were analyzed for associations with 7-month PSA response and 2-year PFS.

From 2014 to 2017, 523 baseline samples were collected. Results showed that, in the 7-month PSA analysis (n = 264), CTCs were detected in 38% of men, with a median of 4 CTCs in those with detectable CTCs. Conversely, in the PFS analysis (n = 336), CTCs were detected in 37% of men, with a median of 3 CTCs in those with detectable CTCs.

Results showed that baseline CTC count was prognostic of clinical outcomes; the odds ratio in the 7-month PSA analysis was 6.06. In the 2-year PFS analysis, the odds ratio was 3.72.

“These data are very promising. It tells us that men with 0 CTCs are more likely to have a favorable disease course and should respond well to hormonal therapy. On the other hand, men with more CTCs may not have a good response and may contemplate a more aggressive therapy,” said Goldkorn. “Ultimately, this trial helps us understand whether CTCs are a beneficial prognostic and predictive factor in this patient population.”

In an interview with OncLive, Goldkorn, an assistant professor of medicine at the Keck School of Medicine, USC Norris Comprehensive Cancer Center, discussed this CTC analysis of the phase 3 SWOG S1216 trial in patients with mCSPC.

OncLive: What was the rationale behind the SWOG S1216 study in this patient population?

Goldkorn: This study looked at CTCs as a biomarker in mCSPC. This is very common in men whose prostate cancer progresses beyond the localized disease and is no longer curable with surgery. In this disease, we usually start by treating patients with hormonal therapies, such as ADT, in combination with another drug, such as chemotherapy or androgen receptor inhibitors. However, we have very few biomarkers that could tell us who will respond and for how long. There is a great need for new biomarkers in this space.

How can you utilize CTCs in clinical practice for these patients with prostate cancer?

CTCs have been evaluated extensively in prostate cancer. The vast majority of studies have been done in a very advanced disease stage, also referred to as metastatic castration-resistant prostate cancer—after men have progressed on initial hormonal therapy. In that state, the most clinically validated and most extensive studies that were conducted used the CellSearch system, an FDA-cleared instrument, that counts CTCs. Those studies have shown that looking at CTCs at the start of therapy in men with very advanced disease can be prognostic of how they will do [on treatment]. There are a variety of studies along those lines.

We’re now focused on taking the CellSearch system and CTC count into this earlier disease state, when men first begin their treatment with hormonal therapies for metastatic prostate cancer.

Could you highlight the design and goal of the SWOG S1216 study?

This trial is being conducted by the National Cancer Institute Southwest Oncology Group, led by Neeraj Agarwal, MD, of Huntsman Cancer Institute. This is a biomarker study where we collected CTCs at the start of the trial. The trial was comprised of 1200 men who were initiated with ADT plus bicalutamide or ADT plus orteronel, which is similar to abiraterone acetate (Zytiga).

The primary end point is OS, which we should be seeing sometime in the next year. We also wanted to determine whether CTCs could predict response to hormonal therapy and [disease] progression, specifically defined as a drop in PSA after 7 months in addition to the 2-year PFS.

What did you find to be most intriguing from the CTC analysis?

When men started therapy, we found that baseline CTC count were, in fact, prognostic with PSA response and progression. For example, if one man had 0 CTCs at the start of therapy [and another] man had 5 or more detectable CTCs in his blood, the man with 0 CTCs were 6.1-fold more likely to attain PSA response at 7 months. Conversely, the men who had 5 or more CTCs at baseline had a 4-fold odds ratio of progressing at 2 years. 

Therefore, no CTCs led to very good responses [to therapy] and long PFS; those with 5 or more CTCs had a lower PSA response and they were more likely to progress at 2 years.

This was even the case when we looked at a variety of cut-points, such as 0 CTCs versus any CTCs, and less than 5 CTCs versus greater than 5 CTCs. In all of those cases, it was statistically significant and [more CTCs] correlated with both PSA response and disease progression.

Additional studies will be looking at multiple time points of CTCs. We're also analyzing factors such as CTC DNA from mutations and CTC RNA for gene expression. It will be very interesting to see the correlation of all of these biomarkers in order to create this multiparametric liquid biopsy profile.

Reference

Goldkorn A, Tangen C, Plets M, et al. Baseline circulating tumor cell count as a prognostic marker of PSA response and progression in metastatic castrate sensitive prostate cancer: results from SWOG S1216, a phase III randomized trial of androgen deprivation plus orteronel or bicalutamide. J Clin Oncol. 2020;38(suppl 15; abstr 5506). doi:10.1200/JCO.2020.38.15_suppl.5506