We did a large study of circulating tumour cells in metastatic castrate sensitive prostate cancer. In this disease setting, basically a disease where men with metastatic disease are first being treated with hormonal therapy, the standard of care right now is a combination of hormones with chemo or with other androgen receptor targeting agents like abiraterone or enzalutamide but we have very little in the way of biomarkers to guide our knowledge of who will respond to these treatments and for how long they would respond.

So basically the purpose of the study was to develop such a biomarker based on circulating tumour cells. CTC analysis has been done using the FDA cleared CELLSEARCH platform in a more advanced disease state, in castrate resistant prostate cancer after men have progressed on hormonal therapies, but very little has been done in this first line metastatic castrate sensitive prostate cancer. So our hypothesis, what we asked in the study, is whether the CTCs on CELLSEARCH could be used as a biomarker to tell us who will respond and for how long.

What were your methods?

study was conducted in the context of a large phase III prospective randomised clinical trial, SWOG S1216. It’s a study run by the NCI Southwest Oncology Cooperative Group. The PI on the clinical trial component is Dr Neeraj Agarwal at the University of Utah. What we did in this study is we looked at the CTCs at baseline for these men, 1,200 men, who were randomised to start treatment with androgen deprivation therapy, hormonal therapy, with either bicalutamide or orteronel, which is a a CYP17 inhibitor similar in class to abiraterone.

The readouts or the endpoints that we looked at in the data that has now been presented is PSA response at seven months, meaning after seven months of therapy how much did the PSA drop. We wanted to look at whether it dropped below 0.2 or between 0.2-0.4 or above 0.4 which are categories of PSA response that have been previously shown to be an intermediate biomarker for overall survival. So we looked at how well the PSA dropped. The other thing we looked at was the two year progression free survival, in other words who had progressed on their hormonal therapy and who had not at two years.

What did you find?

The findings were actually quite dramatic. We found that there was a significant difference in response and in progression between men who had CTCs and men who did not have detectible CTCs at baseline when they started on study. As one example that we discussed in our results, taking two men, if a man had five or more CTCs at the start of the study versus had zero, no CTCs at the start of the study, the man who had no CTCs had a more than sixfold odds ratio of having this complete PSA response to hormonal therapy by seven months relative to the man who had five or more CTCs. Conversely, that man who had five or more CTCs at the start of the study had a nearly fourfold odds ratio of progressing by two years on the study with hormonal therapy relative to the man who had no CTCs at baseline.
We actually found these types of relationships to hold across the board. We looked at a variety of cut-off points, for example men who had less than five versus greater than five CTCs, which is the traditional CELLSEARCH cut-off that has been used in more advanced disease. We looked at men who had any CTCs detectible versus zero CTCs and in all cases there was a very significant difference in their response to treatment and in their progression whether they had CTCs or not.

What were your conclusions?

In our study CTC counts at baseline were indeed very much prognostic of PSA response and progression at two years in men first going on therapy for metastatic castrate sensitive prostate cancer. This, to our knowledge, is the largest prospective trial looking at CTCs in this setting so we’re very enthusiastic about these results. We think they could definitely lead us to some important next steps. For one thing, even where we are right now one might consider clinical scenarios where, for example, a man comes in to the clinic and perhaps is an older or more frail gentleman but has no CTCs. So in a gentleman like that we might feel comfortable treating him with more traditional or more conservative hormonal therapies knowing that he should have a pretty good favourable response to these treatments. Versus another gentleman who might walk in and have many CTCs where the indication would be that his disease maybe a little bit more aggressive, may not respond as well to just hormonal therapies and we might lean more towards more aggressive combination therapies or even clinical trials as they become available.

So basically helping us to identify which men are likely to have a good response and a protracted response versus men who are less likely to have that kind of favourable response. That’s kind of where the data are now. In addition, as the final trial results mature in the coming year we’ll also have final overall survival data and we’ll also unblind the trial completely. So we’ll be able to do an analysis of whether CTCs are actually associated not only with overall how the patient does but also how the patients do on each of the two treatment arms. So the idea there is that potentially CTCs might be able to help us choose between specific different treatments so that in future trials we can build on that and say, ‘Let’s see if having many versus few CTCs can help guide us to choose between hormonal therapy and chemo or hormonal therapy and abiraterone or enzalutamide,’ because these are the choices we need to make in the clinic today and this is where we need the biomarkers to help guide our decisions.