Thank you for this opportunity to present our work. This is a very big question and we had challenging results as well. So the idea is in the context of metastatic breast cancer that is hormone receptor positive HER2 negative… I’m sorry, I forgot a conflict of interest. This study has been partially funded by Menarini which is the provider of the circulating tumour cell count.

So in the context of first line therapy for hormone receptor positive metastatic breast cancer we know that we can treat these patients with either endocrine therapy or chemotherapy as front line therapy when they relapse. Since the ‘90s we had trials showing that chemotherapy was not superior to endocrine therapy and since that time endocrine therapy is the preferred first line for most patients in the absence of either primary endocrine resistance or serious concern about the patient prognosis. Here the issue is that the patient prognosis is evaluated by clinician and it’s poorly reproducible from one clinician to another, one centre to another and one country to another. In that context the CTC count is highly relevant because we know it is a reproducible biomarker that has been FDA cleared for clinical use so it has reached a level of evidence 1 in terms of prognostic biomarker. We also know that it is superior to all the usual prognostic markers that are taken into account by clinicians to choose between hormone therapy or chemotherapy.

So here the aim of the trial was to compare head to head a CTC-driven choice between hormone therapy and the clinician choice between hormone therapy and chemotherapy. The main inclusion criteria were hormone receptor positive HER2 negative metastatic breast cancer without any prior therapy for metastatic breast cancer; they had a condition compatible with hormone therapy or chemotherapy and they were randomised one by one into the two arms of the study. In the first arm, which is the standard arm, the clinician had to choose and to decide and the CTC count remained blinded. So some patients were to be treated by hormone therapy and I will refer to that subgroup as a clinical low subgroup, and some patients were treated by front line chemotherapy and are the clinical high group. In the other arm the clinician’s choice was just dismissed and we based the patient treatment only on the circulating tumour cell count. Low CTC count, which means a good prognosis, patients were treated with endocrine therapy; high CTC count and a worse prognosis patients were treated with chemotherapy. In both the study arms maintenance hormone therapy was allowed after front line chemotherapy.

We were comparing head to head a very single biomarker, standalone biomarker, versus a multiparametric clinical estimate.  So we therefore aimed at demonstrating a non-inferiority in terms of progression free survival between the CTC driven arm and the clinically driven arm. Secondary endpoints included overall survival, which is very important in first line trials, and pre-specified subgroup analysis. We randomised nearby 800 patients and the results have been reported with a median follow-up of 30 months.

Here is the primary objective of the trial. The primary objective was met. We are demonstrating a non-inferiority for using this single biomarker, your single biomarker, your decision is not inferior to a very heterogeneous clinically driven decision and the median PFS was 15.6 months in the CTC arm, 14 months in the control arm. So we also had very similar overall survival curves.

The main point here is just to say that if you are hesitating about which drug to give, for example in the overall study population you’re not making any mistake if you rely on the CTC count to take a decision.

Then we will dive into planned subgroup analyses which are quite interesting. We have basically four kinds of patients: clinically low CTC low; clinically low CTC high; clinically high CTC low and clinically high CTC high. You see that the characteristics highlight that the clinically low CTC low subgroup has a rather more favourable profile whereas the patients with clinically high, so patients for which chemotherapy was a favourite treatment option for the clinician, and they also had a high CTC count, these patients have very poor characteristics. Most patients when treated with chemo were treated by taxane or capecitabine, so the standards for first line treatment.

Here are the concordant groups. The first subgroup here is about half of the patients, they were treated by hormone therapy in both arms so there is no difference between the two curves, it was fully expected, but we do have a very good progression free survival in these patients of 18 months. Without CDK4/6 inhibitor, single agent endocrine therapy in CTC low patients can give you a very good PFS. On the other side of the prognosis you have these patients that are really at high risk of short PFS and the PFS was not that bad with about 12 months achieved with chemotherapy front line, maybe followed by endocrine therapy. So for this concordant subgroup basically CTC can complement the prognostic estimates and may highlight patients with excellent or poor outcomes. What is interesting, and the most interesting of course, is patients with discrepant - when the clinicians here in that subgroup were thinking to endocrine therapy first but these patients had also at the same time a high CTC count. These patients if randomised in the control arm were given endocrine therapy, if randomised in the CTC arm were given chemotherapy front line. What we see here is superiority for the CTC arm in terms of hazard ratio and it was statistically significant. We also see a trend towards statistical significance in terms of overall survival. So, again, if you have a patient that we think that endocrine therapy is appropriate for the patient but a high CTC count, chemotherapy might achieve a better result.

What is also very interesting and a bit unexpected was the other subgroup, the clinically high CTC low patients, so patients for which chemotherapy was the most appropriate option according to the clinician but they had a low CTC count so they were treated by chemo in the control arm, by hormone therapy in the CTC arm. What we see here is that first there is no statistically significant difference between the two arms yet if you look at the curve you can see a difference and the lack of statistical significance might be due to the limited number of patients included in that subgroup. But apparently the chemotherapy is doing better.

To conclude, what we did, this is an exploratory, so unplanned, analysis. We took together the two discordant subgroups and we analysed them not by their allocated arm but by their allocated therapy. So we compared in patients who have discrepant prognostic estimates whether it is better to treat them with endocrine therapy or front line chemotherapy. We find that it is worth to treat these patients with chemotherapy, not only because we achieve a longer progression free survival in these patients but also you achieve a longer overall survival for these patients and overall survival is extremely rare in a first line trial. This is a really interesting result highlighting that basically the use of front line chemotherapy, which has been dismissed for a long time, can be a very suitable option for patients with unfavourable characteristics such as high CTC count.

So the conclusion was in the overall population CTC count was tested as a standalone biomarker, again it is reproducible and you don’t need a clinician or is my clinician a good doctor, a bad doctor, we just do a test and it can tell you what treatment to receive. In the overall subpopulation this biomarker is clinically reliable. There is room for improvement, of course, by combining CTC count and integrating it into multiparametric decision tools. Invasion with more than five CTCs, so your elevated CTC count, chemotherapy is better than single agent endocrine therapy and now we have CDK4/6 inhibitors in first line so it is worth doing further study with these new drugs. In the discordant subgroup analysis I showed you, this is the first contemporary study that reports a significant reduction in the risk of death. Again, this is an extremely rare result with front line chemotherapy. So this result was obtained in about 300 patients, so it’s not a small subgroup, and these patients were randomised between hormone therapy and chemo. I think this result challenges our current standards.
To conclude, CTC count should be included in the decision algorithm to choose between front line chemo or single agent hormone therapy in hormone receptor positive HER2 negative metastatic breast cancer patients.