Good morning. What I’m going to describe to you involves patients who enrolled in a prospective clinical trial which was co-ordinated by the ECOG-ACRIN Cancer Research Group. This is a federally funded group supported by the NCI, and we were fortunate enough to receive funding from the Breast Cancer Research Foundation and the Susan G. Komen Foundation to obtain biospecimens from patients participating in the trial which made this analysis possible.
Late recurrence in oestrogen receptor positive breast cancer is a huge clinical problem, and some experts in the field have dubbed this to be the next frontier for research. It accounts for about half of all recurrences in ER positive breast cancer and it’s generally defined as recurrence occurring five or more years after diagnosis. That’s a historical precedent that’s based on the fact that generally, in the past, we’ve recommended at least a five-year course of endocrine therapy. In a recent meta-analysis that was reported in The New England Journal of Medicine, the ten-year recurrence risk after a five-year course of endocrine therapy was 5%, 10% and 22% for patients who had zero, one to three, and four to nine positive lymph nodes.

With regard to the adjuvant therapy we use to prevent recurrence and biomarkers for late recurrence we know that adjuvant chemotherapy predominantly decreases early recurrence within the first five years of diagnosis. More recently, we’ve come to recognise that extended adjuvant endocrine therapy beyond a course of five years, either with tamoxifen or with aromatase inhibitors, can provide some incremental benefit which ranges from 2-4%. There are some tests that can provide information about late recurrence risk – these are generally gene expression assays obtained on the primary tumour at the time of diagnosis and in general in those assays that have shown the clinical validity for this specific association there is about a 2.5-fold higher risk of recurrence for those who are classified as high versus low risk by these assays. We hypothesized that the detection of circulating tumour cells in patients who are five or more years after diagnosis might be a clinically useful way to apply this technology for a couple of reasons. First of all with analytic validity – that is, the technical issues related to the reliability of the test – this is an FDA CLIA blood test for enumerating CTCs in metastatic breast cancer. With regard to clinical validity, we know that the presence of CTCs and their CTC burden is associated with prognosis in metastatic breast cancer. We also have some information indicating that the presence of CTCs and CTC burden is associated with recurrence in early breast cancer from three studies shown here, where the CTC detection rate at the time of diagnosis, after surgery, and generally before systemic adjuvant chemotherapy was in the range of 20-25%, and after a median follow-up of anywhere from three to five years those patients who were CTC-positive had anywhere from a two- to a five-fold higher risk of recurrence. Although these studies show clinical validity, that is an association between a positive CTC test and recurrence, there was little clinical utility from this information, that is, it wasn’t necessarily useful in implementing a change in treatment because all of these patients were going to receive systemic therapy.

We applied this assay to patients who had previously enrolled on a prior trial, they had stage 2-3 breast cancer, they had received standard modern chemotherapy including an anthracycline and a taxane, and that they had oestrogen receptor positive disease and they received at least a five-year course of endocrine therapy. We went back to these patients, who we were following anyway as part of this clinical trial, and we asked them: ‘Would you be willing to provide a blood sample for us to do this test and then we’ll continue to follow you as we would normally do, and we would not report the test result back to you?’ We also were collecting other blood specimens from them at the time that they enrolled on the trial and then annually for five consecutive years. Over a course of about 3½ years we enrolled 547 patients, this accounted for about 9% of all of the patients who enrolled on the original trial, and the characteristics are shown here. 56% were fifty or older; 59% had a tumour of at least two centimetres; 73% had node-positive disease; about two-thirds had hormone receptor positive disease; a little more than half had high-grade tumours and, of the 330 patients who had hormone receptor positive disease, where we had information, 88% of them were taking endocrine therapy at the time of the CTC sampling.
About one year after completion of accrual we analysed the data, and at the time of the analysis, the median follow-up was 1.8 years, with a range of out to 3.9 years, and we found that about 4% of patients who had hormone receptor positive disease had a recurrence. This was among the 353 patients who had hormone receptor positive disease. In addition, we found only one recurrence among the 193 patients who had hormone receptor negative disease. This was actually not even a distant recurrence, this was a local regional recurrence, whereas all of the recurrences in the hormone receptor positive group were distant recurrences. This actually confirmed what we had known in that the hormone receptor positive disease is associated with a higher risk of later recurrence.

We also looked at the proportion of patients who were positive for CTCs. Overall this was about 4.8%, and there didn’t seem to be much of a difference in the hormone-receptor-positive group - 5.1% versus the hormone receptor negative group 4.1%. But because we had essentially no recurrences in the hormone receptor negative group, the subsequent analysis I will show you will be restricted to the 353 patients with hormone receptor positive disease. Here are the main results: this is a Kaplan-Meier curve showing the time to recurrence for those who were CTC negative on the top, and those who were CTC positive on the bottom. Those patients who were CTC positive had nearly a 22-fold higher likelihood of having a recurrence which was highly statistically significant. When we did a multivariate analysis that adjusted for clinical covariates such as age, tumour size, nodal status, and grade, this very strong association persisted with a hazard ratio of 18.1. The median time to recurrence in those who were CTC positive was 1.6 years. This is a fairly long lead time and for some it was as high as at least 2.8 years, with a low range of six months. If one looks at the likelihood of having a recurrence per person-year, it was approximately 25% for patients who were CTC positive, and only 1.5% for those who were CTC negative. The positive predictive value of having a positive test for recurrence by two years was 35% for those who had a positive CTC assay, and perhaps more importantly, the negative predictive value was 98% for recurrence at two years, in other words, if someone had a negative CTC test, there was less than a 2% chance that they would have a recurrence within the next two years.

So, what are the implications here for clinical practice, and for clinical research? First of all this provides proof of concept of the clinical validity. This provides level one evidence, the highest level of evidence, supporting CTCs as a biomarker prognostic for late recurrence in the specific subgroup of patients with hormone receptor positive, HER2 negative early breast cancer which accounts for about two-thirds of all breast cancers. Secondly, it demonstrates a very high level of risk stratification, a twenty-fold higher risk, which far surpasses other biomarkers that are currently available. Very importantly, it supports the concept of perhaps a new paradigm and that is to have a second decision point to tell if therapy, for individual patients, based on a biomarker that’s obtained not at the time of diagnosis, and not making all decisions at the time of diagnosis, but making a second decision point five years after diagnosis that’s based not only on the patient’s tolerance to prior therapy and their risk of subsequent recurrence based on clinical pathologic features, but also integrating a biomarker, perhaps this biomarker, to assist in making a decision.

However, we still have work to do. We need further study to really nail down what the clinical utility of this information is – that is if we have a negative CTC assay, can we really effectively spare continued extended adjuvant endocrine therapy beyond five to seven years? Secondly, what do we do for those patients who have a positive assay? We know that they have a high recurrence rate, what can we do to prevent it? In order to address this issue, we will need to design clinical trials where we will randomise patients to new therapeutic strategies that are approved, perhaps, in the metastatic setting, such as CDK4/6 inhibitors. Also there are new oral selective oestrogen receptor downregulators, drugs such as fulvestrant that are currently available that we can only give by intramuscular injection, that are effective in patients who’ve had progressive disease on a prior aromatase inhibitor. Finally, we need to compare this with other new and emerging technologies like circulating tumour DNA, and perhaps use these in combination with these assays.