This study really addresses the feasibility of a randomised ctDNA-directed interventional trial in estrogen receptor positive/HER2 negative patients during their follow-up. So ctDNA tests, and the particular one that we used here is the Signatera assay, it’s been reimbursed by CMS and also by several insurance companies for higher risk breast cancer patients, that is stage 2b and 3, as a monitoring tool to look for early signs of molecular relapse or recurrence before it becomes clinically detectable.

So in this study we wanted to ask basically three questions – what is the rate of ctDNA positivity in this patient population over time and what is the rate of ctDNA positive/imaging negative results that would qualify as molecular relapse? The most important question that you’re not going to talk about at ASCO but you’ll talk about in the future when the study is completed, whether an intervention could reduce the rate of distant recurrence.

What was the study design?

So this is a randomised phase II study. The eligibility criteria is to be ER positive/HER2 negative and completed curative intent local therapy, and patients receiving endocrine treatment for longer than six months but less than seven years. We selected this timeframe because there is a randomisation included in the trial if somebody is ctDNA positive/imaging negative, and we wanted to stay within the maximum of ten years of endocrine therapy. We also excluded patients who are ctDNA positive very early on because they may be endocrine therapy resistant.

So high-risk was defined as the usual set of high-risk criteria – four or more positive lymph nodes or a primary tumour greater than 5cm with 1-3 positive nodes and other high risk features that have to be present also at the same time like a grade 3 histology or high molecular risk. These were the entry criteria for screening. So the study includes a screening phase with ctDNA monitoring, surveillance, every six months and if they turn ctDNA positive they undergo a systemic staging with a CT scan of the chest, abdomen, pelvis to determine whether they have asymptomatic but imaging detectable metastatic disease. If they are imaging negative, ctDNA positive then they get randomised to continue with their endocrine therapy versus starting fulvestrant and palbociclib which should be a first-line ER positive  metastatic regimen.

What were the results of this study?

The most important results [??] we found that approximately 12% of patients were ctDNA positive some time during this surveillance period. The median follow-up time from entry to the trial was about 27 months. So we also saw that patients who started out with a ctDNA negative result during serial testing do turn positive at a very steady clip. So there is no real break in the ctDNA positivity findings over time. So patients turn positive at six months, 12 months, 18 months, 24 months up to 5-6 years after their initial entry into the study which is consistent with the clinical behaviour of ER positive disease. We know that many of the recurrences happen in the first five years but even after five years there are many later recurrences. Of course, the premise of ctDNA testing is that you identify this before the patients start to complain about their disease and the symptoms from the metastatic disease.

The second and really important finding was that 71% of patients who were ctDNA positive remained imaging negative at the time of the ctDNA detection. The vast majority of the published literature in the ctDNA field is from archived tissues or archived blood, really. So there is really a paucity of data about what is the asymptomatic imaging positivity rate. We know that in triple-negative disease the asymptomatic imaging findings are much more frequent – it’s closer to 50-60%. But in ER negative disease, in this study, which is the largest study in that space actually, the imaging positivity rate is much lower, it’s about 30%. So this basically establishes the fact that there is such a thing as molecular relapse in ER positive disease.

The third important finding was that really patients were willing to be randomised. This came up that could you actually do a randomised trial in that setting when you know that turning ctDNA positive but being asymptomatic and no evidence of disease but ultimately disease will recur in most of these patients. Would they be willing to change treatment and get randomised to basically continue what they are on versus starting something new? 93% of patients actually were willing to be randomised so this is really an impressive result and we are grateful for the patients who were willing to do that. Because this is the only way to establish clinical utility. Currently, this is the only trial in the United States that actually tests clinical utility of this amazing new tool.

Finally, so the fourth important finding, we looked at ctDNA dynamics. So we didn’t unblind the arms but did report out that in one arm there is a much higher ctDNA clearance rate, about 60% versus 10% in the other arm. Those patients who actually cleared their ctDNA have a low risk, significantly lower risk, of recurrence during our follow-up period than those who do not clear their ctDNA. So the recurrence rate in patients who turn ctDNA positive and remain ctDNA positive is about 90%. Those who actually clear their ctDNA, their recurrence rate is around 40%, not recurrence, the event rate, yes.

The other corollary of this, that those who remain ctDNA negative throughout the surveillance period, as you would expect and what we have known about this assay, these patients who remain negative throughout actually have a very low risk of recurrence. So their recurrence free survival is around 98%.

What is the significance of these results, and what is next for this study?

This is important in a couple of ways. First, it establishes that molecular relapses do exist in ER positive disease and it really raises this really important and tantalising possibility that early intervention, that this study is testing, could reduce the risk of recurrence in patients who have an impending symptomatic and imaging detectable recurrence. It’s very important because the majority of breast cancer death is actually from stage 1 disease and low-risk stage 2 disease. Even though these patients individually have a low risk of recurrence, maybe 5%, 6%, 7%, 8%, but because of the very large number of this population, because of the screen detected cancers, the large number of patients implies a large number of actual deaths, even if the personal risk is only 6% or 7%.

So more than 60% of breast cancer death each year is from early stage disease. One of the rational ways to improve this is monitoring and early intervention, rather than applying the very aggressive, very expensive, often very toxic adjuvant therapies that we use in higher risk patients like stage 2 high-risk or stage 3 disease.

We are very excited about the relationship between ctDNA clearance and improved risk of recurrence and that one arm actually shows a higher ctDNA clearance rate. So that bodes well for the future of the trial. So we would like to complete accrual, the goal is actually to accrue 100 patients, randomise 100 patients. We so far randomised about 40; the results that I will present at ASCO are based on the first 38 patients. But this trial is open for any patient who tests ctDNA positive and imaging negative, even in the context of commercial testing or routine testing. So patients could join the study in the absence of the initial screening phase, so they can just join at the randomisation phase if they are coming in with a ctDNA positive, imaging negative result from routine care.

The trial is open at 15 sites and we would welcome referrals because it’s an important trial to determine the clinical utility of ctDNA monitoring and we still need to accrue about 50 more patients.

Is there anything else you would like to add?

Again, I just want to emphasise that this is the only trial in the United States that tests this really fundamental question that is there clinical utility in ctDNA monitoring and early intervention at the stage of molecular relapse. So we established that molecular relapse exists and patients are willing to be randomised, and the initial results suggest that ctDNA dynamics, clearance of ctDNA, really bodes well for avoiding future recurrence.