Minimal Residual Disease: The 'Holy Grail' in Adjuvant Chemotherapy

Prof Charles Swanton


September 29, 2019

Prof Charles Swanton from the Francis Crick Institute and University College London (UCL) Hospitals has been presenting the latest on TRACERx and MRD at ESMO 2019.

This transcript has been edited for clarity.

MRD stands for minimal residual disease. It is a term borrowed from haematological malignancies, applied to solid tumours, and is really the Holy Grail in the adjuvant setting, that is following surgery when patients have been radically treated with curative intent, where they've had their primary disease resected. We don't know who has been cured by surgery alone, and who hasn't.

So if you take a typical random selection of 50 patients from TRACERx, about 30 of those patients would be eligible based on stage criteria alone for adjuvant chemotherapy. We know from the latest meta-analysis and other studies, and the literature, that adjuvant chemotherapy is associated with an absolute survival benefit of about 5%.

Patients with stage 3 disease actually will benefit even more, probably 12 to 15% absolute survival benefits. But stage 1 disease is probably in the region of about 4 to 5% absolute survival benefits. So that means if you take a typical stage 1 or stage 2 patient, we're treating many, many patients who will derive no benefit from adjuvant chemotherapy.

So if you took, as I said, a random selection of 50 patients from TRACERx, 30 would be eligible for adjuvant chemotherapy, of which 20 of that 30 would be cured by surgery alone. So that leaves a very small proportion, only 10 of those remaining 30 patients, who we deem are not cured by surgery who we would hope to be cured, or have a chance of cure, by adjuvant chemotherapy, but clearly, we don't know which ones of those 10 patients will benefit from adjuvant chemotherapy because we've at the moment got very few stratification tools to separate patients into high-risk and low-risk patients over and above clinical and pathological staging criteria.

So, us and many others in the field, have been developing what are called minimal residual disease assays in one shape or another.

And the premise of these minimal residual disease assays, is that we can take a blood test from patients and spot the evolution of the disease from circulating tumour DNA in blood during the follow-up periods when patients attend to clinic for blood tests as part of routine clinical care, and along with chest X-rays, etc, to see if we can spot evidence of recurring disease.

Now the idea is we take blood draws from patients every 3 months as part of their routine care and analyse that blood for signs of mutations in that blood that are concordant with the original mutations we found in the primary tumour. So how we do this is we create a bespoke multiplex PCR assay, led and generated by the phylogenetic trees of the patient's original tumour. That means that each patient has a bespoke assay, a distinct and unique combination of multiplex PCR primers that will enable us to amplify mutations specific to that patient's own tumour and detect potential evidence for the evolution of recurrence of the disease, based on that patient's original primary tumour genomic makeup.

And over the last 3 years or so, we've collaborated with diagnostics companies in the US to be able to leverage this knowledge and put this idea to the test. And in our early foray into this field we found that this minimal residual disease bespoke assay, where we were targeting 20 mutations, would give us a minimal residual disease lead time of about 70 days.

And lead time is essentially the time between diagnosing or identifying the signals of clonal and sub clonal relapse by minimum residual disease, PCR detection, and the interval between that time and clinical detection of minimum residual disease on a CT scan.

Now, over the last 2 years, we've been optimising this technology for several reasons. The first reason is that there is a sensitivity issue, that we need to maximise our limited detection to get down to lower variant allele frequencies to be able to maximise those clinical lead times.

In our original foray into the field we had limited detection of about 0.1% variant allele frequencies, that is, the mutant could be present in the blood at about 0.1% and we would be able to detect it in a blood tube.

Now, we are looking at trying to advance that limited detection by further 20 fold, down to 0.005%, to maximise that lead time between detection of the mutation and recurrence of disease.

We are busy working hard on optimising those assays in order to improve that lead time. And the principal way we're doing this is by targeting not 20 mutations, but hundreds of mutations. So this has two advantages we hope. The first is that we hope it will improve our limited detection ability. The second is that by tracking hundreds of mutations per patient, we will get enormous resolution, we hope, over the subclonal evolution of disease over space and time through liquid biopsy assays, through a simple 20 ml blood draw from the patient every 3 months.

We hope we'll be able to start to track branched evolution in real time and start to see the hallmarks of neoantigen loss as tumours evolve, as an indicator of immune evasion, and most importantly, the rise of circulating tumour DNA as a hallmark of impending relapse, to allow us to think about structured adjuvant clinical trials that will for the first time, I hope, in lung cancer enable us to ask a very simple question: If we detect disease very early before there's evidence of disease on a CT scan, can we increase the chances of cure and direct adjuvant chemotherapy to those patients who most need treatment?


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.