Circulating tumour DNA (ctDNA) can be used to identify many patients who relapse after being treated for lung cancer. It can also be used to monitor how effective various treatments are for the individual subtypes of cancer cells and to change course if the treatment is not effective.
Blood samples taken from patients with lung cancer can be used to analyse ctDNA to identify those who relapse.
Analysing ctDNA can thus enable timely action within 4 months after initial treatment for lung cancer to identify the patients in need of additional treatment.
These perspectives resulted from a major international study in which researchers explored the clinical potential of monitoring patients with lung cancer by taking blood samples every 2 weeks to determine whether treatment is effective or whether they relapse. The samples also enabled researchers to monitor the development of cancer cells driven by various mutations and determine how treatment work on different subsets of cancer cells.
In the long term, the goal is to be able to perform adaptive treatment, in which doctors monitor the development of cancer in parallel with treatment response, making it possible to constantly treat the cancer with the most effective drugs.
The research has been published in Nature.
Monitoring tumour development in patients’ blood
The researchers developed a panel of 200 mutations that are characteristic of tumours among patients with lung cancer and that are present in circulating ctDNA. When tumour cells divide rapidly, some die, and ctDNA is released into the blood, which the researchers can detect.
In the TRACERx study, based at the Francis Crick Institute in London, the researchers examined ctDNA in blood plasma samples from 197 patients with early-stage non–small cell lung cancer, which comprises 85% of all lung cancer cases.
The researchers took and analysed plasma samples every 14 days preoperatively and at 3-monthly postoperative intervals during the first 2 years, followed by 6-monthly intervals between years 3 and 5.
“The purpose of analysing blood plasma samples for ctDNA is that we can do this often without unduly inconveniencing the patients or putting them at high risk. Other ways of monitoring tumour development among patients with lung cancer are repeated scanning or tumour biopsies from the lungs, neither of which are as practical or safe for monitoring patients with lung cancer as closely as we could in this study,” says Nicolai Juul Birkbak.
Relapse identified much more rapidly
People with lung cancer who have been treated usually have follow-up examinations, typically after 6 or 9 months, to determine whether the cancer has returned. Considerable time can thus elapse before the doctors discover that it has returned and the patients are treated again.
The researchers investigated whether they could identify the patients who experienced relapse after 4 months, and they succeeded for 49% of them after 4 months. About half the patients treated for lung cancer experience relapse.
“This means that we found signs of relapse among 25% of our patients. It is also interesting that we did not identify half of the relapse cases,” explains Nicolai Juul Birkbak.
More effective for some patients than others
In the next part of the study, the researchers tried to determine whether they could identify which patients tested positive for ctDNA in their blood in connection with relapse and which did not. The researchers examined patients for differences in the release of ctDNA between those with squamous cell carcinoma and those with adenocarcinoma.
All patients with squamous cell carcinoma released ctDNA into the blood, making relapse easy to identify. Patients with adenocarcinoma differed; even some of those with large tumours did not release detectable ctDNA and thus could not be identified through blood tests.
The researchers then examined the patients with adenocarcinoma for differences and found that fast growing tumours with cancer cells that both grow and die quickly released enough ctDNA to be detectable but not the tumours in which the cells grew slowly.
The researchers already found this difference between the tumours before initial treatment began.
“This means that we can determine which type of tumour the patients have based on the type of adenocarcinoma. We can use this knowledge later when we monitor the patients postoperatively to identify those with relapses. Thus, using ctDNA from blood plasma samples to identify early relapse works for some patients, but not for others. And with these new findings, we now already know upfront for whom ctDNA analysis will be relevant before treatment starts,” says Nicolai Juul Birkbak, who adds that about 70% of all cases of non–small cell lung cancer are of the adenocarcinoma type. Among the patients included, 62 had adenocarcinoma, of which 28 were of the slowly growing type.
Advancing knowledge on mutations
In the third part of the study, the researchers implemented phylogenetic tracking technologies for the tumours: not just monitoring the tumours but also examining them genetically during treatment.
Tumour mutations can be categorised as clonal or subclonal mutations. The clonal mutations are present in all tumour cells, and the subclonal ones are only present in a subset of tumour cells.
This part of the study showed that not all subclonal mutations were identified at relapse. Some of the cancer-specific mutations that were present initially had disappeared during treatment, and others had survived the treatment and were still present at relapse.
The phylogenetic tracking also showed that the mutations with the most copies preoperatively also had the greatest probability of being present at relapse.
“Again, this enables a risk profile to be created for each individual patient. ctDNA analysis of a baseline blood plasma sample among patients with lung cancer enables us to identify which mutated cells are most likely to metastasise. We can also monitor how effective treatments are on specific mutations, since taking repeated samples enables us to determine that some mutations have been eliminated and others have become more prevalent. We can use this knowledge to identify which treatments will be optimal for individual patients with specific mutations,” says Nicolai Juul Birkbak.
Specific combinations of mutations influence survival
The researchers used phylogenetic analysis, grouping mutations into individual cancer clones representing cancer cells that share a unique set of mutations, and found that prognosis differed among patients who differed in disease development.
The researchers monitored the development of these cancer cell clones on a phylogenetic tree over time and during treatment. The further the individual clones are apart on the phylogenetic tree, the less related they are and the more unique subclonal mutations they have.
The study revealed that the patients in which only one clone survived treatment and was present at relapse had the best prognosis.
The patients who had cancer cells from several clones but in the same branches of the tumour’s phylogenetic tree had a slightly worse prognosis, but the patients who had cancer cells from clones from several different branches in the phylogenetic tree at relapse had the worst prognosis.
The difference in survival after relapse was almost 1 year.
“Although we expected this, we can now prove it for the first time. This means that we know more about tumours and can help to organise better treatments in the long term,” explains Nicolai Juul Birkbak.
May soon be used in clinical practice
Nicolai Juul Birkbak does not think that phylogenetic ctDNA analyses of blood plasma samples from patients with lung cancer will become part of clinical practice in the immediate future, but he thinks that the technology has so much potential that it will probably be adopted quite quickly, perhaps already within 5–6 years.
Monitoring the development of the tumours in real time by analysing blood samples for ctDNA will enable doctors to constantly be at the forefront and adjust the treatment according to the results.
For example, one treatment may be really good at knocking out many clones but may not work on others. In this scenario, a doctor can quickly switch and use another more effective drug on the clones that survive.
“So far this is basic research, but this could become promising in a clinical setting,” concludes Nicolai Juul Birkbak.
