Liquid Biopsies Predict Early ER+ Breast Cancer Progression

Liam Davenport

June 07, 2019

CHICAGO — Analysing circulating tumour DNA (ctDNA) in liquid biopsies at the start of treatment could reveal which patients with advanced estrogen receptor-positive (ER+) breast cancer will develop early resistance to cyclin-dependent kinase (CDK) 4/6 inhibitors, say UK researchers.

The study drew on the PALOMA-3 trial, in which ER+, human epidermal receptor 2 negative (HER2-) advanced breast cancer patients who had stopped responding to endocrine therapy were randomised to palbociclib (Ibrance, Pfizer), plus fulvestrant (Faslodex, Astra Zenica), or placebo plus fulvestrant.

Overall survival data presented at the European Society for Medical Oncology (ESMO) 2018 Congress showed that adding palbociclib to fulvestrant improved survival by 7 months versus fulvestrant alone, but that the difference was not statistically significant.

However, some subgroups, such as patients who were sensitive to prior endocrine therapy and postmenopausal women, did show an overall survival benefit with the CDK4/6 inhibitor versus fulvestrant alone.

Genomic Markers

The new study, presented at the American Society of Clinical Oncology (ASCO) 2019 annual meeting, looked at whether genomic markers in circulating tumour DNA (ctDNA) in blood samples would be able to predict early progression.

Dr Ben O'Leary of the Royal Marsden Hospital and The Institute of Cancer Research, London and colleagues examined progression-free survival (PFS) in 310 patients with a baseline blood sample that could be assessed for gene mutations, circulating tumour fraction, and gene copy number gain.

They found that, for every 10% increase in the tumour fraction in ctDNA, the risk of progression increased by 20%, while mutations in the TP53 gene, which were associated with visceral and soft tissue metastases, increased the risk by 84%.

In addition, copy number amplification of the FGFR1 gene was associated with an almost three-fold increase in the risk of disease progression, although the patient number in this analysis was small.

Dr O'Leary told the audience that the association between a high risk of early progression and tumour faction on ctDNA, FGFR1 gain and TP53 mutations was "prognostic" and that there was no significant interaction with treatment.

He said that the results "could inform future clinical trials in advanced ER+ breast cancer to identify higher risk patient populations that might require escalation of therapy".

Real Impact on Survival

Co-researcher Prof Nicholas Turner, also from the Institute of Cancer Research, said in a news release that while targeted treatments such as palbociclib are "exciting" and are having "a real impact on survival for women with breast cancer", many tumours respond initially but then develop resistance.

He continued: "Our study found that a new genetic test could detect right at the start of treatment those women whose cancers were most likely to develop resistance quickly to palbociclib."

Prof Turner explained that their treatment plan could be adjusted accordingly, by "trialling additional treatments from the outset to try and prevent resistance, or planning for a switch to another treatment as soon as resistance develops".

For him, the crucial next step is to "assess in a clinical trial whether helping direct women's care with this new test can offer improved survival and quality of life".

Key Part of Toolkit

Prof Paul Workman, chief executive of The Institute of Cancer Research, emphasised that the development of treatment resistance by cancers is "the greatest challenge we face in improving patients' survival and quality of life".

"So-called liquid biopsy tests like this one are a key part of our toolkit in staying on top of cancers' adaptability and evolution, and picking up the earliest signs of drug resistance."

Dr O'Leary began his presentation by noting that, while adding CDK4/6 inhibitors to endocrine therapy has improved clinical outcomes in advanced ER+ breast cancer, some patients have early relapse while on treatment.

However, no biomarkers identifying which patients will have early relapse have been established.

There is also limited data on whether the multiple genomic features associated with a poor prognosis in early stage ER+ breast cancer translate into the advanced setting.

Dr O'Leary said that ctDNA fraction has nevertheless been associated with a poor clinical outcome in patients with triple negative breast cancer, and the technique is a promising method for genotyping cancer.


He and his colleagues therefore turned to the PALOMA-3 trial, studying plasma samples collected at baseline in 459 of the 521 advanced ER+, HER2- breast cancer patients who took part.

They used two custom amplicon sequencing panels to perform ctDNA analyses on the samples.

Specifically, they looked for mutations in 17 targetable driver and CDK4/6-related genes and calculated the tumour fraction, based on 827 common germline single nucleotide polymorphisms.

In addition, copy number gain was determined by assessing the amplification status of 11 well-known genes.

Excluding samples with insufficient DNA or in which there was assay failure, the researchers had 331 baseline samples for the mutation assessment, as well as 401 samples to determine the tumour fraction and copy number gain.

Both analyses could be performed in 310 samples.

Dr O'Leary underlined that, importantly, the effect of palbociclib in these subsets of patients was comparable to that seen in the overall trial.

The researchers found that tumour fraction >10% on ctDNA was associated with early progression in both the palbociclib and placebo groups.

Among patients who received placebo and fulvestrant, the median PFS was 2.8 months for those with a tumour fraction >10% versus 5.5 months for those with a lower fraction, or a hazard ratio for progression of 1.767 (p=0.0035).

In the palbociclib plus fluvestrant arm, the median PFS in patients with a tumour fraction >10% was 9.2 months, while it was 13.6 months in patients with a lower fraction, at a hazard rate of 1.616 (p=0.0038).

The most common mutations in the baseline samples were in ESR1, PIK3CA, TP53, and NF1.

As in previous studies, ESR1 mutations were significantly associated with postmenopausal status, endocrine sensitivity, previous aromatase inhibitor therapy, and the number of lines of treatment for metastases. In addition the team found an association with bone metastases (p<0.05 for all).

In contrast, mutations in TP53 were significantly associated with visceral metastases, soft tissue and lymph node metastases, and the number of disease sites (p<0.05 for all).

Analysis showed that, in the placebo group, TP53 mutations were associated with early progression, at a median PFS of 1.8 months versus 5.4 months in patients without such mutations, or a hazard ratio for progression of 2.262 (p=0.0038).

In the palbociclib group, patients with TP53 mutations had a median PFS of 2.7 months versus 12.7 months in those without, at a hazard ratio of 1.996 (p=0.0024).

Dr O'Leary pointed out that, while the prevalence of copy number gain in patients with a tumour fraction >10% was, at 38.9%, similar to that in primary breast cancer, a limitation is that a higher tumour fraction is needed in the first place to detect such gains on ctDNA.

Nevertheless, the data revealed that gain of FGFR1 was associated with early progression, at a median PFS in the placebo group of 1.8 months versus 4.8 months in those without FGFR1 gain, at a hazard ratio for progression of 3.609 (p<0.0001).

In patients treated with palbociclib, FGFR1 gain was associated with a median PFS of 3.9 months versus 12.0 months in patients without, at a hazard ratio of 3.398 (p<0.0001).

Dr O'Leary noted, however, that the numbers of patients with FGFR1 gain, especially in the placebo group, were small, and so the results should be interpreted with caution.

Multivariate Analysis

Next, the researchers conducted a multivariate analysis on the 310 patients from both treatment arms to identify factors significantly associated with PFS.

To confirm the validity of the exercise, they examined the impact of palbociclib after adjusting for baseline clinical and pathological features, finding that the drug was associated with a significant reduction in the risk of progression, at a hazard ratio of 0.43 (p<0.0001).

They found that copy number gain of the FHFR1 gene was associated with a significantly increase in progression risk, at a hazard ratio of 2.9 (p=0.004).

Mutations in TP53 were also linked to an increased risk of progression, at a hazard ratio of 1.8 (p=0.0011), as was tumour fraction, at a hazard ratio per 10% unit increase of 1.2 (p=0.0001).

Overall, 5.3% of patients had FGFR1 gain, while 14.5% had a TP53 mutation, and 37.7% a tumour fraction >10% in ctDNA, indicating that, with some patients having more than one risk factor, 42.3% could be identified as being at increased risk of early relapse.

Additional Mechanisms

Invited discussant Dr Philippe Bedard, from the University of Toronto and the Princess Margaret Cancer Centre, Toronto, Canada, noted that there are many mechanisms by which breast cancer can become resistant to CDK4/6 inhibitors.

He said that the current study shows that "there are additional cell-free markers in the pre-treatment specimen that may be associated with prognosis".

The advantage of tumour fraction, he explained, is that it does not require knowledge of tumour variants, and that a high tumour fraction is associated with visceral metastases. As such, it may reflect the overall tumour burden.

Dr Bedard noted, however, that a raised tumour fraction was identified in only approximately one third of patients in the current population, compared with over half in a previous study of triple negative breast cancer.

Next, he said that TP53 mutation is a marker of aggressive biology that is associated with visceral metastases and a high tumour fraction.

Moreover, previous studies have indicated that it is associated with a more than two-fold increased risk of mortality in metastatic ER+, HER2- breast cancer.

He pointed out that, again, TP53 mutation has been detected in a third of patients but in only 13% here, "which may reflect differences in the patient population but may also be technical".

Dr Bedard explained that this may be because the amplicon assay only covers exons 5-8 but TP53 is a large gene, covering many different exons.

Finally, he said that detecting FGFR1 gain is a "technical challenge", although he noted that, here, the results were similar to those seen in MONALEESA-2, which he described as "reassuring".

Consequently, the data support the inclusion of FGFR alteration in, for example, genome-wide resistance screens and its detection in post-endocrine progression biopsies.

Dr Bedard concluded: "This is an important therapeutic area that may be the source of future clinical trials that can target this particular patient population."

PALOMA-3 was sponsored by Pfizer.

O'Leary reports Research Funding - Pfizer (Inst). Turner reports Consulting or Advisory Role - AstraZeneca; Bicycle Therapeutics; Bristol-Myers Squibb; Lilly; Merck Sharp & Dohme; Novartis; Pfizer; Roche; Tesaro; Research Funding - AstraZeneca (Inst); Bio-Rad (Inst); Clovis Oncology (Inst); Guardant Health (Inst); Pfizer (Inst); Roche (Inst).

Bedard reports Consulting or Advisory Role - Sanofi (Inst); Research Funding - Sanofi (Inst).

American Society of Clinical Oncology (ASCO) 2019 Annual Meeting. Presented June 1, 2019. Abstract 1010.


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.