Liquid Biopsy Likely to Have 'Major Impact' in Breast Cancer

Theodore Bosworth


October 20, 2017

In a recent trial in advanced breast cancer, liquid biopsies, also known as circulating tumor DNA (ctDNA) analysis, identified patients most likely to benefit. If the experimental agent in this trial, buparlisib, had not been sidelined subsequently by unacceptable toxicity, ctDNA analysis probably would have gained its first routine indication in breast cancer management. This setback, however, appears to be temporary.

It is relatively clear now that circulating DNA analysis will have a major impact in breast cancer management.

"It is relatively clear now that circulating DNA analysis will have a major impact in breast cancer management," said Nicholas Turner, PhD, a consultant medical oncologist at the Royal Marsden Hospital in London, United Kingdom. Speaking in a special symposium devoted to precision medicine in breast cancer at the 2017 annual meeting of the European Society for Medical Oncology (ESMO), Dr Turner acknowledged, "No indications for ctDNA analysis have entered clinical utility at this time," but he explained why several different types of indications are likely to emerge.

In advanced breast cancer, the potential indications for free cancer DNA retrieved from the plasma include early signs of treatment failure, early signs of relapse after a complete remission, and evaluation of evolving mutational status in primary tumors or metastases, according to Dr Turner. He estimated that approximately 90% of patients with advanced breast cancer have detectable ctDNA. Very sensitive assays are required for most of these indications, but highly sensitive assays are available and will become routine if clinical utility is proven.

As cancer cells proliferate and die, the DNA is released and some of it makes its way to the circulatory system. According to Dr Turner, the level of circulating DNA in the blood is a surrogate for tumor bulk. Because published data associate high levels of ctDNA with poorer outcomes relative to lower ctDNA levels, ctDNA is thought to be a relatively simple and noninvasive tool for monitoring disease and its therapies.

"When trying to work out whether patients are responding to treatment, it is possible that ctDNA will perform better than we can do with current technologies such as CT scans and tumor markers," said Dr Turner. In studies with paclitaxel, for example, ctDNA "drops very rapidly in patients who are responding—much more rapidly than any traditional biomarkers of response—and then rises again some time before clinical progression on CT scan, and this potentially gives you a lead time that the tumor is becoming resistant to therapy."

In a substudy of the PALOMA-3 trial, ctDNA assays were built into the prospective design as a potential tool for differentiating those with the best response from those who did less well. The trial associated the CDK 4/6 inhibitor palbociclib with improved progression-free survival (PFS) relative to fulvestrant alone.[1] The ctDNA assays were taken the day before therapy was initiated.

"Those with residual ctDNA at high levels at 2 weeks did poorly, whereas those with suppression did well," said Dr Turner. He called this a proof of principle for the value of ctDNA in monitoring response to treatment, but the problem was that this experiment had "good evidence of clinical validity but low evidence of clinical utility." In PALOMA-3, no action was taken on the basis of this information.

What we need to show is whether we can alter the treatment in those predicted to do poorly in order to improve outcome.

"What we need to show is whether we can alter the treatment in those predicted to do poorly in order to improve outcome," Dr Turner explained. Although that step was not taken in this study, he indicated that it is likely to be taken in forthcoming studies.

The recent phase 3 trial with buparlisib, a PI3K inhibitor, may be a better demonstration of ctDNA clinical utility, according to Dr Turner. In this trial, called BELLE-3, ctDNA analysis showed that response rates were substantially better in patients with a PIK3CA mutation than in those who were PIK3CA wild-type.[2]On the basis of this finding, it seemed possible that ctDNA confirmation of a PIK3CA mutation might emerge as a routine tool in determining who receives buparlisib.

"But this is still provisional data because at this time, buparlisib is not going forward to registration because of issues of tolerability," Dr Turner said.

In another demonstration, the presence of estrogen receptor mutations was evaluated retrospectively with ctDNA to evaluate the potential of this tool for guiding endocrine therapy in breast cancer. Using plasma samples from the phase 3 SoFEA study,[3] which compared aromatase inhibitor strategies in second-line therapy for advanced breast cancer, it was shown that specific estrogen mutations as detected with ctDNA could differentiate likelihood of response of one aromatase inhibitor relative to another, according to Dr Turner. Because this was drawn from a single retrospective study, the finding was also called "provisional," but Dr Turner called it a proof of principle.

"ctDNA analysis in patients who might have estrogen mutations acquired through prior aromatase inhibitor therapy may be a particularly good use of ctDNA," Dr Turner said. "Because here, unless you have a contemporaneous biopsy, you may be potentially missing these mutations, which are definitely not present if you look back at the primary tissue sample."

In advanced cancer, having fresh DNA is critical.

In advanced cancer, having fresh DNA is critical because "the more we look at ctDNA, the more evidence we see of heterogeneity" in terms of the mutational evolution of the tumor, said Dr Turner. "We can potentially find resistant mutations with implications for therapy."

The full potential for this approach may not be realized in the near future, according to Dr Turner, but he listed a variety of examples of ongoing work isolating mutations that are becoming increasingly important for guiding treatment. These include not only mutations involving estrogen receptors, but also those involving vascular endothelial growth factor (VEGF) receptors, KRAS, and BRCA. He cited one recently published paper that suggested that specific BRCA mutations might identify patients particularly sensitive to PARP inhibitors or other therapies that target this DNA repair defect.

So far, the data demonstrating that ctDNA can improve outcomes by directing treatment to targetable mutations are "anecdotal," said Dr Turner, but this may change soon. He described a clinical trial in which patients with metastatic breast cancer are being matched to targeted therapies according to mutations identified with ctDNA. The trial, being conducted in the United Kingdom, is already enrolling patients.

In another trial that will be launched soon, moderate- to high-risk triple-negative breast cancer patients are being screened with ctDNA following neoadjuvant chemotherapy and surgery. Patients with high levels of ctDNA will be randomly assigned to the checkpoint inhibitor pembrolizumab or observation. The goal is to determine whether liquid biopsy can be used to deter or prevent relapse. It is this type of study that is critical for taking ctDNA forward into routine clinical use.

"We already know that high levels of ctDNA after neoadjuvant therapy have predictive value for risk of relapse," Dr Turner said. "But what we need to know is, is it useful? It is not enough to know that patients are likely to relapse. What we need to know is how this information can alter therapy to improve outcomes."

Dr Turner reports financial relationships with AstraZeneca, Novartis, Pfizer, and Roche.


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.