Masha G. Savelieff, PhD
Q: What are liquid biopsies?
A: Liquid biopsies refer to biomarker assays that query evidence of cancer cells in biological fluids. In particular, the most commonly studied liquid biopsies are circulating tumor cells (CTCs), which are intact cancer cells in the circulation, and cell-free DNA (cfDNA) or circulating tumor DNA (ctDNA), which are fragments of DNA that are outside of cells and present in the circulation.
Q: What advantages can liquid biopsies offer over pre-existing testing technologies? What disadvantages do they suffer from?
A: There’s a lot of excitement about liquid biopsies because they provide tumor genotype information through a simple blood draw, without requiring an invasive biopsy. Since an ever-increasing number of drugs target specific tumor mutations, having a way to identify these non-invasively is an exciting new opportunity. Additionally, because biological fluids are relatively easy to obtain, we can perform these assays repeatedly, allowing us to monitor how tumor burden and a tumor’s molecular properties change with treatment. A disadvantage is the amount of cancer DNA present in a blood draw, which is very small compared with that in a tumor biopsy. This represents a challenge from an analytical standpoint. Also, liquid biopsies, particularly the next-gen sequencing assays, can be quite expensive. However, taking into account the cost of surgical tumor biopsies, liquid biopsies can actually be quite cost effective.
Q: Can you tell me a little bit about your research focus in liquid biopsies?
A: Our focus has been on developing novel methods of detecting ctDNA, such as CAPP-Seq (CAncer Personalized Profiling by deep Sequencing), and on applying these technologies in the clinic. We’re also interested in measuring aspects of a tumor from a blood draw that go beyond the presence of specific mutations. One particular area I’m very excited about is using ctDNA to measure minimum residual disease (MRD), which refers to a small amount of tumor cells that can remain in a patient after treatment. Currently, we can’t distinguish patients with MRD from those who are cured. But if we could do that, we could better personalize treatment. Patients who are cured could be spared the toxicity and cost of additional treatment, whereas patients with detectable MRD could likely benefit from further treatment to try to eliminate the residual cancer.
Q: In what areas have liquid biopsies seen the most progress? What’s on the horizon?
A: I think the most successful tests have been for ctDNA, which are already in clinical use and part of standard-of-care for certain cancer patients. As for what’s on the horizon, I’m most excited about the data emerging on MRD and its potential for personalized treatment for early stages of the disease. I also think there are going to be improvements that will lead to novel applications, including for early cancer detection.
Q: What is the concordance like been between tissue and liquid biopsies?
A: The concordance between liquid and tumor biopsies depends on the method used and specific study—but ranges from 75 to 95 percent. There are two main reasons for discordance. The first occurs when there isn’t enough ctDNA in a blood sample. In this case one may not be able to detect all mutations present in a patient’s tumor. Those are the false negatives. On the flip side, you can detect mutations in the blood that you didn’t detect in the tumor biopsy. Those are often real tumor mutations that weren’t present in the tumor deposit that was biopsied due to tumor heterogeneity.
Q: Do you think liquid biopsies are ready to replace tissue biopsies?
A: I think in certain cases, we’re already there. In non-small cell lung carcinomas patients who progress on certain epidermal growth factor receptor tyrosine kinase inhibitors, we routinely do liquid biopsies for resistance mutations to determine whether we can start the patient on the next line of treatment. Only if we don’t detect the mutation by liquid biopsy do we perform a tissue biopsy.
For the majority of patients, however, we absolutely still need a tumor biopsy to make the diagnosis and measure things like protein expression and histological subtype. So, I think for most patients, tissue biopsy will continue to be a critical aspect of their initial workup. But I do foresee in the future that we will be able to test more aspects of tumor biology using liquid biopsies, so that more patients will be able to avoid or minimize the number of tissue biopsies.
Q: What are the technology’s future prospects as a diagnostic tool?
A: I think the prospects are bright and it’s likely that we will have clinically useful ctDNA screening tests, at least for some cancer types. It remains to be determined which patient populations will benefit most. Importantly, we need to perform large prospective randomized studies to prove that patients who are screened have better survival outcomes.
Q: How is health insurance coverage for liquid biopsies? What is needed to increase clinical use?
A: Several liquid biopsies have been approved by the Centers for Medicare and Medicaid Services in recent months, so it is possible to get these tests covered, at least for patients with advanced disease. Some of the earlier stage applications such as MRD will require evidence from trials to prove their clinical utility before they will be reimbursed. Therefore, a lot of clinical trial work remains to be done to bring liquid biopsies fully into the standard of care.
CAPP-Seq: An Economical, Ultrasensitive, Broad-Coverage Liquid BiopsyIn 2014, Diehn published a study in Nature Medicine on CAPP-Seq—a liquid biopsy technique designed to identify recurrent tumor drivers, including single nucleotide variants and fusions, from ctDNA in patient plasma. Unlike single target liquid biopsies for a specific mutation, CAPP-Seq comprises “selector” biotinylated probes designed to amplify ctDNA at 521 exons and 13 introns from 139 recurrently mutated genes in non-small cell lung carcinomas (NSCLC), with a coverage of approximately 125 kilo base pairs. That means it can detect somatic mutations in over 95 percent of NSCLC patients. When Diehn and his colleagues applied it to patient plasma samples, CAPP-Seq detected ctDNA in 100 percent of patients with stage II–IV NSCLC and in 50 percent of patients with stage I, at a specificity of 96 percent for mutations present at fractions as low as about 0.02 percent. The technique also employs features to minimize false positives from the potential impact of biological variability and stochastic noise. The study’s authors found that the levels of ctDNA quantified by CAPP-Seq varied proportionately with the tumor volume and could be used to monitor treatment response. It was also capable of detecting residual disease more accurately than radiographic means. Although originally designed to detect the most frequent mutations in ctDNA from NSCLC patients, the method can be extended to any other type of cancer characterized by established, recurring mutations.
Maximilian Diehn, MD, PhD, is an associate professor of radiation oncology at Stanford University. He received his bachelor’s degree in biochemical sciences from Harvard College and his MD/PhD in biophysics from Stanford University. He is a board-certified radiation oncologist and specializes in the treatment of lung cancers. Dr. Diehn’s research program spans laboratory, translational, and clinical studies. His areas of interest include cancer genomics, liquid biopsies, and lung cancer biology.