June 2016, Vol. 5, No. 5
Early Recurrence Detection: Analyzing DNA from Circulating Tumor Cells and Circulating Cell-Free DNA
An Interview with Paul W. Dempsey, PhD, and Paul Y. Song, MD, of Cynvenio
Cynvenio is a cancer diagnostics company offering their LiquidBiopsy technology to provide molecular analysis of cancer biomarkers in blood. The company purports that molecular diagnostic approaches to cancer detection and monitoring can benefit from their multiple-template strategy that analyzes DNA from circulating tumor cells (ctcDNA) and circulating cell-free DNA (ccfDNA) in blood, as well as from tissue biopsies. They recently published the results from their National Cancer Institute–supported clinical trial showing that data gained through next-generation sequencing of ccfDNA and ctcDNA liquid biopsy templates differ but complement each other and should be considered simultaneously. Analysis of each template enables the capture of different moments in a tumor’s evolution, providing additional and complementary information to that acquired through tissue biopsy alone. Moreover, the information gained from examining circulating tumor cells (CTCs) enables the identification of both relevant mutations and other biomarkers (such as proteins and RNA expression) that can drive a particular cancer.
Tumor-derived samples from tissue biopsy, CTC populations, and ccfDNA represent different moments in a cancer’s progression and may not derive from the same biologic sources. Research suggests that ccfDNA samples genomic DNA fragments released from all tumor sites and may capture mutations not specifically associated with the disease, as well as disease-relevant ones. In contrast, CTCs reflect the mobile subset of tumor cells in blood; they are clearly related to the disease process and predict more aggressive as well as metastatic disease.
Cynvenio is applying their multiple template approach in a new clinical research study in triple-negative breast cancer. To date, they have enrolled 150 women. Enrollment is still open.
The publishers of PMO had the pleasure of speaking with Drs Dempsey and Song about their multiple-template strategy that analyzes ctcDNA and ccfDNA in blood as well as the initiation of their clinical trial in triple-negative breast cancer.
PMO Good morning, gentlemen. To begin, how do you define personalized medicine, and how does your company fit into this definition?
Dr Song I believe personalized medicine has been incorrectly defined as synonymous to genomic profiling. It was over 15 years ago that Human Genome Sciences was able to clone the human genome and yet the lack of advances since that time really points to the limitations of focusing solely on genes. I spoke recently at the Personalized Medicine World Conference and addressed the frustrations of clinicians trying to obtain approval from insurance companies to get a “personalized analysis.” But if you obtain this report, many times it doesn’t translate into any meaningful improvement for the patient. There may not be a targeted therapy for the identified gene, or maybe the genetic mutation is multifactorial, or perhaps you give a targeted therapy that still doesn’t work because of some other aspects of the patient.
I’ve tried to redefine or question what the medical field has defined as personalized medicine. I think personalized medicine goes beyond genes to include consideration of exosomes, RNA, proteins, and additional markers.
Dr Dempsey Personalized medicine comes in many different forms, and it’s not necessarily driven by the latest technologies but by information that’s most useful to a patient and his or her doctor. Tamoxifen is a great example of a targeted personalized treatment, and it’s been around for decades.
Our goal is to generate tools and information that can help doctors make the best decision for each of their patients. That has always been the doctors’ aim, but in oncology we’ve been unable to leverage information the way we do in other fields—infectious disease, for instance. In infectious diseases you have specific information that you can collect longitudinally and understand the progress of an infection and treat the infection as it moves.
I think therein is a gray line between group dynamics and individual personalization. We’re trying to provide information to the physicians so that they can make evidence-based decisions for the patient they have in front of them.
Dr SongThat describes our major objective with our technology. We want to move beyond current next-generation sequencing in genomics and offer the ability to find circulating tumor cells, to look for expression of surface targets, to interrogate the cells for RNA analysis, to look at proteins, to look at the serum for a secretome analysis, and the exosome.
Ultimately, we would really like to redefine what personalized medicine is, but to do that, we need to provide the community with enough tools to consider other markers beyond just DNA fragments.
PMOCan you describe your liquid biopsy test, the multiple-template approach, and how the 3 components of the test complement each other?
Dr Dempsey Our test is a monitoring tool that considers multiple different tumor-derived events. The components reveal different, but complementary, information.
They’re complementary in the biomarkers you can analyze. RNA, proteins, and DNA segregate differently in different compartments, but also, these are invariably rare events. The tumor-derived template in blood is a tiny fraction of all the information that’s actually there and, frequently, you are going to be at or near the limits of detection for these.
Having multiple templates that give you an increased chance of being able to generate useful information is incredibly important. Because the information comes from different mechanisms, they tell you different things about the disease.
It doesn’t serve the patient and the doctor to be restricted to 1 class of biomarker in any circumstance because there’s always additional information to be considered. So we’re looking at many different technologies in addition to what we’ve already built.
For example, as part of a study we’re doing now in triple-negative breast cancer, we’re using a technology that Dr Song brought to the United States and to Cynvenio from South Korea. It is a tool to measure the critically important immune response in patients. We’re successfully using this tool in our triple-negative study. This is a great example of how we’re incorporating technologies beyond the current platform to give us a broader picture of the disease.
PMO Can you describe your ongoing trial for triple-negative breast cancer and how Cynvenio’s technology is utilized in this study?
Dr Song Our platform has primarily been used in research, so my challenge as Chief Medical Officer was to show real clinical utility. I wanted to select a disease that would really highlight our platform in a way that was clearly unequivocal.
Triple-negative breast cancer is a very aggressive type of breast cancer. It affects about 18% of all women who are diagnosed with breast cancer. The sad part about this particular disease is that 1 of 3 women will develop metastatic disease in about 2.8 years after completion of therapy and, unfortunately, there’s no biomarker to monitor these patients.
There was a trial done in London last year that looked at 55 women with advanced breast cancer. They did traditional tumor profiling of specific genes, and they looked for evidence of those gene mutations in circulation with PCR [polymerase chain reaction] pretreatment and posttreatment. In those women who had disease in their blood posttreatment, 86% went on to develop a recurrence about 7 months after they found evidence of disease.
I wanted to expand upon their findings with our technology in a high-risk population. In triple-negative breast cancer, once these women complete therapy, they go back to their physicians every 3 months. The doctors poke around their nodes, then maybe draw some simple blood tests like liver function tests. But if the results are abnormal it’s too late, because that means the cancer has already spread to the liver.
What we decided to do was to check these women, monitor them, and look for evidence of disease in circulation. But unlike the London trial that used PCR and was very limited in what they looked for, we actually built our own triple-negative breast cancer–specific gene panel of 27 genes. We scoured the literature for any gene that had been reported in major publications that was associated with triple-negative breast cancer, and we created a customized panel based on our findings. We use our platform to not only look for cell-free DNA but also to look for circulating tumor cells. If we find circulating tumor cells we are doing next-generation sequencing, but we’re also planning to then look for other potential markers that could benefit the patient.
We’re looking on the surface for PD-L1 [programmed death-1 ligand 1] expression as immunotherapy is now emerging as a possible choice for these women once they recur. We believe that looking for PD-L1 expression on circulating tumor cells in the metastatic setting will be much more indicative of potential response to immunotherapy than the primary biopsy model, which looks at PD-L1 expression on the primary tumor specimen, because PD-L1 expression is an adaptive changing response. There may be little to no initial expression on the primary tumor, but as the disease progresses and mutates over time and eventually metastasizes, there may be a far different degree of expression that is much more indicative of the actual current disease state and a more accurate predictor of potential response.
The other thing that we are very close to rolling out is our ability to take RNA from the CTCs for subsequent analysis. Another advance that we’re very excited about is the ability to take that RNA and look for androgen receptors and splice the variants like AR-V7.
AR-V7 has primarily been reported for prostate cancer, but 90% of all breast cancers actually express androgen receptors. One of the emerging areas of breast cancer is the use of antiandrogen drugs that are typically used in prostate cancer. The AR-V7 analysis cannot be done with DNA or cell-free DNA; it has to be done strictly with RNA. The problem with the current methodology is that they have to use fresh blood, and many times patients and doctors cannot get the blood processed in a timely fashion, so the overall utilization of this information is not employed enough.
With our fixative process, we’re able to make it much easier for patients to get their blood drawn and for doctors to order this test, so we believe it’s going to actually increase the ability for people to incorporate this into their practice.
We’re looking at cell-free DNA, we’re looking at CTCs, and we’re also looking at incorporating another test that looks at natural killer cell activity with a simple 1-cc ELISA test. There is literature showing that when natural killer cell activity is low, there’s a much higher incidence of circulating tumor cells in patients with metastatic disease.
We opened the trial on March 3, which was national Triple-Negative Breast Cancer Day, and in less than 3 months we have enrolled 150 patients.
PMO How are the patients doing thus far?
Dr Song This is a patient population that’s highly motivated because 1 of 3 will recur. Sadly, we have found many patients with recurrence in the blood but with no clinical evidence of disease yet. We’ve also had 1 patient at the time of enrollment with no clinical evidence of disease, but 2 weeks after she enrolled she presented with a lump in her neck. We not only found evidence of disease in her blood on that first sample, but we also identified actionable mutations that her oncologist could then use to personalize her treatment.
PMO Can you treat based on the recurrence in the blood, before clinical evidence is present?
Dr Song No, once we find evidence of disease in blood, we follow the patient serially. Let’s say we do your analysis and we find you have a 4% TP53 mutation frequency or a PIK3CA mutation, and then in another 3 months it’s gone up to 12%. Part of our trial is to prove that we can detect recurrence. But once they recur, we want the doctors to use our information to help guide the next salvage treatment.
A long-term goal we have with hospital partners is to randomize patients once they’re found to have disease in the blood; giving half of them treatment and monitoring the other half to determine if early treatment translates into better outcomes.
Dr Dempsey Your question is actually about breaking the circle. We have a system right now where women are treated for triple-negative breast cancer, and then they’re told to wait. If we say we’re building a capability of detecting the recurrence earlier, the doctors will question what to do for these patients because the standard of care says they can’t treat until the patient presents with clinically manifest disease.
That’s rather barbaric. For the patient, it’s difficult to wait when you know that a third of your group is going to have disease recurrence.
So in order to break that circle, you have to start either by treating early or detecting early and then figuring out what best treatments you can use. This trial is initially about showing that we can detect the disease earlier, and then we’ll take on the steps of using that information to show that it can inform practical, specific, and personalized treatment decisions for those patients as they go into their care cycle again.
We published a study earlier this year where we show that both cell-free DNA and CTCs can be detected in blood; they can be shown to be related to the tumor tissue, and for any given sample you may see a signal in one compartment or the other or both.
So we’ve taken that technology, paired it with the immune monitoring test, and are looking to initially demonstrate that we can apply those technologies to capture and define the recurrence of disease at a time that’s earlier and therefore more likely to have a lower disease burden.
PMO Can you speak to the complexities of the multiple-template strategy?
Dr Song We’re really showing how all components of our platform work together to complement and really provide personalized medicine in a tangible way to help the patient.
So let’s say we have a patient with triple-negative breast cancer who recurs, and we’ve been able to find an EGFR mutation, which is normally associated with lung cancer, for which there is targeted therapy that the doctor can try. We can also look at their cell for PD-L1 expression if they want to consider immunotherapy. Or, if we find the AR-V7 mutation, then they may plan to switch to androgen therapy. AR-V7 can predict resistance to certain hormonal therapies. You can’t do all this with just 1 liquid biopsy that focuses strictly on cell-free DNA.
Dr Dempsey We also sequence the germline DNA in every patient so as to not confuse polymorphisms with acquired somatic mutations.
Now, we’re not making prognostic conclusions here, we’re making molecular conclusions based on various different biomarkers that we can access. Now when you put that together in a given example, you do see the difference of the different templates. So we have one example; it’s a clinical example where we had a patient with prostate cancer who had failed a series of chemotherapy cycles and presented for the liquid biopsy using the multiple-template approach.
We found mutations that were shared in both the cell-free DNA and the CTC compartment. But in the cells, there was a specific mutation in BRAF that has never been published before in prostate cancer. But it’s a targetable mutation that’s seen in other cancers and for which there are drugs available. The mutation was specifically present in the cells because the cells are mobile, moving and driving the disease process, and the cell-free DNA hadn’t caught up yet. Down the road you might see it in the cell-free DNA, but it wasn’t available at the time that the patient needed the information. So he was given a drug that targeted that mutation and had a progression-free survival window of over 9 months as a result of that specific information.
So it all builds on the technology and these monitoring tools that reveal biomarkers that are useful to the doctor and help to make clinical decisions.
Dr Song That’s the argument I’ve been making to clinicians. I think we are doing patients and physicians a disservice by limiting choices. Some people say cell-free DNA is the best way to go, or CTC is the best way to go, or exosomes. We’re agnostic. We want to provide as much information as possible to physicians and patients so that collectively it can complement one another. That’s the fight, and that’s our theology that we’re trying to evangelize.
PMO How accessible is this test for patients?
Dr Song Patients can simply go to any doctor’s office or even take the kit to a lab, and they can draw blood, and there’s a fixative process that stabilizes the sample for 96 hours before it needs to be processed. It really makes it very easy.
Now as far as third-party payers, it has been approved for breast cancer in the metastatic setting. Most third-party payers have reimbursed that. What we’re hoping for in the triple-negative trial is that we have compelling data that they will then allow approval for monitoring posttreatment in a very high-risk disease.
PMO Where do you see this technology fitting in the long-term of patient care?
Dr Song Speaking as a clinician, I think when more and more people realize what’s possible with the platform, including the ability to interrogate exosomes and analyze RNA, and more clinically relevant information is discovered from that platform, it will be used more frequently. And it’s important to note, our platform allows all of that to happen from the same 1 blood collection from the patient.
Dr Dempsey Over the next few years, Cynvenio will focus on 2 efforts. Having completed all the heavy lifting in terms of building the tech and the workflow, now we are focusing it in specific settings so that we can produce the clinical evidence that will ensure the uptake of the technology and the application of this ability in the clinical setting. This is where Dr Song’s clinical trials fit in.
So that’s one whole effort. And the other focused effort is to support distribution of the platform through our partners at Thermo Fisher Scientific so that the platform will go into institutions around the country and around the world. That will build many more applications by tweaking the system for different research settings, different clinical settings, and different cancers that we can’t reach ourselves. We’re supporting the distribution and education of this capability so that other people can access this technology as well and build their own tests.
PMO Thank you for your time today, and best of luck in this endeavor.
Dr Dempsey is Chief Scientific Officer of Cynvenio. He is an immunologist with more than 20 years of experience in biomedical research and analysis. Before joining Cynvenio, Dr Dempsey was an Assistant Research Professor in the department of microbiology, immunology, and molecular genetics at the University of California, Los Angeles.
Dr Song is Chief Medical Officer of Cynvenio overseeing translational and clinical research programs. He also serves as Chief Medical Officer at ATGen Co. Ltd, where he oversees all clinical aspects of the company’s NK Vue, an ELISA-based blood test that measures natural killer cell activity. Dr Song was formerly a faculty member at Cedars-Sinai Medical Center in the Samuel Oschin Comprehensive Cancer Center department of radiation oncology and biomedical sciences.
Dear Colleague, This issue of Personalized Medicine in Oncology (PMO) is a testament to how far we’ve come in our understanding of, and novel approaches to, the treatment of cancer. Armed with knowledge of molecular biomarkers, genetic mutations, genomics, immunotherapeutics, and targeted therapies, we are better able to treat patients [ Read More ]
Compared with other cancers, ovarian cancer is relatively rare, accounting for just 1.3% of all new cancer cases in the United States. However, it has a high death rate—the highest of any female reproductive system cancer. Only around 1 in 7 women with ovarian cancer are diagnosed at the local [ Read More ]