By Diagnostics World Staff
March 16, 2016 | The 23rd annual Molecular Medicine Tri-Conference convened last week in San Francisco, Calif., featuring hundreds of presentations on issues facing scientists in molecular diagnostics, drug discovery, and computational biology. The six-day event covered too much ground to easily summarize, but Diagnostics World staff was there to catch pressing conversations on the design and use of molecular tests. Of particular note this year were the growth of point-of-care diagnostics, leading to more pointed discussions about their practical use; and swift advancements in analyzing biomarkers from whole blood, including circulating tumor cells and cell-free DNA.
New and Developing Products
James Lim, Chief Scientific Officer of Xcell Biosciences, noted that this February, the National Cancer Institute decided to terminate its use of 60 cell lines that had become entrenched test beds for cancer research. “These established cell lines had evolved and adapted genetically and phenotypically due to the steady diet of tissue culture plastic, animal serum, and normoxic culturing conditions,” Lim said. The inability to keep cells—especially from primary tissue—for any significant length of time in a state similar to natural cells in the body has consequences for the relevance of molecular tests, both in basic research and diagnostics. At the Tri-Conference, Xcell launched a new cell incubator, the Avatar System, meant to provide more natural microenvironments for growing and maintaining cell samples.
Joe Beechem, Senior Vice President for Research and Development at NanoString, presented new applications of his company’s optical barcoding technology for analyzing cell samples. Recently, NanoString has been previewing data that shows its instruments can be used for a unique form of DNA and RNA sequencing. But at the Tri-Conference, Beechem was focused on a new, more comprehensive version of his company’s targeted molecular tests, encompassing DNA, RNA, and protein targets. “The problem in cancer is, how do you get more biology out of smaller and smaller clinical samples?” he asked. With a forthcoming “molecular profiling microscope,” to be demonstrated in the coming weeks, NanoString expects to be able to profile tissue slides as small as six cells by six cells, illuminating both DNA mutations and antigens at specific sites on the tumor, primarily for immuno-oncology applications.
Kevin Coker, CEO of MolecularMatch, described his company’s vision for an online network connecting cancer patients with clinical trials through the results of molecular tests. “There is so much technology here [at the Tri-Conference] that’s being advanced very quickly,” he said. “But we’re also tied to a system that is failing at a great rate—the clinical trials system, which is responsible for bringing new products and therapies to market.” By developing a new ontology and search function for clinical trials that is sensitive to molecular test results, and placing those capabilities in hospital labs around the country, MolecularMatch hopes to increase enrollment and make it easier for patients and sponsors to connect.
Grace Zhao, Director of Research at AccuraGen, presented on the Firefly method, which her company uses to detect mutations in cell-free DNA with high sensitivity at the very low abundance found in serum samples. Firefly involves circularization of gene fragments in order to perform rolling circle amplification, building up sequencing-friendly levels of target DNA from nanogram samples. Addressing the same challenges, Maximilian Diehn from Stanford University presented CAPP-seq—closer to a brute force method with sequencing depth of 1,000 or even 10,000x over targeted gene regions, but with some refinements to both chemistry and bioinformatics to increase sensitivity to very rare mutations.
A very different use for cell-free DNA came from John Sninsky, Chief Scientific Officer of CareDx. His company performs testing for patients who have received organ transplants, to monitor for signs of transplant rejection—and, conversely, to make sure patients who are responding well to their transplants are not prescribed excessive doses of immunosuppressant drugs. Working with Stephen Quake of Stanford, CareDx has concluded that cell-free DNA levels can meaningfully track transplant response, with more successful transplants shedding less DNA into the bloodstream. CareDx has designed a targeted assay covering areas of high diversity in the genome, to maximize the chance of catching DNA that differs from donor to recipient, and measures the percentage of cell-free DNA from the donor organ as a marker for the patient’s immune response. This process, said Sninsky, increases the power of traditional tests based on direct immune markers: “The information from the recipient’s activated immune system, and the information from cells that are undergoing apoptosis from the transplanted organ, are indeed complementary to each other.”
Not all innovations in diagnostics on display at the Tri-Conference came from industry. Kamlesh Patel, a Senior Scientist at Sandia National Laboratories, spoke about work his agency is doing on a project called RAPIER, to sequence and identify bacterial DNA in the field. The project combines nanopore sequencing—a highly portable system commercialized by Oxford Nanopore Technologies and increasingly used for this type of microbial analysis—with in-house microfluidics innovations and bioinformatics.
Despite the incredible mobility of the nanopore sequencer itself, said Patel, “it still has a manual prep. You’re still doing pipettes, you’re still using benchtop systems to prepare all your samples.” At Sandia, his group has developed a system called the “digital microfluidic hub,” which lets users control the movement of droplets across a microfluidic path using embedded electrodes. In a video, Patel showed a lab technician sliding droplets along a tablet computer screen, while a microscope showed the same movements occurring inside the actual hub. The system forms an interface for sample preparation steps—mixing reagents, washing, and porting samples between other devices—and can even do thermocycling.
As a result, prototypes in the RAPIER workflow allow Sandia to extract bacterial-specific gene fragments from raw DNA samples and amplify that DNA for sequencing in a portable format. Sandia has also developed its own software to quickly match that sequence data to specific bacterial strains. The system is still in development, but Patel hopes it might someday form the basis for a commercial solution, with many more applications for field diagnostics.
Developments at the Point of Care
Tests that can be performed outside a laboratory setting, in doctors’ offices, clinics, and occasionally even at home, are growing more common and increasingly include some of the same complex molecular analyses that are mainstays of laboratory assays. While it has long been hoped that these tests will change the practice of healthcare, by expanding access and speeding up treatment, they are now prevalent enough to start assessing their effects. Several sessions at the Tri-Conference raised both the promise and potential drawbacks of point-of-care testing.
Elsie Yu, Associate Director of Chemistry at Geisinger Health System, gave a talk on “The Inconvenient Truth About Near-Patient Testing,” another term for point-of-care testing. Yu warned that diagnostic failures and misuse are more common at the point of care than in traditional lab tests, whether tests are used by patients themselves or by care providers. “Home use devices don’t have to be as accurate as hospital devices,” Yu said, because the consequences of errors are not as severe when a doctor is not making immediate treatment decisions. Nonetheless, errors can cause real harm. Pregnancy tests, for example, can give false negatives when used too late in the day or too early in pregnancy. Urine drug tests measure antibodies, which can exhibit cross-reactivity with non-drug substances; though they should only be used for screening purposes, employers may make hiring and firing decisions based on these tests without confirming the results with a mass spectrometry assay.
Newer trends, said Yu, are beginning to overcome accuracy issues. The ability to perform more precise molecular tests, especially based on polymerase chain reaction (PCR), with automated point-of-care devices is an important development. Even mundane changes, like using screens instead of colored sticks to display results, can greatly reduce the risk of operator error. Yet these developments also come with costs: point-of-care molecular tests are often far more expensive than traditional lab tests, and they can disrupt the flow of the clinic. “One of these devices we’ve used takes 18 minutes,” Yu said. “I know 18 minutes is not very long for a lab person like me, but 18 minutes is very long for a nurse.”
Of greatest concern to Yu is operator training in offices that are not accustomed to running molecular tests. Without naming specific care centers, she cited several instances she has seen of staff changing test protocols, failing to document results correctly, or running assays in areas with too much opportunity for contamination. This is not a matter of professionalism, but of a lack of training on lab standards in point-of-care settings. Yu recommended that individual offices partner with providers to work on best practices for point-of-care testing. “We believe that will improve operations and regulatory compliance, enhance clinical utility… and ensure accountability to quality of care and cost,” she said.
Following Yu’s presentation, Gyorgy Abel, Director of Clinical Chemistry at Lahey Medical Center, specifically addressed the cost problem in point-of-care testing. These newer tests, Abel argued, could save the healthcare system money despite their higher price tags―but only if their use in physicians’ offices does in fact encourage better preventive care that stops illnesses from progressing. Providers need to be attentive to how efficiently point-of-care testing is used, and which tests are preventing serious medical problems from developing. This kind of economizing might well be given a boost by the growth of value-based payments to providers, as opposed to fee-for-service. “This will put pressure on laboratories to use these services in a very rational and cost-effective manner,” he said. “You need to minimize the testing, up to the point that outcomes still do not suffer.”
In a separate session, Bela Matyas, Deputy Director of the Solano County Public Health Laboratory, discussed how point-of-care tests are affecting the thinking of public health agencies. Advantages that may be mere conveniences in major hospitals and affluent areas, like reaching a diagnosis and starting treatment on a patient’s first visit, are far more powerful for labs and clinics serving public health―especially in isolated rural areas where patients can be hard to reach. Public health officials also have responsibilities beyond the patients they are immediately dealing with. “One of the most difficult aspects of public health is finding the contacts of individuals who have communicable diseases,” said Matyas. “If the person’s right there in front of me, I can interview them to find out who their contacts are, and begin the contact tracing part right away.” It is also much less likely that labs will lose track of patients before they receive test results, making it easier to know where to concentrate resources to prevent diseases from spreading.
For public health laboratories, especially if they have a presence in the public clinics where patients most disconnected from the wider healthcare system seek care, the question may be not so much whether to adopt point-of-care testing as how to make the most effective use of these devices. In particular, Matyas advocated the automated processing and reporting of test results to agencies responsible for public health. “We can actually create, with these point-of-care devices, a distributed public health laboratory network,” he said. Devices placed for the convenience of patients, in pharmacies or grocery stores, could still feed data in real time to public health authorities, to respond to disease trends. Most importantly, direct links between public health labs and devices ensure that cases do not fall through the cracks, as often happens when many separate care centers are expected to do manual reporting.
Nonetheless, there are big challenges to making an automated system like this a reality. Providers need training to use point-of-care tests appropriately; there is still only a limited menu of tests that address the most critical illnesses for public health; and as with all issues in public health, it is very difficult to coordinate systems between the many different providers serving a given patient population. Fortunately, cost may be less of an obstacle to the adoption of point-of-care testing in public health than in other fields, given the huge savings that can be realized when health risks and local outbreaks are addressed early.
The Molecular Medicine Tri-Conference concluded on Friday, but Diagnostics World will continue covering some of the interesting molecular testing solutions we saw at the event. More information on the conference can be found at www.triconference.com.