AMP 2018: Interview with David Hillyard, Chair of AMP’s Infectious Diseases Subdivision
An important part of the Association of Molecular Pathology (AMP) is its Infectious Diseases Subdivision, made up of AMP members who have a professional interest in infectious diseases. The subdivision seeks to ensure the continuing education and professional development of those who use molecular diagnostics in the diagnosis and management of infectious diseases.
We spoke to Dr. David R. Hillyard MD, Chair and Medical Director of Molecular Infectious Diseases at ARUP Laboratories, about his work within the realm of infectious diseases and the increasing role of molecular diagnostics in combating pathogens.
FLG: What kind of work do you do in infectious diseases?
DH: I am currently the Medical Director for Molecular Infectious Disease (ID) Testing at ARUP Laboratories. Our institution is lucky to have a diverse team of ID specialists, including experts in transplantation medicine, bacteriology, virology, parasitology, fungal disease and mycobacteriology. In addition to existing methodologies, our team has recently focused its attention on the potential for using next-generation sequencing (NGS) methods for ID diagnosis and management. We’re working to increase the scope of these advanced methods and integrate them into existing test algorithms. Like many scientists, we’re very focused on the ever-critical question: ‘what is the value of this testing, how do we measure that value, and where do we take it in the future?’
It has been my great fortune to practice clinical pathology in an era that has seen the introduction of molecular infectious disease testing, its technical evolution and its dramatic impact on medicine. In the early 1980s there were essentially no molecular ID tests, not even polymerase chain reaction (PCR)! Like many, I’ve been a strong advocate for molecular ID testing across a broad spectrum of application and for careful clinical validation of these tests to appreciate both their strengths and vulnerabilities.
Today, pure molecular technologies are run in parallel with classic virology and protein-based detection methods, including serology and mass spectroscopy. Molecular ID testing has recently progressed from single pathogen detection and quantification to multi-pathogen panel testing and pathogen genome sequencing.
FLG: How are molecular diagnostics developing in light of genomic studies on pathogens?
DH: As is the case in many areas of science, new technologies are brought to clinical practice by the pioneering work of basic researchers. In the case of genomics for infectious diseases, one of the key initiatives has been to determine what can be learned from genomic information to create a more accurate microbial taxonomy. Genomics have also been key to the discovery of new viruses that may be causing disease, better understanding the organisation of genomes, and even simple applications like identifying the best regions for PCR targeting. The exploding field of microbiome research and clinical application has depended on having well-validated sequencing resources. Imagining other potential impacts of next-generation technologies is a central focus for individuals practising in this clinical space.
A recent progression we’ve seen with traditional molecular technologies has been an evolution from monoplex to multiplex panel testing. The many available syndromic panel tests have probably been as revolutionary as any recent development in infectious disease diagnostics. They employ several different technologies and allow clinicians to look much more quickly and with greater sensitivity for suspected pathogens. A critical follow-on effort is to better understand the clinical and financial value of these tests.
Now, in the age of advanced NGS platforms including Ion Torrent, Illumina, and other emerging modalities, clinical pathologists are exploring the possibility of using genomic analysis to establish a clear diagnosis. We need to be able to look for every potential pathogen in a body compartment rather than using a limited panel of 20 or 30 pathogens we suspect may be causing disease. That’s been going forward full steam for at least five years in several labs. The goal is to bring NGS testing into routine clinical practice as common orderable tests with clear indications for utilisation and with FDA approval.
Researchers are betting there will be a continuing ‘Moore’s Law’ of technical progression with rapidly increasing speed and lowering cost for NGS. That would invite the question: ‘why order a syndromic panel test when NGS has the potential to reveal any potential pathogen?’ For now, I don’t think there is enough compelling evidence to suggest that the technology is as reliably sensitive as current PCR methods. In addition, there are certainly other issues including the cost, complexity, and reliability of these technologies. However, once a high level of performance is achieved, sequence-based testing could certainly have a big impact. It would likely not be as affordable as current PCR-based tests, but could potentially obviate some of the need for multiple sequential test ordering and provide a more comprehensive and impactful answer.
NGS is currently being positioned as a test to run when everything else fails. The idea is to perform traditional syndromic testing on various body compartments which can be completed in one to two hours. The success of this strategy and its potential use as a more ‘first line’ test is a point of great interest for ID clinicians and laboratorians. These are the ideas that are increasingly being presented at the infectious disease meetings. Right now, no one really knows how far this technology can go.
In addition to advanced pathogen detection, an important problem to solve would be figuring out what accounts for mild versus severe disease, and identifying good versus very poor outcomes in infected individuals. Chris Woods’ group at Duke has been looking at what can be learned predictively from measuring molecular markers of host response. They’ve been using genomic technologies at the discovery level and are working to translate this knowledge into advanced clinical tests to guide patient management. Dr. Gregory Storch at Washington University is another leader in this field, asking similar questions for different populations and disease scenarios.
Right now, we don’t have rapid and highly informative tests for these assessments. As a field, investigators are doing what has been done in related research fields. They are pairing down candidate transcripts and other biomarkers to a few that are most predictive to allow for informative clinical testing in practical formats.
FLG: What kinds of changes do labs have to make, and have had to make, to integrate molecular diagnostics for infectious diseases?
DH: Looking back 20 years, all phases of molecular ID testing development have evolved and improved. Not that long ago, the preponderance of clinical ID testing relied on classic serology, enzyme immunoassays, and a variety of culture methods. The introduction of PCR, multiplex PCR, and sequence-based testing came gradually, and required the development of not only new technologies but guidelines for testing and test utilisation.
This is where AMP (the Association for Molecular Pathology) really made great contributions. AMP identified as its primary mission the creation of a society of experts to help transition molecular testing into the real clinical space across the spectrum of clinical pathology. It has been a challenging journey that is becoming more formalised and routine but also much more complicated given the challenges of advanced instrumentation, informatics and clinical algorithms. We now have much more reliable instrumentation, some of which is highly automated in formats which are smaller and sometimes even less expensive. The high demand for molecular ID testing has resulted in the development of much more sophisticated machines, including random access instruments akin to those used for clinical chemistry. Samples can be triaged based on urgency for quick testing, and with the capability of performing many assays on the same test run.
Major manufacturers like Hologic, Abbot, and Roche are moving very strongly in this direction. Given the continuing shortage of laboratory technologists throughout the country, the availability of these instruments also allows for greater productivity with fewer human resources. Even at high volume labs laboratories like ours, which have traditionally developed their own laboratory developed tests, there is great interest and motivation to utilise these platforms using their ‘open channels’.
Early in the molecular ID journey, a small subset of laboratories adopted Sanger sequencing principally for viral resistance testing. Most labs still use this methodology as the backbone of resistance testing though the large reference and academic labs are looking to adopt NGS for this and other applications. Early indications are that NGS can be a reliable and cost-effective platform for clinical ID testing. Our lab is very optimistic about this prospect but is preparing for the challenge of a much more complicated format of testing. NGS requires a whole new standard of care to avoid contamination and manage the large number of steps required to generate a test result. Although new pore-based methods of NGS sequencing have the potential to simplify the test process, they are not yet ready for routine testing. In addition, all NGS tests require a whole new informatics process to assure rapid and accurate mutation assignment.
FLG: What is the scale of costs associated with all of that?
DH: We live in a flourishing age of invention and manufacturing which has led to the creation of many new test platforms and increasing industry competition. In addition, many technologies have come off-patent, allowing laboratories that are capable of building their own tests to perform them much more cost-effectively than in the past. However, while the absolute cost of testing has fallen, test reimbursement has fallen as well. There are inconsistencies in what a test result costs to produce versus what a test can be reimbursed for these days and there is strict attention to the price and value of clinical tests, including molecular ID assays. However, comparing medical resources, laboratory testing across its broad spectrum represents a small fraction of total health care costs. Indeed, its proper use including molecular ID testing is well proven to reduce health care costs. Increasing pressures on molecular testing are expected and innovative approaches to make this sustainable are underway in many labs.
FLG: Once the laboratory is up and running, what is its overall economic impact?
DH: Historically, labs have been resource centres for hospitals and clinics. This has changed dramatically in recent years, especially for small laboratories which struggle to provide full resources at competitive prices. This has led to the amalgamation of small labs to achieve efficiencies of scale with reliance of reference labs for highly specialised work. We expect to see new disruptive technologies that shape the cost and utilisation of testing. Another important phenomena will be the continuing expansion of point of care testing. It potentially affords a rapid and cost-effective alternative to sending samples to hospital or reference laboratories, and when well-managed, a high-quality test result. It has even been envisioned that NGS can eventually be positioned as near point-of-care assays, particularly those employing nanopore technology.
A general question for NGS will be its competition with increasingly affordable and diverse multiplex tests. The case has been made that for a patient in an ICU bed costing $5,000 a day and potentially harbouring a life-threatening infection, that strong NGS test reimbursement is justified if results can provide a clear diagnosis and rapidly identify appropriate therapy. If this kind of performance is achieved, there will be competition for more and more NGS pathogen tests which will hopefully lead to lower test and patient management costs.
FLG: What are you most looking forward to seeing within the realm of infectious disease as a result of integrating molecular diagnostics into the lab?
DH: I’m very excited about point of care testing. If properly guided it allows for very rapid diagnosis and effective clinical decision making. Moreover, there is a positive psychology having patients empowered to utilise rapid testing that provides results during the clinical encounter and even privately at home. Developers have imagined near point of care testing for NGS as well using pore-based technology. We’re still a long way from that that scenario but it is a real possibility. I’m also excited and optimistic about having versions of current state-of-the-art molecular multiplex testing available in smaller and smaller boxes, closer to point-of-care and available for point-of-care testing.
Finally, I’m very excited about the emerging work on host response. Individuals vary dramatically in ability to deal with a serious infection due to differences in host response. Progress in this field would provide a whole new approach for managing patients that is personalised based on their unique individual biology.