Driving Discovery

Drug discovery needs all the help it can get. NGP talks problems and solutions with four gene genies: IQuum’s Shuqi Chen, Sequenom’s Elizabeth Dragon, Cellumen’s D. Lansing Taylor, and X from Assuragen.

NGP. Pharmacogenetics is still in its infancy but what do you think is causing the biggest buzz in the industry right now on the road to personalized medicines?
ED. The first area that I think is causing a lot of buzz is the changing regulatory environment. Historically, diagnostic tests have received less regulatory scrutiny than drugs and devices. As diagnostic tests are more broadly adopted and integrated into healthcare decisions, the FDA will increasingly want to review these tests. Things are still early on and it is unclear what the final regulatory environment will look like but this area will undergo substantial change in the next two to three years.

Another area of interest is the expanding pipeline of diagnostic tests. In areas with a long history of diagnostic tests such as oncology, we are seeing tests becoming more sophisticated. Instead of single marker tests, we are seeing marker panels and algorithms that incorporate many types of information (e.g. SNPs, gene expression levels, and epigenetic changes). We are also seeing molecular diagnostic tests expand beyond the traditional areas such as oncology and infectious disease. There are now molecular diagnostics in development for diabetes, stroke, transplantation medicine, and pre-natal diagnosis to name a just a few promising areas.

SC. Pharmacogenetics has the potential to improve drug therapy by preventing adverse events and improving patient outcomes. This very exciting, and initial developments in this area have shown promise. Yet, this is still a first step towards personalized medicine. To truly realize the vision of personalized medicine, we must go beyond variations in drug metabolism and take into account the broader application of genomic technologies to drug discovery. This would truly enable the physician to match a patient with a drug, and therein lies the potential to change the practice of medicine as we know it.

To achieve this goal, the two areas that deserve the most attention are bioinformatics and molecular diagnostics. We need to have a broader and deeper understanding of the molecular basis of disease before we can truly start designing drugs based on genetic information. Additionally, pharmacogenomics accentuates the need for novel therapeutic strategies that combine drugs and diagnostics. As such, we must also develop new genetic detection technologies and systems that can be used for this purpose.

ASSURAGEN. Economics – the fact that pharmacogenomics could dramatically improve the number of compounds being found safe and effective, particularly in Phase I and Phase II. By using the detailed molecular knowledge that comes from the discovery phase with virtually every new small molecule or biologic, appropriate biomarkers can be identified and used to better target populations of responders, or screen out patients with potential for adverse effects, or get clearer markers of response. Any one of these advances can not only reduce the number of patients or study duration, but may mean the difference between success and failure. Molecular biomarkers and their potential impact as companion diagnostics will become the standard for drug development and future pharmaceuticals. Without biomarkers, drug development will become increasingly difficult. The message is clear: No biomarker - no drug.

DT. A major issue is whether the best diagnostic/theranostic tests will involve correlative measurements of gene expression profiles from patient DNA or selected protein levels in serum versus more functional panels of biomarkers in patient cells and tissues.

NGP. The complexity of finding gene variations that affect drugs is a major challenge but what else are your clients increasingly citing as some the most significant obstacles they need to overcome?
DT. The two key obstacles are the existence of more relevant cellular systems that better model the disease that can be used early in the discovery process, and better cytotoxicity profiling that can be applied at least at the lead optimization step.

SC. Currently, there is still a large gap between finding valid genetic variations and effectively treating a patient. Filling this gap remains a key challenge. We all realize that finding genetic variations, correlating and validating the variations to a disease, and developing a drug that can exploit these variations is highly complex. Yet, how to utilize that new found information in the clinic to treat a patient is also very challenging. Physicians must have not only the drug, but also the tools and the information to guide the use of the drug. As personalized medicine develops, it becomes increasingly important for physicians to identify a patient’s genotype at the clinic and rapidly receive actionable diagnostic information. At present, the lack of these capabilities is a significant obstacle.

ED. Twenty years ago, it was thought that much of human disease would be explained by simple genetic mutations. For the diseases being studied at the time such as sickle cell anemia, this was true. However, as we’ve tried to understand more complicated disease such as diabetes, hypertension or cancer, we are finding that things are much more complicated. For most people, these diseases are probably the result of multiple mutations whose impact is modified by both epigenetic phenomena, and the environment. The real challenge is developing clinically relevant tests that incorporate all of these data and can be done in an inexpensive, robust, and high-throughput way.

ASSURAGEN. The first is absolutely data quality. Too many studies, particularly using RNA, have been plagued with problems because of sample handling errors, lab inconsistencies, or simply poor experimental design. Asuragen’s success depends on our customer’s success so we are motivated to deliver the best possible data to every customer on every job. The second, particularly in the clinical domain, is value. The cost/benefit ratio of pharmacogenomics has not been adequately considered in many studies. While data quality is essential, we further strive to understand the clinical context of a study to provide a cost-effective solution. Finally, although the FDA has repeatedly issued requests for genomic data, many pharmaceutical companies have perceived risk with submitting data they cannot fully explain. However, the FDA is continuing educational programs to demonstrate that in fact the opposite is true – more knowledge, earlier in the drug development process, will reduce risk and speed development.

NGP. Molecular diagnostics has the potential to help combat a host of diseases but in which area in particular are you seeing the most exciting or promising developments taking place?
ASSURAGEN. Oncology has the most potential and is the only area with proven success. For example, Genentech/Roche’s Herceptin and the companion marker Her2. The expression of Her2/Neu, a biomarker detected by fluorescence in situ hybridization and RT-PCR, is used to determine whether a patient will benefit from the drug. For those 25-30 percent of patients with ERBB2 (Her2) amplification, Herceptin plus other chemotherapeutics reduce recurrence by 50 percent and the risk of death by 30 percent. Another example is Novartis’s Gleevec. During clinical trials of the drug, patients with Human Philadelphia chromosome-positive (Ph+) Chronic Myeloid Leukemia (CML) in late chronic phase were treated with Gleevec after failing conventional therapy. An estimated 88 percent of patients were maintaining their major cytogenetic response 2 years after achieving it.

Extensive economic analysis has been performed to determine the cost-effectiveness of the application of Herceptin with its accompanying companion diagnostic, HercepTest (immunohistochemical test) and/or PathVysion (fluorescent in situ hybridization or FISH test). Approximately 20 percent of patients with Her2/neu amplification have been shown to benefit from Herceptin treatment. Conversely, patients with no amplification do not respond to Herceptin and consequently should not receive this therapy.

One study showed that the initial HercepTest followed by FISH confirmation of all positive results yielded an incremental cost-effectiveness ratio (ICER) of US$125,000 per quality-adjusted life-years (QALY) gained. The incremental cost-effectiveness of performing FISH in lieu of HercepTest was US$145,000 per QALY gained, thereby demonstrating that a molecular companion diagnostic proved optimal. Her2/neu analysis by FISH has since become the gold standard for prescribing Herceptin therapy. FISH is more costly than HercepTest; however, economic modeling shows that it is the more successful and cost-effective total healthcare option. This strategy of releasing a therapeutic along with a companion diagnostic that adds prognostic, disease stratification, or other pertinent information, can lead to fast tracking with the Agency, rapid adoption by clinicians, better patient outcomes, and reduction in overall healthcare costs.

This kind of compelling science and economics will continue to drive molecular diagnostics. And new tests, such as Genomic Health’s Oncotype DX, will further show the value of RNA-based tests, which for oncology provide better molecular characterization than DNA or protein-based tests.

SC. The potential to take molecular diagnostics out of the specialized laboratory and into the near patient setting is perhaps the most exciting trend. As a medical tool, molecular diagnostics has so much potential and its use at the point-of-care will benefit both the clinical diagnostic and the pharmaceutical industries. As drug discovery takes advantage of biomarkers, pharmacogenetics and pharmacogenomics, molecular diagnostics become critical. To be able to genotype a patient on the spot or precisely determine their disease state would give physicians the ability to determine suitable drug regiments in a single office visit and would truly allow the pharmaceutical industry to enter the personalized medicine arena. Yet, this capability is virtually non-existent right now. Currently, nucleic acid testing can only be reliably conducted in specialized laboratories by highly trained technologists using complex systems or manual processes. Therefore, developments that can automate the process as well as ensure the quality of results when such testing is performed outside the lab hold the greatest potential.

DT. The most exciting area today is in oncology. There is the opportunity to compare the genomic, proteomic and functional biomarker profiling in cells and tissues. Samples from tumor banks are available for a number of important cancers including breast cancer.

ED. While molecular diagnostics are spreading into many therapeutic areas, I believe that oncology and infectious diseases will continue to shape the leading edge of molecular diagnostics.

In both of these areas, molecular diagnostics will play increasingly important roles in diagnosis, treatment selection, treatment monitoring, and disease recurrence.

Within infectious diseases, HIV viral load and genotyping is a classic example of how molecular diagnostics have been integrated into treatment selection. As new diseases emerge, new drugs are developed, and organisms develop resistance to established drugs, molecular diagnostics will become increasingly important. Infectious diseases are unique in that the sheer numbers of organisms allow for very rapid changes in their genetic make-up. Flexible genomes require flexible technologies that allow for rapid, sensitive and quantitative measurements of these organisms.

In many ways, cancer is similar to infectious disease. Although derived from a person’s own cells, like an invading organism, tumor cells are present in large numbers and mutate in response to therapy. Similar to the genome of a pathogen, the genome of a cancer cell is critical to determining how to treat the cancer. In fact, therapies such as Gleevec work because of a specific mutation within a cancer cell. In addition to targeting specific genes, drugs that inhibit methylation or change histone acetylation target gene expression on a global level. Beyond treatment selection, better methods to detect cancer early are critical to increasing survival rates. Novel technologies that are sensitive, quantitative, and versatile will be key to advancing cancer therapy.

NGP. What solutions does your company offer and how are they benefiting your clients?
SC. IQuum is developing automated and rapid nucleic acid testing solutions that will satisfy the need for decentralized molecular diagnostics. Based on the company’s proprietary lab-in-a-tube platform, our Liat Analyzer will enable minimally trained healthcare practitioners to perform sample-to-result nucleic acid testing at the point-of-care in 30 minutes to one hour. The system reaches a level of automation, integration, miniaturization, and assay uniformity that is currently unmatched in the diagnostic industry. Additionally, the Liat Analyzer has incorporated robust features to reduce the potential for user error and ensure the quality of test results. These features are especially important to enable reliable testing by non-specialized personnel in non-laboratory environments.

Indeed, we believe that the Liat Analyzer is the ideal system for nucleic acid testing at the point-of-care. For pharmaceutical applications, the system can be readily used for on-site genotyping to provide physicians with immediate actionable information to guide drug regimen selection, or viral load quantification for therapeutic monitoring. The system may also be used at clinical trial sites to determine a potential participant’s genetic profile, and help the clinical trial investigator make inclusion or exclusion decisions, as well as monitor a participant’s response after treatment. By enabling such testing to be done while the patient is in the clinic, the system is expected to offer significant benefits to patient therapy and drug development.

ASSURAGEN. Our capabilities are broad, but our clients generally use us for either biomarker discovery or to develop a mechanistic understanding of a compound or gene. We have aligned our capabilities in experimental design, RNA isolation, RNA analysis, and bioinformatics to provide fully-integrated, partner-level support in these two areas. This focus combined with our client-oriented approach and RNA expertise keeps us best-in-class. As a spin-off of Ambion, Asuragen benefits from its extensive experience developing RNA technologies including RNA labeling and detection methods, bioinformatics, and diagnostic assay development.

The two keys to biomarker discovery are choosing an appropriate analyte (DNA, RNA, or protein) and being able to discriminate subtle changes in the target analyte. For example, within our Discovery group, we continuously look for potential diagnostic analytes for specific disease indications. We rely on the informatics expertise within the services unit to guide those choices, often based on published data. Then when we undertake an experimental program to identify a biomarker, services’ brings together experts in informatics, RNA isolation, and RNA analysis to ensure we maximize the power of the experiment by minimizing variance and ensuring our experimental design will provide a definitive result. For example, if a client wants to develop a biomarker or signature for molecular staging of a tumor based on archival fixed tissues, we can suggest how to compensate for the increased inherent variability in these samples by selecting a sample size, RNA isolation, and RNA analysis methods that will be able to support the biomarker before we undertake the study. If the client then wants to validate the signature in prospectively collected fresh frozen tissues, we would expect reduced variability and could therefore suggest a smaller sample size. We might also guide a client to using different analysis methods between fixed and fresh tissues, and be able to provide equivalency information between the two methods so that the results can be interpreted together.

In the case of mechanistic understanding, our solutions take on a different character. These experiments are often go/no-go decision points, so turn-around time is critical. In this case, we guide our clients to design their study and use analysis methods that are widely accepted, as demonstrated by published studies that may be similar to our client’s, saving precious time in experimental design and deliverables. The typical experiment here would be an in vivo experiment, such as comparing a cell line with a gene knocked-down by siRNA or comparing untreated primary cells to those treated with a compound. Again, by properly defining the experiment upfront and powering it with sufficient replicates, we apply the right RNA isolation and analysis techniques to get the best quality data in the shortest possible time. Then we apply our understanding of the experiment to properly focus on the genes of interest, set within the context of biological pathways. We do this using the industry-leading pathway analysis tool, Ingenuity Pathways Analysis. Often these experiments can move from planning to completion – with full pathway analysis – in under a month. Knowing, within the bounds of a model system, how you’ve affected entire pathways can save months in discovery-stage go/no-go analysis.

DT. Cellumen is addressing the key problems of efficacy and toxicity from early drug discovery with ‘cellular models of disease’, through drug development with ‘cytotoxicity profiling’ and into the clinic with ‘patient sample profiling’. All of these are addressed through our cellular systems biology approach that recognizes that cells are integrated and interacting networks of genes, proteins and metabolites that bring about normal and abnormal functions.

ED. The MassARRAY is our core platform. Using proprietary MALDI-TOF technology, we have developed a system that we believe is ideal for applied pharmacogenomics. The system is extremely sensitive with a wide dynamic range, quantitative, high-throughput, cost-effective and versatile. We have done multiple head-to-head studies against competing platforms and found that the MassARRAY is as good or better than competitors. Further, while competitors were highly tailored for specific applications, our MassARRAY out-performed them using our standard off-the-shelf system. We have also launched a service business that increases our ability to provide tailored solutions for our customers. Customers now have the option to buy our platform or outsource specific projects to our Genetic Services group. Our Genetic Services group can also work with customers to translate basic science findings into diagnostic tests.

Our technology can be used for many different applications, for example genotyping, gene expression and methylation. For genotyping we use our proprietary iPLEX technology. This technology allows us to rapidly develop custom SNP panels comprising a few SNPs to several hundred SNPs at industry leading cost effectiveness. The platform is extremely robust and hundreds of SNPs in thousands of samples can easily be processed every day.

For gene expression analysis we offer our QGE (quantitative gene expression) technology. QGE can simultaneously measure up to 25 genes in a single reaction and allows us to offer custom expression panels ranging from several to several hundred genes of interest. This flexibility allows us to quickly validate microarray studies and our ability to compare with multiple external standards can dramatically improve data quality. QGE is also more sensitive and precise than competitors and is up to 80 percent less expensive. We have validated our QGE technology against multiple competitors and found that we can consistently offer the same or better data quality at higher throughput at a fraction of the cost.

For DNA methylation analysis we offer our EpiTYPER technology. This allows researchers to study methylation in ways that haven’t been possible before because of technical limitations. With EpiTYPER researchers can quantitatively assess the methylation status of multiple CpG islands in amplicons up to 600bp long in a single reaction. A researcher with a single MassARRAY could analyze tens of thousands of CpG islands in a single day. Like our other technologies, EpiTYPER is highly accurate, high-through put and cost-effective.

Using these core technologies we are preparing to launch multiple diagnostic tests in several therapeutic areas. Our SSI (signature sequence identification) application will allow for extremely sensitive, quantitative, and high throughput pathogen identification and subtyping. SSI’s mixture analysis capabilities will also allow us to quantitatively characterize pathogen resistance patterns in patients being actively treated.

We are also developing a suite of pre-natal diagnostic tests that would allow for the genetic analysis of a fetus from a simple blood sample taken from the mother. This technology has the potential to revolutionize the field of pre-natal diagnostics

NGP. In such a competitive environment such as life sciences, where does your company have the cutting edge?
ED. Our cutting edge is based on our proprietary MassARRAY platform and on our IP position for specific markers and marker classes. Fundamentally, the MALDI-TOF technology employed in our MassARRAY is more sensitive than optical readouts that require fluorophores. This sensitivity allows us to offer diagnostic tests that are extremely quantitative, low-cost, and high throughput. The sensitivity of the test also allows us to develop tests for very difficult tissue sources (e.g., paraffin embedded sections).

On the IP side we have developed a broad IP portfolio that covers our core technology as well as specific diagnostic tests. For example, we have secured root IP for non-invasive prenatal diagnostics. This IP is platform independent and will allow us to offer proprietary non-invasive tests such as Rhesus D, cystic fibrosis and Down syndrome. In addition to this IP, we are developing a broad portfolio that includes many other diagnostic tests.

SC. IQuum is the leader in the field of decentralized molecular diagnostics. Our Liat systems are specifically designed for clinical applications and set the standard for simplicity in molecular diagnostics. The advantages of our system include:

• Sample-to-results automation. The Liat Analyzer automates all testing processes including reagent preparation, target enrichment, nucleic acid extraction, inhibitor removal, amplification and real-time detection. This high level of automation and integration significantly simplifies molecular diagnostics and allows testing to be performed by non-specialized personnel.
• Rapid testing. This technology uniquely uses liquid flow and mixing to enhance reaction rates and allows for relatively large reaction volumes. For example, a 50µl rapid PCR can be accomplished within 12 minutes for 50 cycles. A genetic test from whole blood takes as little as 30 minutes from sample to results and a test for an infectious agent in plasma takes less than one hour.
• Truly closed system. The analyzer uses a disposable Liat Tube that contains all reagents pre-packed in unit-dose quantities. Once the sample is introduced, the tube remains closed for all test processes - no sample, reagent, or reaction mixture needs to be added or removed from the tube. The closed system approach avoids cross-contamination, reduces biohazard risks and allows nucleic acid tests to be conducted in non-laboratory setting, such as the point-of-care or physician’s office.
• Assay uniformity. Liat assays utilize real-time polymerase chain reaction and other proven assay chemistries. This ensures the quality of the test and provides comparability of test results to that obtained from laboratory automation systems or bench-top methods. The analyzer is also highly flexible and can accommodate a variety of assay formats, reaction volumes and sample matrices.
• Robust error prevention features. The Liat Analyzer incorporates such features as automated sample metering, remote monitoring, integrated error diagnostics and self-corrective capabilities. These features ensure the quality of test results and allow testing to be done reliably in non-laboratory settings.

With such advantages, the Liat Analyzer will allow, for the first time in the industry, any hospital, clinic, or physician’s office to perform nucleic acid testing in a fast and reliable manner.

ASSURAGEN. While DNA and protein-based biomarkers have preceded RNA biomarkers, RNA offers an increased value as compared to other kinds of biomarkers. RNA technologies such as microarrays and PCR are highly sensitive, robust and adaptable to high throughput analysis. It is also clear from clinical tests, including the Roche HIV viral load test, that RNA-based tests can be more sensitive and specific than protein-based tests.

What people may be less aware of is that RNA is actually a better interrogator of molecular processes than DNA testing alone. For example, the Genomic Health Oncotype Dx test, which assesses the risk of recurrence of breast cancer, examines the expression signature of twenty-one genes in cancerous tissue rather than specific mutations in tumor suppressor genes, DNA repair enzymes, etc. Bergh, et al., in a September 2005 PNAS paper discovered a 32-gene panel in breast cancer tissues that, in their words, “outperforms sequence-based assessments of p53 in predicting prognosis and therapeutic response”.

We believe that the difficulty of handling RNA and getting reproducible laboratory performance with RNA have prevented more similar discoveries. As a full-service partner, we can overcome those technical issues and deliver meaningful biological results with similarly powerful clinical impact. An RNA biomarker is both more sensitive and specific than a protein biomarker, and while DNA may predict response, RNA confirms response.

DT. Cellumen has assembled an outstanding group of experienced biotechnology scientists that originally developed High Content Screening, have developed the approach of cellular systems biology and have great experience with multiplexed fluorescence. In addition, Cellumen has assembled a terrific group of key advisors including Rosemary Mazanet, M.D., Ph. D., the CEO of Access Pharmaceuticals, Leroy Hood, M.D., Ph.D., President of the Institute for Systems Biology, and Franklyn Prendergast, M.D., Ph. D., the Director of the Mayo Clinic and Foundation Center for Individualized Medicine.

NGP. Regulations and compliance also dominate the industry. How do you provide the most effective tools without compromising on the stringent quality they demand?
DT. All profiling assays for cellular models of disease, cytotoxicity profiling and patient sample profiling are validated to the most stringent levels including standard statistical tests and the inclusion of measurement standards and QC data.

ED. In contrast to other technologies, MALDI-TOF directly measures molecules. This allows us to offer extremely accurate tests that avoid the pitfalls of other detection technologies. Further, since the it is extremely sensitive we are able to offer this accuracy without sacrificing sensitivity. We have also initiated a software redesign that will further accelerate product development cycles and ensure our ability to maintain compliance with FDA regulations for medical device manufacturers.

SC. Compliance to any particular set of Standards or Regulations is usually the minimum level of performance one must demonstrate in order to have a successful and effective product. Industry competition and medical demands require manufacturers to continually evaluate and improve their products. Static compliance to regulations and standards will not, alone, result in truly effective products in today’s market place. The most effective products are those that have identified a very specific need that is not currently being satisfactorily addressed.

IQuum’s Liat systems are designed to address the need for decentralized molecular diagnostics. For this market, quality is especially important, as tests are performed by non-specialized personnel in non-laboratory environments. Our systems take this requirement into account by incorporating many features that increase the reliability of the instrument, reduce the likelihood of user error, and ensure the quality of test results. We have also followed the FDA IVD development guidelines in the development our Liat systems. As a result, we believe that these systems will not only comply with regulatory standards, but also meet the market expectations in terms of performance and providing actionable diagnostic information within a competitive cost structure.

ASSURAGEN. Asuragen has been an FDA QSR and cGMP-compliant manufacturer of Active Pharmaceutical Ingredients for several years. In addition, we are certified compliant with ISO 13485:2003 and ISO 9000:2001. Consequently, as we contemplate existing and future regulations in our Service business, such as 21 CFR Part 58 (GLP) and ISO 17025, we already have the regulatory infrastructure necessary to build an efficient and compliant organization. To some companies, regulations are a set of rules to overcome or circumvent. But since we’ve lived with the stringency of cGMP for many years, regulations represent a set of principles we now embrace. There is no paradox in your question – compliance with regulations and providing our customers with the best data quality possible are perfectly harmonized in our business.

Shuqi Chen, Ph.D CEO and Founder, IQuum, Inc.
Dr. Chen has over 20 years of experience in blood research and medical device development and was previously a faculty member at Harvard Medical School. He is the inventor of the lab-in-a-tube technology and is currently leading IQuum in the development and commercialization of systems based on this innovative platform. Dr. Chen has a Ph.D. in Medicine, a M.S. in Biophysics and a B.S. in Theoretical Physics.

Elizabeth Dragon, Ph.D. Senior Vice President, Research and Development, Sequenom
Dr. Dragon joined Sequenom as SVP of R&D in May 2006. She has over 25 years of diagnostics R&D, management, and leadership experience. Prior to joining Sequenom, Dr. Dragon was at Roche Molecular Systems from 1990 to 2006, where she held many leadership roles of increasing responsibility, most recently as Senior Vice President of Global Standardization and Vice President of Diagnostics Development.

D. Lansing Taylor, Ph.D. President and CEO, Cellumen, Inc.
Dr. Taylor is the Founder, President and CEO of Cellumen, the Systems Cell Biology company. Cellumen offers products and services to accelerate the successful progression of candidate therapeutics from the discovery laboratory through the clinic. Before founding Cellumen, Dr. Taylor was the founder, Chairman, President and CEO of Cellomics, Inc., the company that created High Content Screening, from 1996 until 2004.