BB&T Contributing Editor
SAN DIEGO – During the past two decades, molecular diagnostics has progressed from a fledgling market of less than $5 million in the mid-1980’s to one that exceeded $2 billion worldwide in 2005. This growth has been driven by a number of technological innovations, such as polymerase chain reaction (PCR) amplification of nucleic acids, automated DNA sequencing, advanced high-sensitivity labeling and detection methods such as chemiluminescence, and, most recently, microarray technologies enabling the analysis of thousands of molecular markers in a single experiment.
The San Diego Conference, an event sponsored by the American Association for Clinical Chemistry (AACC; Washington), held here in mid-March on its 20th anniversary, has chronicled the development of molecular diagnostics from its earliest beginnings. During those two decades of innovation, a number of new companies have emerged that now are substantial players in the clinical diagnostics market, and major diagnostics companies have out-paced overall market growth in the diagnostics space by expanding into molecular diagnostics.
As discussed by leading experts in the field at the San Diego Conference, the expansion of molecular diagnostics is far from over. In fact, the potential exists for a breakout of the field over the next 10 to 20 years, similar to the developments in computers and communications.
According to Randy Scott, PhD, CEO and chairman of Genomic Health (Redwood City, California), one of the key factors that will drive this transition is the rapidly declining cost-per-unit of genomic information. As shown in Table 1 below, the cost of analyzing the human genome is projected to drop – based on current trends in analytical technology – from $300 million in 2000 to less than $300 by 2040. Furthermore, for clinical applications it is only necessary to analyze a small part of the genome of an individual, the functional genome, which comprises only about 1% of the total DNA, bringing the cost for clinical whole genome testing to under $1,000 as soon as 2010.
As costs continue to decline, the demand for genomic data will increase dramatically, ushering in the era of personalized medicine and consumer-driven testing. An additional factor expected to accelerate the adoption of widespread molecular testing is the increasing utility of genomic data as the total amount of information expands. As the amount of information available grows, the utility of the network of biological information increases faster than its cost.
In addition, because of the finite nature of biological systems, the time and effort needed to arrive at a solution to a biological problem diminishes rapidly as the solution is approached. That may allow an understanding of a clinical condition, such as tumor development, that now appears intractable, to be reached much more rapidly than expected.
As the field of molecular diagnostics has expanded, it has diversified from a clinical perspective from its initial focus on infectious disease testing to include cancer diagnostics, genetic testing, pharmacogenetics and hematopathology. A variety of new applications in cancer testing were described at the San Diego Conference, and numerous other developments in molecular cancer diagnostics have recently been announced elsewhere, that demonstrate the movement of molecular diagnostics into the mainstream in oncology.
Genetic testing for other diseases such as coagulation disorders and inherited diseases is also expanding rapidly. And the field will continue to be driven by technological advances in areas such as specimen preparation, amplification, sensors and microarrays.
Expanding applications in cancer diagnostics
New developments in molecular testing for cancer were discussed by a number of presenters at the San Diego Conference. Scott described the tests now being performed by Genomic Health to help guide breast cancer treatment. The company’s Oncotype DX assay measures a panel of 21 genes to characterize the probability of recurrence of an individual’s tumor, and also assesses the likely benefit from certain types of chemotherapy. Over 7,000 Oncotype DX tests were performed in 2005, accounting for 5% to 10% of the early-stage breast cancer market in the U.S.
As shown in Table 2, an estimated 190,000 patients in the U.S. are potential targets for use of the test. The Oncotype DX assay offers a major advance compared to conventional receptor assays for determining stage and response to anti-cancer drugs. Genomic Health has found that receptor assays such as estrogen and progesterone receptor tests are unable to classify receptor expression with sufficient resolution to allow clinically meaningful differences in response to be predicted. For example, estrogen receptor expression levels in breast cancer specimens can vary by up to 1,000-fold and within the ER-positive group expression varies by 100-fold, making a simple classification of patients as ER positive or negative an imprecise way of deciding upon the course of therapy.
The Oncotype DX assay provides a quantitative score for risk of recurrence that allows treatment to be individually optimized. Scott said that with conventional receptor tests and associated treatment protocols, clinicians are now giving chemotherapy to 100 breast cancer patients to benefit only four, at a cost of several hundred thousand dollars per patient benefited.
The Genomic Health test provides information that allows physicians to change that equation, more than paying for the cost of the test and sparing many women from needless treatment that can have quite debilitating side effects. An additional feature of the Genomic Health technology is that it can be used on archival specimens up to 10 years after a biopsy was taken.
Another new molecular analysis technology with applications in cancer diagnosis was described by Jian-Bing Fan, PhD, of Illumina (San Diego, California). Illumina has developed a microarray platform that provides an integrated system for genotyping, gene expression analysis and DNA methylation analysis. The system has been used for expression profiling in formalin-fixed, paraffin-embedded samples to identify markers for prostate cancer diagnosis, and has also been used to classify cancers by methylation status.
The microarray platform is based on the company’s BeadArray technology, which employs three micron diameter beads with nucleic acid probes synthesized on their surface. The beads are then applied to the end of a fiber optic bundle or distributed on the surface of a BeadChip substrate with etched micro wells in its surface. A single analyzer can serve as the platform for performing a wide range of different tests by switching the probe sets contained on the beads. For prostate cancer diagnosis, a 16-gene expression panel has been developed that has been shown to provide significantly better prediction of relapse than Gleason Score, the standard prognostic method used clinically.
The assay is an example of Illumina’s DASL gene expression profiling assay for use with paraffin-embedded, formalin-fixed samples. Illumina has developed a comprehensive panel of 502 genes from ten publicly available cancer gene lists that can be analyzed using its microarray platform.
Illumina also has developed a DNA methylation assay that is being evaluated for applications in lung cancer. DNA methylation changes, which affect gene expression, are the earliest detectable changes in precursor lesions in colon cancer, according to Jorge Leon, PhD, chief scientific officer of Orion Genomics (St. Louis, Missouri), who discussed the topic at an AACC Southern California Section meeting following the San Diego Conference.
DNA methylation and histones that bind to DNA and alter its expression are examples of epigenetic changes that occur without changes in the primary DNA sequence. Methylation results in reduced expression of genes, whereas reductions in methylation cause genes to be turned on or expressed. Cells containing hyper- or hypomethylated DNA are shed into the circulation, and can consequently be detected in blood samples and analyzed to provide an early indication of malignant transformation. DNA methylation changes are also observed in other diseases such as cardiovascular disease, autoimmunity and psychiatric disorders. According to Leon, the number of publications in the academic literature on DNA methylation analysis has increased exponentially over the past five years, indicating a surge in research in the area.
Illumina is developing a panel of methylation markers for use in lung cancer diagnosis, and has identified a panel of 10 expression loci that can be used to reliably separate normal vs. colon, breast, lung and prostate cancer cell lines. The markers are now being evaluated in an assay for detection of lung adenocarcinoma.
Orion Genomics is evaluating methylation assays employing a 20-marker panel based on RT-PCR analysis with applications in breast, ovarian, and cervical cancer diagnosis. A second tier of products will target screening tests for lung cancer, ovarian cancer, colon cancer, and bipolar disorder. Orion has already found that very early changes in DNA methylation can be observed in smokers who go on to develop lung cancer.
Another new development in molecular testing for cancer was recently announced by ExonHit Therapeutics (Paris, France) and bioMerieux (Marcy l’Etoile, France). As discussed by Richard Einstein, PhD, vice president research for ExonHit’s U.S. business unit based in Gaithersburg, Maryland, the company is a leader in the analysis of alternative RNA splicing, another means by which the same primary gene sequence can produce different proteins depending on expression mechanisms. ExonHit has employed microarrays from Agilent Technologies (Palo Alto, California) and Affymetrix (Santa Clara, California) to develop products called SpliceArrays that can be used to analyze alternative RNA splicing.
According to Einstein, alternative RNA splicing is controlled by proteins, but at present it is not possible to predict gene expression patterns based on knowledge of the regulatory elements that are present. There is evidence, however, that alternative splicing changes are associated with certain motor disorders as well as with breast cancer and some genetic diseases. The SpliceArrays designed by ExonHit, which are based on the Agilent microarray platform, are unique in that they produce many-fold changes in signal for different RNA transcripts, whereas conventional microarrays produce only small signal changes and thus require highly precise quantitation to detect expression differences. Precise quantitation is typically very difficult with microarray technologies.
ExonHit and bioMerieux recently announced results of studies with a breast cancer detection assay that uses ExonHit’s Differential Analysis of Transcripts with Alternative Splicing (DATAS) technology that indicate the test could represent a significant advance in early detection. A 54-gene panel was employed to classify women with Stage I/II breast cancer vs. normal controls with an accuracy of 86.7%. The two companies have initiated a 1,000-patient multi-center trial as the next stage in validation of the test.
Another new development in molecular diagnostic cancer testing was recently unveiled by Quest Diagnostics (Teterboro, New Jersey).
As discussed by Feras Hantash, PhD, of the Quest Diagnostics unit in San Juan Capistrano, California, at the San Diego Conference, the company has developed a number of molecular tests with applications in genetic testing as well as cancer diagnostics. The LUMETA family of cancer assays test for both protein and nucleic acid markers in blood samples that provide diagnostic information for major types of leukemia and lymphoma, including tests for bcr-abl fusion protein expression in Chronic Myelogenous Leukemia and for ABL gene mutations that are indicative of aggressive or treatment-resistant leukemia. The LUMETA tests are now available to physicians through Quest’s reference laboratory testing service.
Advances in genetic testing
Hantash also described molecular tests developed by Quest for genetic disease testing. Quest already performs tens of thousands of molecular tests annually to screen individuals for mutations associated with cystic fibrosis. Another genetic test is used to detect gene rearrangements associated with beta-thalassemia. Quest employs reagents licensed from Tm Bioscience (Toronto) along with the multiplex assay platform from Luminex (Austin, Texas) for many of its genetic tests.
Another molecular genetic test was described by Iris Schrijver of Stanford University Medical Center. This test employs the Arrayed Primer Extension (APEX) array technology from Asper Biotech (Tartu, Estonia) to test for 15 genetic diseases that commonly occur in the Ashkenazi Jewish population, including Tay-Sachs disease, Bloom syndrome, Canavan disease, Gaucher disease, and Factor XI deficiency. The panel includes 57 genetic sequence variant markers. The APEX array can accommodate up to 6,000 different probes, but so far the applications addressed by Asper have required use of only about 200 probes.
The analysis is performed using Asper’s Generama-003 system which consists of a confocal laser scanner and a cooled CCD camera. Despite the complexity of the equipment required, Asper says the cost can be equivalent to that of existing, widely used molecular screening tests for cystic fibrosis but will allow a much larger number of markers to be analyzed at once.
Other suppliers of microarray-based systems for use in molecular genetic testing include Nanogen (San Diego), AutoGenomics (Carlsbad, California) and Roche Diagnostics (Indianapolis). Nanogen is planning to launch a new, fully automated test for cystic fibrosis that will run on its $75,000 NanoChip 400 instrument employing the company’s Analyte Specific Reagent (ASR) microarrays. The microarrays are packaged in the NanoChip 400 cartridge and allow up to 400 samples to be analyzed on the same array. The new CF test will be launched within the next few weeks.
AutoGenomics also markets a microarray-based system, the Infiniti, which now includes a number of ASR microarray tests. The company markets ASR tests for 12 genetic diseases; for infectious diseases, including a Human Papilloma Virus genotyping assay that provides results for each HPV subtype; cancer tests including assays for mutations associated with chronic and acute myeloid leukemia; pharmacogenomic tests including an ASR test for warfarin resistance; and infectious disease tests including Chlamydia, Neisseria Gonorrhoeae, and a test for drug-resistance tuberculosis. A test for assessment of Epidermal Growth Factor Receptor mutations with applications in chemotherapy management is under development.
Roche Diagnostics, the leader in molecular diagnostics worldwide with 2005 sales of $937 million, markets the AmpliChip CYP450, which is manufactured by Affymetrix, for performing pharmacogenetic testing on whole blood samples. The AmpliChip is the first microarray test to be cleared for in vitro diagnostic use by the FDA, and provides tests for variations in the CYP2D6 and CYP2C19 genes that are related to metabolism of an estimated 25% of all prescription drugs as well as some OTC drugs. Roche also is targeting development of assays that will be useful in psychiatric testing to guide drug therapy.
Assessing drug response
The expanding capability provided by molecular genetic testing to stratify patients according to their response, or lack of response, to drugs and perhaps to other types of therapy is expected to drive a fundamental change in the balance of the roles of diagnostics and therapeutics in health care. At present, laboratory diagnostics comprises only about 2% of healthcare spending in the U.S., as shown in Table 3, while spending on prescription drugs accounts for 10%. Molecular diagnostics accounts for less than 0.1% of spending. There is clearly a need, however, for diagnostics that can enable improved approaches to selecting therapy.
As discussed by Ron van Shaik, PhD, affiliated with Erasmus University Medical Center (Rotterdam, The Netherlands), who is implementing pharmacogenetic testing in oncology, current drug treatments are effective in only 25% of patients who receive cancer chemotherapy, and effectivity is only about 60% for antipsychotic and anti-depressant drugs as well as for drugs used to treat asthma. With advances in pharmacogenetic testing to improve drug targeting, along with new developments in high sensitivity molecular screening tests that allow detection of disease at an earlier stage when treatment is most effective and less expensive, molecular diagnostics is expected to play an increasingly important role in healthcare, altering the balance between diagnostics and therapeutics.
New molecular test systems for lab, POC use
In addition to providing a forum for presentation of the latest developments in clinical applications of molecular diagnostics, the San Diego Conference also is a venue for introduction of new molecular diagnostic technologies. Several new developments were described in the areas of specimen preparation, amplification, microarray technology and sensors for nucleic acid assays.
Owe Oelmueller, PhD, of Qiagen (Hilden, Germany), described recent developments with the PAXgene nucleic acid collection and stabilization system developed by PreAnalytiX, a joint venture between Qiagen and BD (Franklin Lakes, New Jersey). Specimen preparation and stabilization is an important aspect of the molecular testing process, particularly as molecular diagnostics transitions from the research setting to the clinical lab. In gene expression testing, for example, the use of the PAXgene stabilization system in determining expression levels of the bcr-abl gene results in a dramatic improvement in specimen stability over time. A 99% correlation was achieved between expression levels determined after three days of storage at room temperature vs. the levels measured at two hours after specimen collection using PAXgene, whereas for unpreserved blood specimens the correlation was 0.65.
Such differences could produce completely misleading results according to Oelmueller. The PAXgene kit was cleared for marketing in the U.S. and Europe last year. The company is now developing a PAXgene tube for stabilization of RNA in bone marrow specimens.
A nucleic acid amplification technology that may offer some advantages compared to PCR was described at the San Diego Conference by Huimin Kong, PhD, of BioHelix (Beverly, Massachusetts). The company’s Helicase-Dependent Amplification (HDA) employs a helicase enzyme to separate the nucleic acid strands during a polymerization reaction instead of heat, as is employed in pcr thermocycling. The helicase technology is highly insensitive to contaminants in the reaction mixture, and has been employed in a low-cost lateral flow assay to achieve detection of very low levels of infectious organisms. For example, 5-copy sensitivity has been achieved with prototype assays for anthrax and Group B streptococcus.
HDA amplification has also been used to develop sensitive assays for Herpes Simplex Virus, Chlamydia Trachomatis and the p53 oncogene. HDA assays are less complex than PCR-based tests because thermocycling is not required, and the capability to perform multiplex testing has been demonstrated for up to five simultaneous targets. A whole genome amplification method has been developed using HDA that can be performed in one hour at 37 C.
Another new amplification technology has been developed by Nugen Technologies (San Carlos, California), its Single Primer Isothermal Amplification (SPIA). SPIA technology is particularly well-suited for applications in gene expression analysis and for analysis of small-volume samples. The technology employs DNA/RNA hybrids as primers and uses the enzyme RNAse H to cleave the RNA portion and promote exponential amplification of a target sequence. SPIA offers an alternative to quantitative PCR for performing gene expression analysis, but is an isothermal technique. The technology is marketed as the Ovation amplification system. A new version, WT-Ovation Pico, is under development that will provide subnanogram sensitivity.
Various developments in microarray technology additionally are helping to move microarray-based testing into the mainstream of clinical diagnostics. Microarrays have proven to be a challenging technology to convert from the research setting to the clinical lab, partly due to the difficulty in obtaining standardized results using different microarray platforms and specimen preparation methods. Andrea Ferreira-Gonzalez, PhD, of Virginia Commonwealth University (Richmond) described current efforts to improve quality control and sample preparation methods for microarray testing. At least three organizations have implemented programs aimed at improving standardization for microarray analysis, geared initially to resolving issues with use of microarray data in pharmaceutical development programs.
The FDA has established the MicroArray Quality Control (MAQC) program to provide quality control tools to the microarray industry, including reference RNA samples. The External RNA Control Consortium is a working group composed of representatives from academia, commercial entities, and government scientific organizations sponsored by the Biotechnology Division of the National Institute of Standards and Technology that is also developing tools for control of molecular assays and for performance evaluation of molecular tests including microarray tests. The Clinical and Laboratory Standards Institute (Wayne, Pennsylvania) also has developed a guideline (MM-16P) for use of external quality controls in microarray analysis.
Ferreira-Gonzalez has studied the effects of sample condition on microarray test results and found that gene expression patterns obtained with microarrays such as those supplied by Affymetrix can vary substantially depending on the how the tissue is sampled. While it is obvious that a specimen consisting of 90% tumor cells will give different results than one for which the sampling technique collected no tumor cells, expression patterns also can vary substantially between samples that have 90% vs. 45% tumor cell content.
The use of amplification prior to analysis of nucleic acids from tissue specimens also affects expression patterns observed on microarrays, as does the quality and quantity of RNA present. Improvements in standardization of microarray tests, including better specimen collection and processing techniques, will be crucial for microarray technology to become widely adopted for clinical diagnostics. Issues with sampling and microarray result standardization have already had a significant negative impact on progress in using gene expression pattern analysis in cancer diagnostics. Once quality and reproducibility issues are resolved, however, microarrays promise to allow a quantum advance in molecular diagnostics for rapid, automated analysis of large numbers of genetic markers.
Microarray technology already has achieved commercial success, primarily for use in pharmaceutical research and development, as indicated by recent trends in product sales for Affymetrix, the leading supplier of microarrays, as shown in Table 4. While clinical applications represent only a small part of the total microarray market at present, that segment is expected to grow.
Development is now underway of microarray-based systems that provide complete integration of all steps of the analytical process. For example, Combimatrix (Mukilteo, Washington), a supplier of custom microarrays for research, is developing an integrated microfluidic/microarray lab-on-a-chip that allows hybridization, washing, labeling, and scanning to be performed in one self-contained device. The integrated microfluidic device takes the place of three tabletop or floor-standing systems used in most labs to perform molecular genetics analysis. It incorporates miniaturized electrochemical pumps, valves, reagent and reaction chambers, and a 12,000-feature microarray chip in a single cartridge. It also performs mixing using a microbubble generation method. Two versions have been developed, one using fluorescence labeling and a second employing electrochemical detection.
Combimatrix manufactures a benchtop synthesizer that can be used by researchers to synthesize a customized array of nucleic acid probes on the surface of the silicon substrate. Using fluorescence detection, a microarray having up to 90,000 features can be scanned, whereas with electrochemical detection an array with up to 500,000 features is feasible. The cost of a typical microarray is $99, but because the device can be reused up to four times, the cost per analysis is about $25. The microfluidic components cost less than $3.
Because all reagents are stored in the cartridge, interface issues are minimized. A thermal gradient technique has also been developed that provides a significant acceleration of the hybridization process. Combimatrix has developed prototype assays for avian influenza A, genotyping, and nucleic acid sequencing.
Piia von Lode of Abacus Diagnostica (Turku, Finland) described at the conference another integrated nucleic acid analyzer that employs time-resolved fluorometry to achieve high-sensitivity detection. The company has patents pending on multiple components of the system including sample preparation, rapid temperature cycling, and all-in-one dry reagent PCR chemistry. The analyzer uses low-cost plastic/metal foil chips as the disposable elements, and total turnaround time for a typical clinical assay is 30 minutes. Initial applications are being developed for food diagnostics. A prototype salmonella assay detects two salmonella cells in a 25 mg food sample. The use of time-resolved fluorescence enhances sensitivity by 10-100 fold compared to standard fluorescence detection. The company plans to launch the product for food diagnostic applications in early 2007.
New nucleic acid sensor technologies were discussed at the conference by Joseph Wang, PhD, director of the Center for Bioelectronics and Biosensors, Biodesign Institute, and a professor at Arizona State University (Tempe). Wang described ultra-sensitive electronic detection technologies employing nanomaterials that may eventually allow direct detection of nucleic acid targets at levels approaching those obtained using PCR amplification. Metal nanoparticles are used as labels in hybridization assays performed on low-cost screen printed electrodes for electrochemical detection via potentiometric stripping analysis. The use of different metal labels such as zinc, cadmium, bismuth, and lead allows multiple targets, such as a panel of single nucleotide polymorphisms (SNPs) in a genotyping assay, to be detected simultaneously.
Still another technique employs polystyrene beads as labels, which are filled with ferrocene particles that are released chemically after hybridization is complete, producing thousands of detectable tags per hybridization event. Using this technique, Wang has demonstrated the ability to detect 80 copies of a nucleic acid target in a direct detection electrochemical assay that does not require PCR amplification.