Medical Device Link.

Originally Published IVD Technology June 2002

Commentary

On to the postgenomics era

Richard S. Schifreen

Richard S. Schifreen, PhD, is the director of technology and market development for Promega Corp. (Madison, WI). He can be reached via rschifre@promega.com. This article is based in part on discussions that took place at the 2002 spring meeting of the Analytical & Life Sciences Systems Association (Alexandria, VA).

It is hard to believe how quickly things can change. Less than 18 months ago, the Human Genome Project and Celera Genomics (Rockville, MD) published sequences of the human genome in Nature and Science. For their discovery, Frances Collins, the director of the National Human Genome Research Institute (Bethesda, MD), and Craig Venter, the president of Celera then, were given the honor of meeting with former President Clinton. Everyone in the industry was looking forward to an exciting, new, and profitable postgenomics era.

Who would have suspected that the sale of sequencing instrumentation would decline precipitously, or that the former leaders of the genomics revolution would be seeking new opportunities? Venter has left Celera as the former genomics enterprise seeks to reinvent itself as a drug discovery and diagnostics company. Randy Scott, a cofounder of Incyte Genomics (Palo Alto, CA) and an early leader in genomics, recently left that company to start Genomic Health (Redwood City, CA), a firm focused on developing new approaches to personalized medicine. A number of former genomics or genomic tools companies are reinventing themselves around drug discovery, personalized medicine, proteomics, or other primary missions. It is becoming almost as difficult to find a pure-play genomics company as an Internet start-up.

Is this trend having an impact on the overall molecular diagnostics market? Molecular diagnostics generates $1.0 billion to $1.2 billion in revenues per year and is growing at a healthy average annual rate of 20–40%.¹ Approximately 80% of this market is in infectious-disease assays, with the remaining $200 million to $300 million made up of the 900 or more different assays related to genetic disease, predictive testing, cancer, and paternity testing.² While the infectious-disease segment is poised for rapid growth with the recent approval of nucleic acid tests for blood screening, the noninfectious-disease segments are also showing strong growth at greater than 20% per year, with some assays at 100% or more.

Much of this noninfectious-disease testing is concentrated in about 20 assays performed using home-brew techniques that have been a growth engine for commercial laboratories such as Quest Diagnostics (Teterboro, NJ), Laboratory Corporation of America (Burlington, NC), and Specialty Laboratories (Santa Monica, CA).

Pharmacogenetics has also emerged as a new growth opportunity and is being developed by companies such as Gentris (Morrisville, NC) and Variagenics (Cambridge, MA), along with the established commercial laboratories. In addition, Myriad Genetics (Salt Lake City) has established itself as the leader in predictive testing and has been aggressively expanding its testing menu. With such new discoveries and emerging fields, it is easy to understand how the outlook for growth in genomic testing is optimistic.

Genomic Testing Limitations

In terms of utility for diagnosing or monitoring disease, genomic assays must compete with other molecular approaches such as measuring RNA or protein expression. In some cases, genomic testing is advantageous, in that testing for genetic disease would be an ideal application for genomic assays. However, many diagnostic methods are focused on the detection of abnormal gene products, rather than on the gene sequences themselves.

For example, the use of mass spectrometry to rapidly and inexpensively detect multiple metabolites that reveal genetic-disease phenotypes is providing new capabilities for newborn screening. However, the prenatal testing field still relies primarily on biochemical assays and imaging methods for whole-chromosome analysis. Adoption of genomic testing will depend on the development of reliable methods to isolate and analyze the fetal DNA present in maternal blood, a technology that remains beyond present capabilities.

Cancer-related testing is aimed at detecting tumor-associated changes in gene expression, either due to translocations, loss of heterozygosity, or other abnormal patterns of expression. It is not clear, though, if future growth will be driven by detection of changes in gene expression through the analysis of nucleic acids, or whether protein determinations will be the methodology of choice.

Another consideration is the relative advantage of bulk methodologies versus those based on cellular imaging. The competition between in situ and immunohistochemical assays for Her2 expression (to identify candidates for Herceptin therapy) may be only the first example of the need to identify optimal methodologies based on their utility for guiding medical decisions. Meanwhile, protein determinations may gain popularity as companies such as Ciphergen Biosystems (Fremont, CA) demonstrate success with proteomic approaches for identifying putative markers for the detection of various cancers.

Clinically, genomics-based testing is mostly focused on thrombotic-risk determination, HLA typing, paternity testing, predictive testing in selected populations, and diagnosis of cystic fibrosis and other selected inherited genetic disorders. The genomic testing that is associated with preimplantation genetic diagnosis is still new and limited to a few leading in vitro fertilization centers. Paternity testing is based mostly on measuring the size of selected short tandem repeats in the genetic code, although some believe that this field will eventually adopt testing for single nucleotide polymorphisms (SNPs) as a more cost-effective approach.

Another potential application for genomic testing is demonstrated by assays that have been developed at Exact Sciences (Maynard, MA), which detect oncogenic changes in DNA collected from stool as a means of detecting colon cancer in its early curable stages. If successful, such assays could replace testing for occult blood in stool as a primary screening method.

Market Realities

The availability of suitable analytical systems has historically influenced the evolution of new diagnostic systems. In the case of molecular diagnostics, the research markets have driven the development of single-assay and multiplexed systems for DNA sequencing, SNP analysis, fluorescence in situ hybridization assays, RNA expression, and detection and quantitation of proteins in advance of their application to diagnostics. It appears that suitable analytical platforms will now be available to support whatever direction the field of molecular diagnostics may take. The development of molecular diagnostics may depart from historical trends and create a new paradigm driven by the capability of tests to deliver clinical utility. Instrumentation is still important, but in this case analytical capabilities are well ahead of the validation of useful new diagnostic markers.

For diagnostics manufacturers, genomics-based assays represent a different type of business opportunity.³ The testing volumes of such assays too small, and their reimbursement rates too low, to support the development of traditional, FDA-cleared (Class II) or approved (Class III) diagnostic kits.

In general, a market opportunity of $10 million to $50 million per year is needed for a product to generate a positive return on investment and to justify pursuing FDA approval. It is estimated that only the highest-volume genomic tests—such as Factor V Leiden and cystic fibrosis panels—even begin to approach this market potential. In turn, this market reality has resulted in the emergence of new, nontraditional diagnostic businesses at companies such as Promega (Madison, WI) and Third Wave Technologies (Madison, WI), which offer reagents labeled as Class I general-purpose reagents (GPRs) and analyte-specific reagents (ASRs) to CLIA high-complexity home-brew laboratories.

Contrast this approach with the more-traditional strategies of molecular diagnostics companies that are competing for the larger infectious-disease opportunities, such as viral-load testing, viral genotyping, and detection of sexually transmitted diseases. Purchasing GPRs and ASRs, versus approved IVD kits, has worked well for those commercial laboratories that are capable of developing and validating home-brew assays, and for those patients who benefit from earlier availability of state-of-the-art testing. Laboratories have been able to leverage higher margins for the "esoteric" molecular diagnostic testing that have driven improvements in their bottom lines and market capitalization.

Both manufacturers and laboratories have adapted remarkably well to these changes. Reliable molecular diagnostic testing is available to physicians and patients despite continued confusion regarding interpretation of the ASR regulations, uncertainty surrounding reimbursement for home-brew tests, and aggressive enforcement of technology and gene patents.

Genomic Testing Applications

As noted above, there have been some successful applications of genomic testing in clinical diagnostics.

In fact, it would be difficult to identify a molecular diagnostic test used in routine medical practice that was discovered through the Human Genome Project or competitive private sequencing efforts. By way of example, consider the products and testing being offered for genetic assessment of thrombotic risk and cystic fibrosis—two of the largest and fastest-growing opportunities in genomic testing. However, these assays were not derived from the highly publicized projects to sequence the human genome. The Factor V Leiden and prothrombin polymorphisms were characterized based on initial studies that were related to functional assays and protein biochemistry. Likewise, characterization of the mutations associated with cystic fibrosis predates the sequencing of the genome.

Use of genomic methods for HLA typing has also been successful. These assays represent new genomic tools being applied to biological systems that had been initially characterized by other means. While SNP haplotyping is believed to hold great promise, however, no diagnostically useful models have yet emerged.

Another area of genomic testing is the predictive testing that has been pioneered by Myriad Genetics (e.g., the BRCA 1 and BRCA 2 genes associated with inherited risk of breast and ovarian cancer). The tests being offered by Myriad have resulted from the company's own studies of affected populations using techniques that predated the massive human genome sequencing. Myriad has offered these tests exclusively through a reference laboratory model, and it does not offer test kits. Despite the excitement that has been generated by these assays, they only bring in revenues of approximately $25 million to $30 million per year. It is unclear to what extent predictive assays will be available for other diseases—or how physicians and patients will accept them.

Perhaps the greatest potential for genomic testing is in the area of pharmacogenetics, also called personalized medicine and, most recently, theranostics. These assays will help identify those individuals who are most likely to benefit from particular drugs, as well as those most susceptible to adverse reactions. Ideally, a patient could be given the right dosage of the best drug the first time, instead of undergoing the trial-and-error approach of modern-day medical practice.

FDA and pharmaceutical companies have already accepted pharmacogenetics as a tool for patient stratification and statistical analysis in clinical trials.⁴ Several new drug applications that would combine diagnostic tests and therapeutics are reportedly in progress, although none are currently pending before FDA. FDA has also established a precedent—in the case of Her2 testing and Herceptin therapy—for linking drug approval to pharmacogenetic testing.

The magnitude of the opportunity in pharmacogenetics will depend on successful clinical validation, and the speed with which new diagnostic assay and drug combinations can be moved through the approval process. In addition, the possibility exists that pharmacogenetic profiling could be applied to existing drugs or even to resurrect old drugs that were recalled due to adverse effects in identifiable subgroups. While pharmacogenetic testing is currently being offered by contract service laboratories to support clinical trials, it is generally accepted that the clinical validation and approvals required to move pharmacogenetics into routine diagnostic testing will take 5–10 years. The economic reward is substantial, as the profit potential would include sales of both the diagnostic and the companion drug.

Conclusion

A new wave of investment promises to drive the noninfectious-disease segment of the molecular diagnostic market beyond the potential of current genomic testing, SNP haplotyping, and pharmacogenetics. Companies such as Celera, Millennium Pharmaceuticals (Cambridge, MA), diaDexus (South San Francisco) and Roche Molecular Systems (Pleasanton, CA) are making substantial investments in developing a molecular understanding of disease.

These efforts are different in that they are combining all the available tools of genomics, proteomics, and cellular analysis. The focus is on understanding molecular mechanisms, rather than on applying a particular tool. There is also no bias as to whether the diagnostic products will be targeted toward DNA, RNA, proteins, or metabolites. Ultimately, the diagnostic targets of choice will be those that most accurately reflect the biology of disease.

Fully validated new diagnostic tests that are based on these technologies are probably at least 5–10 years away. Although microarray and labeled-particle multiplex methodologies will play an important role in discovery, the actual diagnostic products may look very much like familiar immunoassays, biochemical tests, or cytological examinations. In one sense, we have come full circle, advancing from the development of protein markers, to genomic sequencing, to gene expression, and finally back to protein identification. But tomorrow's methods will make full use of the power of modern proteomics technology. We have truly entered the postgenomics era.

References

1. DR Lewis, PR Knight, and JC Haroon, Molecular Diagnostics: The Genomic Bridge Between Old and New Medicine, (San Francisco: Thomas Weisel Partners, 2001).

2. R Park, "Genetic Testing Market on the Rise," IVD Technology 7, no. 9 (2001): 18.

3. EG Gorman, "Building a Diagnostics Market: One Molecule at a Time," IVD Technology 7, no. 7 (2001): 59–67.

4. LJ Lesko and J Woodcock, "Pharmacogenomic-Guided Drug Development: Regulatory Perspective," The Pharmacogenomics Journal 2, no. 1 (2002): 20–24.

Copyright ©2002 IVD Technology