Personalized Medicine: A New Medical Paradigm

Recent advances in diagnostic testing have increased the likelihood that our current model of medical treatment will soon be supplanted, at least in part, by personalized medicine. With this change in the medical paradigm will come numerous benefits and opportunities for patients, caregivers, drug developers, diagnostics firms, and MCOs.

Personalized medicine, in which therapies are targeted to patient subgroups or individuals based on biomarkers, is defined by the FDA as medical practice predicated on the defined characteristics of a person or group.

Personalized medicine seeks to provide what today’s treatments, relying to some extent on trial and error, may deliver after a delay or sometimes not at all or accompanied by adverse effects. For example, persons in whom hypertension is diagnosed often will start treatment with several medications in succession before settling on one that lowers blood pressure without significant adverse effects. Similarly, a cancer patient frequently undergoes treatments with successive drug regimens until one is found that works better than the others.

It has been known for some time that medications have different effects in different people. However, the tools for differentiating patients based on such varying responses are only now becoming widely available. Before the advent of genomics, pharmacogenomics, toxicogenomics, and metabolomics, no one could predict which drugs would work in an individual patient.

The mapping of the human genome is changing the practice of medicine and, ultimately, enabling personalized treatments by providing tools for identifying genes and gene activation pathways.

Perhaps the greatest contribution of genomics is the potential to treat genotypes rather than phenotypes. Instead of treating a patient based only on his or her phenotype (consisting of outward appearance, observed psychological development, and behavior), genotyping would give physicians and patients the benefit of knowing how patients might respond to certain medical treatments based on their genes. A more individualized approach would result in medical treatment that is more precise, yielding increased efficacy and reducing the likelihood of adverse effects.

Combining Diagnostics and Therapeutics

One example of using genotypes to guide therapy is Herceptin (trastuzumab), Genentech’s breast cancer biologic drug for women with tumors that overexpress the HER2/neu gene. Herceptin’s label requires that women be tested for the gene before beginning therapy, an example of combining therapeutic and diagnostic products known as theranostics. Herceptin was barely scraping by in demonstrating efficacy in clinical trials before Bayer developed the companion HER2/neu gene test. Today, Herceptin enjoys $1.3 billion in annual sales.

Gene testing is not the only avenue to personalized treatments. Any biomarker will serve the purpose, provided it adequately defines patient subpopulations for whom the risk-benefit ratio for a particular drug is advantageous. In genetics, a biomarker is a fragment of DNA sequence that causes disease or is associated with susceptibility to disease. A biomarker can also be a substance that is introduced into a patient to examine organ function. For example, a radioactive isotope can be used to evaluate perfusion of heart muscles. In general, a biomarker can indicate the risk or progression of disease in an individual or can be used to tailor treatment.

Cell-based assays represent an emerging and increasingly useful service in support of personalized medicine. Several companies now offer to culture cells from a cancerous tumor and test them, ex vivo, against a panel of chemotherapy agents to help predict which treatments are likely to be effective for an individual patient.

Multiple Biomarker Approach

Unfortunately, the most intractable diseases affecting US adults today, such as heart disease and cancer, lack a definitive relationship with one easily identified biomarker. Even when a single biomarker works reasonably well for diagnostics, it may have poor prognostic capability. Numerous studies have shown that the differential expression of groups of genes may provide greater predictive power than that of a single gene. Typically, some of the relevant genes will be up-regulated (more active) while others will be down-regulated (less active).

By knowing the levels of gene expressions, we can understand whether a patient is producing the proteins needed to promote good health. Our bodies are made up mostly of proteins, and gene expression is the process by which our genes use the inheritable information in the DNA sequence to create messenger molecules (mRNAs) as a set of instructions to make proteins.

Gene expression through detection of mRNAs has been the rationale behind the development of diagnostic gene chip products at numerous academic and corporate research laboratories. One of the most advanced products is Iris Biotechnologies' BreastCancerChip, which we hope to introduce this year and follow with similar products for cardiovascular disease and disorders of metabolism and the CNS. BreastCancerChip identifies optimal treatment regimens based on levels of genes associated with the efficacy or toxicology of a particular agent in a specific patient. Thus, BreastCancerChip satisfies 2 of the requirements of personalized treatments: the right drug for the right patient. It remains for the health care system to deliver that therapy at the right time.

Paradigm Shift

Although promising from a patient’s perspective, the shift to a personalized medicine paradigm carries significant risks for the pharmaceutical and biotechnology companies. The pharmaceutical industry’s prevailing economic model relies heavily on blockbuster products to fund innovation, reward stockholders, and finance a product's life-cycle management.

Some experts believe that with widespread genotyping and pairing of “omics” disciplines with new drugs, personalized medicine will cause segmentation of once-huge therapeutic markets into several (or many) smaller ones and the number of patients treated with different drugs will fall dramatically, eventually sounding the death knell for the blockbuster model.

The “value” argument for personalized medicine holds that individualized treatments represent a win-win situation. Patients receive better drugs and experience fewer adverse effects, and because of this (and their ability to predict which patients will benefit) pharmaceutical manufacturers should be able to charge a premium for these products. While the number of patients being treated with each drug will undoubtedly fall, the number of drug approvals should rise dramatically as companion tests that identify targeted patient populations demonstrate the effectiveness of medications that if tested in the general population would otherwise fail to show benefit.

According to this line of thinking, high development costs are the main hurdle to the widespread adoption of personalized medicine. To some degree, the pharmaceutical industry (particularly biotechnology companies) has already begun to come to grips with market segmentation, albeit not as a result of personalized medicine.

Numerous orphan biologics are already on the market, some benefiting as few as 5000 patients worldwide. Since the original Orphan Drug Act was passed in 1982, the FDA has approved more than 250 orphan biologics and small-molecule drugs for marketing. Although the original orphan designation was limited to treatments for conditions affecting fewer than 200,000 persons, several of these have become blockbusters through the sponsors’ continued development efforts. Perhaps the most striking examples are erythropoietin (Epogen, Procrit) and infliximab (Remicade), both originally approved as orphan drugs.

Diagnostics companies have much to gain from the increased demand for high-value “omics” testing products and services to support personalized medicine. Biomarker testing is now poised to break out of its traditional business model of high-throughput, high-volume, commodity-priced tests and become a critical player in the success of new pharmaceuticals and biologics. In addition to providing postapproval companion diagnostics and testing services that should command a premium price, test developers will supply the tools by which drug sponsors choose their study subjects from phase 1 through phase 3 clinical trials. The increased emphasis on postmarketing surveillance, mandated by the FDA Amendments Act of 2007, will open the floodgates for further application of diagnostics, as well as studies to validate the predictive power of diagnostics used during drug development.

Implications for Managed Care

Insurers and PBMs will also benefit from the coming wave of personalized therapies, although not all tests are likely to be covered. While established drug-test combinations such as Herceptin and HER2/neu testing are readily reimbursed, patients typically pay for newer tests, particularly cell-based assays that are not part of a drug’s label.

The economic rationale for personalized medicine is easily demonstrated. A focus on “the right drug for the right patient” decreases the likelihood that the wrong drug will be prescribed, resulting in more effective disease management that nearly always translates to reduced costs. Fewer adverse events portend fewer or shorter hospitalizations and less need for additional pharmacotherapy.

If personalized medicine is successful, the number of drug approvals will rise dramatically, which will expand treatment choices-not with me-too medicines but with products that provide clear benefit to patients and value for health plans.

Access to Information

In addition to diagnostics, the personalized medicine paradigm also requires a complete overhaul of the health care information infrastructure with universal adoption of the electronic health record. By combining advanced data algorithms and demographic data, pharmaceutical marketers and sales aggregators can track disease categories and numbers of affected patients to the 9-digit zip code level-a small cluster of households. For now, patients’ medical records remain scattered in paper file folders in physicians’ offices and hospitals, inaccessible to others, including caregivers who must rely on the patient to provide information on comorbidities and concomitant pharmacotherapy.

Numerous studies have concluded that computerizing prescribing and patient health records could save the US health care system tens of billions of dollars per year, while enabling a higher standard of care. Citing these reports, every major 2008 presidential candidate with a comprehensive health care platform has called for modernization and computerization of patient records as part of proposed cost-saving and quality-enhancing measures.

Powerful as “omics” approaches may be for predicting a patient’s response to specific medicines, testing reveals only part of a person’s health status. The rest of the story is told by influencing factors such as environment, stress, family history, and lifestyle, which put biomarkers into context. Researchers in the field of epigenomics-the study of changes in gene silencing that occur without changes in the genes themselves-are trying to understand the bridge that connects the genome with everything else in life. Cancer and other diseases that typically do not show themselves until later in life are not caused only by inherited or somatic mutations. Also important is how the environment alters the expression of genes during a lifetime, turning them off or on permanently. It is now thought that most cancers are a mixture of genomic and epigenomic changes and that the epigenome plays the greater role in most cases.

Recognizing this, Iris Biotechnologies is currently developing a patient database product called BioWindows that will serve as a comprehensive repository of information on all factors affecting a person's health. BioWindows unifies and harmonizes genomics information from chip-based diagnostics with a patient's demographic, occupational, interpersonal, family, medical/ surgical, and lifestyle influences.

When all is said and done, the success of personalized medicine boils down to access to the right information. Which genomic factors influence effectiveness and safety? How are the genes implicated in a specific disease expressed and regulated in particular individuals or groups of patients? Which treatment options are available, and which combinations offer the greatest probability of success? What is the appropriate balance between risk and reward in treating a person who has a particular disease? And finally, do the patient’s medical history and lifestyle justify a particular medical intervention?

Within this medical information paradigm, we believe that Iris’s genome disease chips, BioWindows analytical software, and other advances that deliver the right information at the right time will usher in the golden age of personalized and targeted medicine.