The Inner Frontier: Exploring the Brave New World of Personalized Medicine

Oncology Live®, August 2007, Volume 8, Issue 8

Currently available pharmaceutical therapies for many chronic illnesses leave much to be desired.

Currently available pharmaceutical therapies for many chronic illnesses leave much to be desired. Although considerable progress in therapeutics has been made in recent years, available treatments usually do not provide a definitive cure. The goal of current therapy is usually more modest—symptom relief or delay in disease progression.

Many of today’s drugs are also often quite expensive and often must be taken over a long period of time. Therapy with currently available medications is often hit-or-miss, because we cannot accurately predict a patient’s response to a drug or if side effects will occur. Thus, selection of a therapeutic agent for an individual patient is usually marked by a frequently time-consuming and expensive process of trial and error. Often, the first drug chosen is not effective or is not tolerated due to side effects.

All medications approved for marketing by the FDA have been shown, through well-controlled randomized studies, to be statistically effective for large groups of test subjects. However, these drugs are often not effective for some individuals in these groups. Absent or incomplete efficacy in individual patients is very common among major drug classes. The cost of this failure to effectively treat is very high.

Into the Future

Due to recent advances in several fields of biomedical research, we will soon be able to identify those patients for whom a drug will be effective. We are on the threshold of major breakthroughs in individualized patient care as a result of new and anticipated advances in molecular biology, genomics, pharmacogenetics, biomarkers, and other biomedical disciplines.

Key to these changes is our growing understanding of the genetic factors that underlie disease and patients’ responses to medications. Pharmacogenomics blends pharmacology with genomic science to study how drug response is influenced by inherited variations in genes and the proteins they produce. Pharmacogenetics is expected to have a profound effect on the individualized treatment of complex, multigene disorders. Recent advances in genomic research have enabled new strategies to determine how individual genetic differences may be responsible for variability in response to drug treatment.

Ongoing studies are finding links between a patient’s specific genotype and the likelihood that a specific agent will be effective or that an adverse effect will occur. Subtle variations in genes can affect expression of specific protein targets, such as receptors, transporters, and cell signaling pathways, all of which play an important role in response to medications. Future therapies will often be based on the identification of which molecular pathway is disrupted in a given patient. This emphasis on diagnostics represents a sea change in the focus of the healthcare industry from a clinical definition of disease and diagnosis to a molecular definition of diagnosis and predisposition.

A major limitation of current therapeutics is the relatively small number of known “drugable targets” at the molecular level. It is anticipated that advances in genomics will greatly expand the amount of targets. The number of targets “hit” by all drug substances marketed today is only about 200, but the number of potential targets for small molecule drugs is estimated to be about 3,000. Therapies that impact biological activity at these targets will greatly expand our therapeutic armamentarium with agents that are selective for the disease process, with limited side effects.

These advances will usher in a new era of effective and truly personalized treatment for major diseases. The considerable “noise” that now exists, in the form of non-response to treatment and adverse drug effects, will be reduced as therapies are channeled to patients for whom they will be safest and most effective, improving

quality and quantity of life.

Individual Drug Effectiveness

Studies of twins and blood relatives have shown that genetic factors are important determinants of the normal variability of drug effects. For example, genetic variations appear to underlie differences in the effectiveness of the beta-agonist albuterol used in the treatment of asthma. Different haplotypes (groups of genetic variations) are associated with patients’ varying responses to the medication.

Haplotype 4 is associated with depressed responsiveness and haplotype 2 with increased responsiveness to this agent. Similarly, there is considerable variation in clinical response to antipsychotic agents. For example, the response rate for clozapine varies between 30% and 60%, a diff erence believed to reflect variation in genes governing neurotransmitter-receptor-drug relations. An optimal combination of six specific gene variations (across multiple genes) has been associated with a maximal (77%) success rate with clozapine. Other antipsychotic agents may be relatively more effective in patients with other gene-variation combinations.

At the drug-receptor level, variations in genes controlling dopamine receptors have been associated with patterns of response to antipsychotics. Some variants are associated with a better response to typical antipsychotics, others with a better response to atypical agents. Several gene variants controlling serotonin receptors may influence the effectiveness of antipsychotic agents in schizophrenia’s negative symptoms (blunted affect, social withdrawal); other variants have been linked to drug effectiveness in positive symptoms (delusions, hallucinations). A dopaminereceptor variant has been implicated in susceptibility to tardive dyskenesia.Weight gain induced by atypical antipsychotics is associated with genetic factors, but the genes involved can differ among agents.

In addition, some gene variants are associated with less weight gain in response to a given agent, even with those

atypicals in which weight gain is often pronounced.

Cancer: A Personalized Approach

Examples abound of known genetic factors that influence cancer drug response. If detected, they can actually be used to predetermine how well a drug will work for an individual patient.

Breast Cancer

• The Oncotype DX 21-gene assay quantifi es the expression of 21 genes linked to the likelihood of breast cancer recurrence and the effectiveness of certain types of chemo- and hormonal therapy.

• A multigene assay can predict the recurrence of breast cancer in patients treated with tamoxifen. Gene expression analysis can be combined with an algorithm for calculating risk for distant recurrence of breast cancer.

• Polymorphisms in the HER2 receptor gene have been found to determine whether a patient with metastatic breast cancer will respond to treatment with Herceptin (trastuzumab).

• A test for the BRCA 1, 2 genes guides surveillance and preventive treatment based on susceptibility risk for breast and ovarian cancer.

Blood Cancers

• The BCR-ABL and c-KIT detect genetic mutations in the tyrosine kinase receptor that are correlated with the efficacy of Gleevec (imatinib mesylate) in chronic myelogenous leukemia (and also in gastrointestinal stromal tumors).

• The thiopurine S-methyltransferase (TPMT) assay guides dosage adjustment of Purinethol (mercaptopurine) in treatment of acute lymphoblastic leukemia. Patients with inherited little or no TPMT activity are at increased risk for severe toxicity from conventional doses of this drug.

• Genetic analysis can be used to assess clinical response to therapy for follicular lymphoma with the monoclonal antibody Bexxar (tositumomab). Analysis of a genetic rearrangement in the BCL2 gene demonstrated a molecular association with clinical response and disease progression.

Melanoma

The p16/CDKN2A genetic test guides surveillance and preventive treatment based on susceptibility risk for this cancer.

Lung Cancer

Iressa (gefitinib) is a genome-based drug for lung cancer that blocks the epidermal growth factor (EGF) receptor kinase. This agent almost failed FDA approval because of low response rate in clinical trials. But some subjects did respond dramatically. In particular, persons in Japan have a higher response rate than US patients. A mutation has been found in a gene encoding the receptor for EGF that is associated with drug response. Japanese people are more likely to carry this mutation than US patients and are three times more likely to respond to treatment with Iressa.

Colon Cancer

The UGT1A1 marker detects variations in the UGT1A1 gene in colon cancer that influence the ability to break down Camptosar® (irinotecan), which can lead to increased blood levels of the drug and a higher risk of side effects.

Target: Adverse Drug Reactions

An estimated two million people develop serious side eff ects from adverse drug reactions (ADRs), and more than 100,000 hospitalized individuals die from ADRs each year in the US,13 resulting in enormous costs to society—the cost of drug-related morbidity and mortality in 2000 exceeded $177 billion.

ADRs often result from variations in the genes that regulate the enzymes involved in drug metabolism. Variations in these genes can result in slower or faster drug metabolism. Gene mutations can result in the production of an enzyme with reduced function, or no enzyme may be produced at all. Depending on whether a person has zero, one, or two normal copies of a gene, he or she may be a poor, intermediate, or rapid metabolizer. The fewer normal copies a person has, the poorer the metabolism and hence higher blood levels of the drug. This, in turn, may result in greater effectiveness or, more likely, greater side effects.

Among 27 drugs frequently cited in adverse drug reaction studies, 59% are metabolized by at least one enzyme with a variant genetic allele known to cause poor metabolism. Identifying persons having such alleles prior to drug administration may result in a clinically important reduction in adverse outcomes. Genetic tests are available to detect variants in the cytochrome P450 series genes that commonly regulate drug metabolism. Testing for genetic variations and metabolic status prior to drug treatment would allow accurate selection of dosages, thereby reducing costly drug interactions and toxic responses.

The Big Question: Will Insurers Pay?

It is likely that payers will apply the principles of evidence-based medicine and pharmacoeconomic analysis to their coverage decisions regarding the new genetic technologies, just as they do for other medical services. Covered services must receive FDA approval and be supported by peer-reviewed evidence on performance and health outcomes. It would be expected that these new technologies demonstrate improved net health outcome, be at least as beneficial as established alternatives, and be cost effective.

Several studies indicate that molecular diagnostic testing might make healthcare more cost effective. A modeling study using insurance claims for asthma patients to determine the economic impact of a hypothetical diagnostic test used to predict drug response found that such a test would offset costs, provided that the test was sensitive and not overly expensive. An examination of genetic testing for TPMT deficiency, which can lead to severe adverse reactions to purine drugs, has confirmed the cost-effectiveness of such testing under certain conditions.

Recently, a model was published detailing the impact of applied genomics testing in cancer on cost savings in specific disease situations. The best-known example of genetic-diagnostic success is Herceptin, coupled with HER2/neu testing for breast cancer. Adoption has been successful despite a small market and concerns about value, because patients and providers were early demanders of this drug, which treats a life-threatening disease with few options and a poor prognosis. Despite the high costs of treatment, payers are covering the drug.

Testing CYP2C9 activity to prevent bleeding due to high blood levels of warfarin (see call-out) is predicted to prevent one major bleed for every 44 patients screened. At $135/assay, this would cost $6,000 (a neutral economic result), but patient outcome and quality of life would improve. A prospective trial at Harvard University is ongoing to test the clinical and economic outcomes of such testing in the real world.

Clinicians Must Keep Pace

The medical community is currently not prepared to meet the challenge of these new technologies and to utilize them in routine clinical practice. There are only 900 board-certified medical geneticists and only 2,000 board-certified genetic counselors in the US.

Seventy-two percent of physicians rate their knowledge of genetics as fair to poor; and 69% of physicians have had a genetic course in medical school but have not applied genetics to their practice. The genetic content in nursing school curricula actually declined from 10.5 hours in 1984 to just six hours in 1996. Clearly, physicians and allied healthcare professionals will have to work hard to stay abreast of developments in genetics-based medical technologies. The effort required will be considerable, but well worth it given the vast potential benefits in the form of improved outcomes, reduced side effects and adverse drug events, and cost savings.

Richard Levy, PhD, is a Senior Research Advisor for the National Pharmaceutical Council.

Personalized Warfarin Dosing: Predicting Dose Requirements and Preventing Toxicity

The oral anticoagulant warfarin is a top-20 drug in the United States, and the most widely used worldwide for the oral treatment and prevention of thromboembolic events, including atrial fibrillation and deep-vein thrombosis. Warfarin has a narrow therapeutic index and high risk of overdose or toxicity, and is among the top three drugs causing adverse drug events leading to emergency visits in older individuals.

Variability in response to warfarin is high; the daily therapeutic dose within a patient population can vary by more than an order of magnitude. Patients are all titrated to the same pharmacological endpoint—a two-to-threefold increase in their clotting time—but stable dosing can take many months to achieve and put patients at risk of excessive bleeding (if the initiating doses are too high) or ineffective clot control (if the doses are too low). It is important to quickly find the correct dose for an individual patient since the therapeutic window for this drug is small.

Non-genetic factors such as age, gender, body mass index, and drug history contribute up to 30% to warfarin dose variability, and these factors are generally taken into account when prescribing initial doses of the drug. Much of the remaining variability is associated with differences in cytochrome P450 metabolism (specifically CYP2C9) and to variation patterns in the VKORC1 gene that codes for an enzyme-controlling response to this agent.17 By combining patient clinical characteristics with knowledge of the CYP2C9 and VKORC1 genotypes, as much as 60% of the variability in warfarin dose requirements can now be explained