Personalized Medicine Program Gets Green Light

Full access

Suppose you’re seeing a new patient with kidney disease, high blood pressure, and high cholesterol. What if you could order a single lab test that would assess all known gene variants that might affect his response to common drugs—not just medications he’s currently taking, but also common drugs that may be prescribed in the future? That’s the approach being studied by The University of Chicago’s Center for Personalized Therapeutics and other centers nationwide.

The goal is to develop a “medical system model” to overcome barriers to personalized medicine, incorporating patient-specific pharmacogenomic results into everyday patient consultations. The model is being tested now in the Center’s ongoing “1200 Patients Project” ( no NCT01280825). Peter H. O’Donnell, MD, is principal investigator.

The 1200 Patients Project is designed to perform “broad, preemptive pharmacogenomic testing” for a large number of germline polymorphisms with known effects on drug responsiveness or toxicity. The test, performed in a Clinical Laboratory Improvement Amendments (CLIA) setting, reports on genes affecting widely used medications including aspirin, hydrochlorothiazide, various classes of blood pressure-lowering drugs, statins, and warfarin. The cost of testing is less than $500 per patient—about the same as for most individual CLIA genotype tests.

An important part of the model is delivery of the results to physicians via an interactive Web portal, or “genomic prescribing system” (GPS). The GPS presents results in a color-coded traffic light system: green means a favorable result, yellow means caution, and red means high risk. Physicians can also access further information and a patient-specific interpretation of the test results, for a “virtual pharmacogenomic consult.”

The 1200 Patients Project has been launched to evaluate the feasibility and utility of incorporating a preemptive pharmacogenomic testing approach into routine medical care. Eligible patients were receiving routine care or treatment for conditions such as heart disease, inflammatory bowel disease, autoimmune disease, or others. All were regularly taking at least one, but no more than six, prescription drugs, with a life expectancy of at least three years.

Last year, O’Donnell and colleagues published results from the first year of the 1200 Patients Project in a special issue of American Journal of Medical Genetics. At that time, 812 patients had participated and 608 had been successfully genotyped. Of 268 clinic encounters at which genotyping results were available, participating study physicians accessed the GPS in 230 visits.

A total of 367 result signals were delivered via the GPS. Green lights accounted for 57 percent of results, yellow lights for 41 percent, and red lights for 1.4 percent. In 100 percent of the high-risk red light alerts, physicians clicked through to access the clinical details. They also clicked on 72 percent of yellow lights, as well as 20 percent of red lights.

The information delivered via the GPS was routinely used in consultations, and patient interest in being tested was “nearly ubiquitous.” O’Donnell and coauthors write, “We demonstrated that delivered pharmacogenomic alerts had widespread applicability to our patient population and to the drugs they are routinely prescribed.” At the time of the report, the investigators were accruing about 30 patients per month, with increased participation expected over time.

So far, the results demonstrate the successful implementation of preemptive pharmacogenomic testing in a program that is appreciated by physicians and patients and routinely used in clinical care. Of course, the ultimate goal will be to determine how preemptive testing and pharmacogenomic decision support will affect key clinical outcomes—including high-risk prescriptions, adverse events, and nonreponse to prescription drug therapy.