Using groupings or clusters of a patient's gene expression to compare to a diseased "test" set that identifies the cause of heart failure, the Hopkins team assembled a 90-gene profile to determine which type of heart failure had most likely developed. Results showed the test profile to be highly accurate, with 90 percent specificity.

The findings could, if affirmed and adapted to a standardized and affordable test format, someday aid physicians in the diagnosis of heart failure and help determine which kind of therapy is best to use for the condition. In ischemic heart disease, the patient's arteries have narrowed and the heart cannot pump normally because blood flow (and thus oxygen) is often restricted to the heart muscle. In nonischemic forms of the disease, the heart cannot pump normally because the heart muscle has often enlarged for other reasons, such as physical deformity or alcohol abuse. Both conditions can lead to cardiac arrest or more gradual heart failure as the muscle weakens over time.

"The gene expression differences between various forms of cardiovascular disease are poorly understood, despite the fact that we know there are major differences in what is happening at the cellular level," said Michelle Kittleson, M.D., cardiology fellow at the Johns Hopkins Heart Institute and lead author of the study to be presented at the American Heart Association's Scientific Sessions 2004 on Nov. 6, as a finalist for the Samuel A. Levine Young Clinical Investigator Award.

"Our study shows that gene expression profiling for heart failure patients is not only possible, but accurate as well. Based on these initial findings, we hope to close the gaps in our understanding of the gene expression patterns underlying heart failure and treatments for the illness. Ultimately, we hope to be able to use genetic profiling to classify patients according to their risk of developing all kinds of heart disease."

To create a gene expression profile, or test, the Hopkins team collected 16 biopsy tissue samples, six from patients with the ischemic form of the disease and 10 from nonischemic cases, all with end-stage heart failure. Most of the test samples came from heart transplant patients at Hopkins in the last 20 years.

Using a biostatistical technique called prediction analysis, the researchers identified the 90 genes that best distinguished the two kinds of heart failure. The large number of genes used also improved accuracy of the test.

This gene profile was later validated by testing it against 38 other tissue samples, including 14 provided from the University of Minnesota. These test samples involved tissue from all stages of heart failure, including end-stage, post-LVAD (a type of heart surgery) and biopsy samples from newly diagnosed patients.

"Now that we know we can genetically profile heart patients according to ischemic and nonischemic heart disease, our next step is to develop a test that can be used in a clinical setting," said senior study author and cardiologist Joshua Hare, M.D., a professor of medicine at the Heart Institute. "Ischemic patients need to be monitored more closely in case they develop drug resistance and require surgery to unblock clogged arteries. Knowing which patients to treat and how closely to monitor them could significantly improve how well physicians manage the disease and, consequently, improve health outcomes."

jhmi/