Online Extra: Meeting Leukemia's Diagnostic Challenge

Tests that distinguish among the disease's many forms either cost too much or don't exist. Dr. Torsten Haferlach may have a solution

Meeting Leukemia's Diagnostic Challenge Tests that distinguish among the disease's many forms either cost too much or don't exist. Dr. Torsten Haferlach may have a solution

Crusaders for personalized medicine are discovering that the single biggest barrier to their quest lies in the mind-boggling complexity of disease itself. For the last 21 years, internationally renowned leukemia expert Dr. Torsten Haferlach has spent seven days a week, 18 hours a day, conducting cutting-edge research into the disease and treating patients.

But over the last decade, the 46-year-old German's job has grown a lot tougher. Huge advances in genetics and molecular biology have revealed that leukemia, a cancer of the blood and bone marrow, is not just one disease but four. Within these exist 50 subtypes, each with its own unique genetic signature -- and requiring a different prognosis and treatment.


  As a result, testing for leukemia is complicated, expensive, and often inaccurate. The best labs rely on as many as a dozen different, labor-intensive technologies, all of which require highly trained specialists. Even so, perhaps 50% of patients are misdiagnosed in regard to subtype.

Part of the problem stems from the steep cost of the various technologies used to diagnose leukemia today. Limited by their budgets, labs use only a handful of diagnostic methods at best.

Take chromosomal analysis. Viewed as crucial in helping doctors choose the right therapy for leukemia patients, it costs several thousand dollars per patient and calls for highly experienced technicians to interpret the results. Because of the price factor, fewer than 10% of leukemia patients in the U.S. undergo it, says Dr. Claudia Schoch, a cytogeneticist and Haferlach's research partner.


  This will soon change, asserts Haferlach, who recently started the Munich Leukemia Lab after seven years of running one of the world's largest leukemia diagnostic labs, at the University of Munich.

He's working with Switzerland's Roche Diagnostics to develop a new genetic test that promises to dramatically improve the diagnosis and treatment of leukemia by offering the first detailed picture of the specific genetic defects behind each of the subtypes.

Using a tiny silicon chip from Santa Clara (Calif.)-based Affymetrix (AFFX ) and a drop of a patient's blood or bone marrow, the chip simultaneously screens up to 400 genes implicated in leukemia and finds the exact variations that cause the different subtypes. In clinical studies, the test has proved more than 95% accurate. Roche expects to launch the test in Europe by the end of 2006.


  Although no price has been set, it will likely be much cheaper and faster than existing diagnostic methods, which cost an average of $2,400 to $4,800. Instead of waiting a week or more, patients would get their results within 48 hours. Haferlach believes this advance alone could make the difference between life or death for many patients. "You don't have time to waste in diagnosis," he says.

Getting the diagnosis right the first time also will help health-care providers and patients save money by avoiding unnecessary or inappropriate treatments. And considering bone-marrow transplants, a commonly used leukemia treatment, cost around $260,000, savings could be substantial.

The genetic test will not only provide more rapid and accurate diagnosis of the subtypes with which researchers are familiar but also eventually be used to identify new ones, says Haferlach. Armed with that knowledge, scientists can begin to develop targeted drugs designed to treat each of the various subtypes.


  The knowledge of the specific genetic mutation causing chronic myeloid leukemia, for instance, led to the development of the Novartis (NVS ) breakthrough drug Gleevec four years ago.

Such a test represents a huge advance in diagnosis, especially for patients with no clear distinction among the various subtypes. Take Haferlach's patient Hermann Michael, the 68-year-old former music director of the Phoenix Symphony. In 1999, he was diagnosed with severe aplastic anemia, a precursor to leukemia.

A few years later, his disease had evolved into a combination of three different ones, including acute myeloid leukemia, which can cause death within weeks without aggressive chemotherapy. Michael has lived nearly three years without it, taking a daily cocktail of 10 different drugs.


  "Because he seems to have three different diseases, I knew that going in with all guns blazing and giving him chemo could hurt or even kill him," Haferlach explains. "Without having the kind of detailed analysis we are able to do here, other doctors might have given him chemo just to do something."

Haferlach is confident that the gene test will enable doctors to accurately diagnose even complicated cases such as Michael's. Until then, Michael remains one of the lucky ones. "Dr. Haferlach is my personal hero," he says. "I owe him my life."

By Kerry Capell in London

Edited by Patricia O'Connell

    Before it's here, it's on the Bloomberg Terminal.