Clinical Research Division investigators and colleagues have reported the first use of a sensitive molecular test to measure precisely the extent of remission in cancer patients undergoing drug therapy and to predict their likelihood of relapse.
In a study published in the Oct. 9 issue of the New England Journal of Medicine, Dr. Jerry Radich's laboratory and collaborators used a molecular test to detect residual cancer cells in leukemia patients enrolled in a clinical trial to compare two drug regimens. The researchers found their method-which is based on a technique called polymerase chain reaction, or PCR-to be far more stringent at picking up traces of cancer that could lead to relapse than the standard test known as cytogenetic analysis, in which chromosomes are examined under a microscope.
The researchers predict that the approach will speed the delivery of promising new therapies to patients by drastically reducing the duration-and cost-of clinical trials, the process used to test and compare cancer treatments.
The study was part of a large international clinical trial to compare the response of patients with chronic myeloid leukemia to Gleevec, a recently approved drug manufactured by Novartis, with a combination therapy consisting of interferon and cytarabine.
Consistent with previously published results, the new study found that significantly more patients entered remission as judged by cytogenetic analysis when treated with Gleevec than with a combination of interferon and cytarabine. But using PCR analysis, researchers could for the first time detect differences in what Radich described as the "depth of remission," depending on which drug therapy patients received. Those differences foretold the likelihood that a patient's disease would worsen within two years of achieving remission.
Radich said that the study, which is the first to use the technique to assess the extent of response to drug therapy, opens the door to a new conceptual approach for the design of future clinical trials.
"Our results tell us that PCR can be used to stratify patients according to their level of remission by measuring the number of cancer cells that remain after patients are in cytogenetic remission, and those numbers correlate with how likely it is that a patient's disease will progress," he said. "We found that patients treated with Gleevec experienced a deeper remission that was associated with a better prognosis than patients treated with interferon plus cyarabine."
In addition, he said, the use of molecular diagnosis of remission as a surrogate endpoint could drastically reduce the duration and cost of conducting large-scale clinical trials.
"Rather than waiting five or 10 years to know how well a new therapy works, we can have answers in a year, which will make the best new therapies available faster and at less cost."
Chronic myeloid leukemia, or CML, is a rare cancer of the blood that affects about 4,300 people each year. CML initially presents as a relatively indolent condition, called the "chronic phase," during which there is uncontrolled expansion of myeloid cells (white blood cell precursors). Over time, the chronic phase can develop into a "blast crisis," an aggressive phase that is characterized by excessive production of blasts, which are abnormal immature blood cells. Blast crisis is commonly fatal, as patients suffer from bleeding and infections.
Currently, bone-marrow or stem-cell transplantation is the only known cure for the disease. Patients who elect not to pursue transplantation, have risk factors that make them ineligible or who do not have tissue-compatible donors are treated with drug therapy.
Unlike all other cancers, CML owes its origin to a single genetic defect: the breakage and swapping of portions of chromosomes 9 and 22, a rearrangement that produces a truncated chromosome 22 known as the Philadelphia chromosome. A single cell with this cancer-causing defect can divide to form millions of copies, or clones.
The bcr-abl gene
The rearrangement results in the formation of a new gene, bcr-abl, that consists of DNA from each intact chromosome - bcr on chromosome 22 and abl on chromosome 9 - juxtaposed at the break-point fusion on the Philadelphia chromosome. The rearrangement can be detected by microscopic examination of blood cells in a procedure known as cytogenetic analysis, which is used to evaluate the extent of the disease and response to treatment. The PCR test, which Radich pioneered in earlier studies to evaluate relapse in stem-cell transplant recipients, essentially counts the number of copies of the bcr-abl gene and can pick up very small numbers of cancer cells that would not be seen by cytogenetic analysis.
The new study involved 1,106 patients between the ages of 18 and 70 with chronic-phase CML who were randomly assigned to a group that received Gleevec or interferon plus cytababine. After a median 19 months of follow-up, 408 patients in the Gleevec group and 47 in the interferon/cytarabine group had a complete remission as judged by cytogenetic analysis (no detectable Philadelphia chromosome by microscopic analysis). Of these patients, 333 patients in the Gleevec group and 37 in the interferon/cytarabine group had samples available for PCR testing.
The researchers measured the reduction in the number of bcr-abl gene copies, as determined by PCR analysis of the amount of RNA produced by the gene, at the time of complete cytogenetic remis-sion as well as at time points up to 18 months later.
At the time of cytogenetic remission, the median reduction in the levels of bcr-abl was 250-fold in the Gleevec group compared to 220-fold in the interferon/cytarabine group. Thirty-two percent of the Gleevec-group patients in cytogenetic remission had a reduction of at least 300-fold, compared to none of the interferon/cytarabine patients in cytogenetic remission. Bcr-abl levels continued to drop during follow-up, with a median reduction of 370-fold in the Gleevec group and 250-fold in the interferon/cytarabine group 15 months after achieving complete cytogenetic remission.
For patients with a reduction in bcr-abl levels of at least 300-fold, the probability of remaining in remission was 100 percent at 24 months, as compared with 95 percent for patients who had less than a 300-fold reduction and 85 percent for patients who did not have a complete cytogenetic remission. Patients will continue to be monitored to determine whether their molecular response to treatment as judged by the PCR test continues to correlate with remission and survival.
Despite the strong response to Gleevec seen in a significant fraction of patients, only 3 percent of those who had achieved complete cytogenetic remission had levels of bcr-abl that dropped below detectable levels, which is equivalent to about a 450-fold reduction.
"The fact that bcr-abl levels rarely drop to undetectable levels with Gleevec is worrisome," Radich said. "This means that that the cancer is not completely going away, and there may be potential for the disease to recur." Relapses after long-term therapy with Gleevec have been increasingly described.
In contrast, he said, most patients who achieve complete cytogenetic remission after bone-marrow or stem-cell transplantation have undetectable levels of bcr-abl activity and are considered cured of their disease.
Trials and collaboration
Radich and colleagues are now initiating clinical trials to further optimize drug therapy for CML using bcr-abl levels as the primary measure of drug effectiveness. The trials will be administered through the Southwest Oncology Group, a national clinical trials cooperative group whose statistical center is housed partly within Fred Hutchinson.
Collaborators on the study include lead author Dr. Tim Hughes at the Institute of Medical and Veterinary Science in Adelaide, Australia, and colleagues at Hammersmith Hospital in London; the University of Heidelberg; and Novartis Pharma of Basel, which funded the study.
