Dr. Estey is an expert in treatment of acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS). He serves as an advisor to the Food and Drug Administration regarding new drugs in AML.
Many patients I have seen have relatively unfavorable prognoses when given conventional standard therapy. I believe that these patients need to be thus informed so they can decide whether to enter into a clinical trial testing a new therapy. The clinical research epitomized by such testing is frequently the best care that can be offered a patient.
Acute leukemia (particularly acute myeloid leukemia; myelodysplastic syndromes)
- Professor, Division of Hematology, University of Washington School of Medicine
- Member, Fred Hutchinson Cancer Research Center
- BA: Yale University 1968,
- MD: Johns Hopkins University 1972,
- Internship & Residency: NYU, Bellevue Medical Center
- Fellowship: M.D. Anderson Cancer Center
- Professor: Department of Leukemia, MD Anderson Cancer Center, 1991-2007
Seeking new treatments through clinical trials
Dr. Elihu Estey was a math major in college before attending medical school, which explains the direction his career has taken.
Graduating from Johns Hopkins University with a medical degree, Dr. Estey did residency training in New York’s Bellevue Hospital in neurology. “It was a real puzzle – very logical and appealing to me in that way,” Estey says. His neurology training led him to an interest in brain tumors and he went on to MD Anderson Cancer Center in Houston to work with oncology pioneer, Dr. Emil Freireich, who in the 1960s with Dr. Emil Freireich, was the first to use combination chemotherapy to cure childhood acute lymphocytic leukemia (ALL).
“Dr. Freireich was there and I studied under him. His main interest had shifted to adult acute myeloid leukemia (AML). Because of that, the size of MD Anderson, and the length of my career, I have probably seen as many adults with AML as anyone in the world,” says Estey.
Dr. Estey worked at MD Anderson for nearly 30 years. He and his family decided to relocate to the Seattle area for a change of scenery and so Estey could help build an AML program at Seattle Cancer Care Alliance.
“I would love to offer more and newer therapies to patients who are unlikely to do well with standard therapies,” Estey says. This includes the majority of patients with AML. Although most attention has been paid to the need for new clinical trials for older patients, younger patients with specific abnormalities in the chromosomes of their AML cells are often not best served by receiving standard therapy and hence are also candidates for clinical trials of new therapies.”
Following the core of his math career, Dr. Estey is very interested in clinical trial design, specifically in making trials more in keeping with the complexities of medical practice and with patients’ wishes. For example, trials typically only monitor one endpoint. Yet physicians are often concerned with several endpoints. Physicians might be willing to tolerate a slight increase in toxicity if that increase was more than counterbalanced by an increase in survival. Hence both survival and toxicity might be monitored, particularly in AML where outcomes are known relatively quickly.
As another example, while it is accepted that randomized trials are the best scientific way to find the answer to a question,” Estey says, these trials are often huge and take many years to complete. Interim analyses are often only performed after many patients have been treated, by which time, it might already be known that there is 80 percent, or even higher, probability that one treatment is better than another. Under these circumstances, most patients would prefer to receive the treatment that is doing better. However standard designs make this difficult to do. This has led to a resurgence of interest in Bayesian statistics.
Developed by Thomas Bayes (1702-1761), an English theologian and mathematician, Bayesian statistics calculate current probabilities based on current data and probabilities before a trial is started. Although not without its problems, Bayesian methodology encourages frequent reviews of the data, presumably something that patients would find desirable. Based on these reviews, randomization probabilities are adjusted to favor a treatment that is performing better.
Estey says that despite being an old method, Bayesian statistics is “up and coming” in the medical field. “It allows you to look at data more frequently and see sooner how a trial is going. Adaptive randomization means if a drug has an 80 percent probability of being better, instead of doing a one-to-one randomization to compare a new drug to an older one, you’d randomize eight patients to the currently superior treatment and only two to the currently worse treatment. This gives patients a higher chance of receiving the treatment that, on average, appears to be superior, which appeals to Dr. Estey, whose ultimate goal is to find better treatments for his AML patients.
During his off hours, Dr. Estey is a huge sports fan. His favorite teams are the Houston Texans, New Orleans Saints, and of course, the Seattle Seahawks.