Todd R. Golub of the Dana Farber Cancer Institute for his work on
The Individualization of Cancer Medicine.
The long-term goal of Todd Golub's research is the personalization of cancer treatment based on genetic principles. Currently, most cancers are diagnosed based on their microscopic appearance to an expert viewer. While this strategy is effective much of the time, the approach is imperfect, and results in mis-diagnoses in the case of molecularly distinct tumors whose appearance under the microscope is identical. Personalized medicine geared toward maximizing cures and minimizing side effects requires a move away from morphology-based cancer diagnosis and treatment planning, toward a system based on the molecular genetic features of a given patient's tumor.
Dr. Golub first approached this through the analysis of chromosomal abnormatities (translocations) present in patients with acute leukemia. He reasoned that such translocations mark the site of disruption of genes critical to leukemia development. He and his colleagues cloned a novel fusion gene that results from a t(12;21) translocation which fuses a gene discovered by Golub, named TEL, to another gene, AML 1. Dr. Golub subsequently demonstrated that the TEL/AML 1 fusion is the most common gene fusion in childhood leukemia, and furthermore, its presence is predictive of an outstanding response to chemotherapy. Eighty-one childhood acute lymphoblastic leukemia patients were studied, and all of the twenty-two TEL/AML 1 postive patients in the series were cured. This result has been confirmed by investigators throughout the world, and testing for TEL/AML 1 is now routine at all major medical centers.
The success of the TEL/AML 1 story indicated that genetic features of cancer can be extremely useful in cancer diagnosis. Unfortunately, unlike leukemias, most other types of cancer do not carry chomosomal translocations that serve to pinpoint the location of cancer-causing genes. To address this problem, Dr. Golub has recently pioneered the use of DNA microarrays ('DNA chips') for cancer diagnosis and prediction of treatment response. He and his colleagues combined the use of microarrays with novel computational algorithms to demonstrate in seventy-two leukemia patients that, for the first time, cancer diagnosis was indeed feasible (with 100 % accuracy) using the molecular genetic patterns in tumors alone. This advance moved the field one step closer to the goal of cancer diagnosis based on the intrinsic genetic properties of each patient's tumor. Dr. Golub's most recent work has taken this one step further, namely to identify genetic patterns in tumors that predict response to therapy. In particular, he has observed that the response of fifty-eight lymphoma patients to chemotherapy is predictable based on the genetic features of each patient's biopsy. Similarly, he has shown that the response of eighty-five children with the brain tumor medulloblastoma to brain radiation is also predictable. This type of approach has great promise for guiding the selection of treatment for patients not on the bahavior of the 'average patient', but rather on the predicted behavior of each individual patient. It is anticipated that this approach will result in patients being more informed about their disease, and thus better able to choose a course of action that is suitable.