Center News Magazine: An Interview with Boris Bastian

Saturday, January 1, 2011
Boris Bastian Boris Bastian

A combined interest in genetics and dermatology led physician-scientist Boris C. Bastian to make significant discoveries about melanoma early in his career. Appointed Chair of the Department of Pathology at Memorial Sloan Kettering in April 2010, Dr. Bastian leads pathologists and clinicians in finding new ways to integrate genetic information with traditional clinical observations to diagnose cancer and guide treatment.

I became drawn to dermatology in medical school at the University of Munich. I liked the Sherlock Holmes aspect of the discipline that allows one to make diagnoses simply through careful observation and by asking a few targeted questions. Also, skin conditions are readily studied and fairly easy to biopsy. I did my training in dermatology, allergy medicine, and dermatopathology [skin pathology] at the University of Würzburg, which has a large dermatology department with a major focus on skin cancer. After my residency, I became an attending physician and led the outpatient clinic and the dermatopathology service of the department.

German dermatology departments are different from those in the United States — they handle the entire range of oncology work in skin cancers, from diagnosis to surgery and treatment of disease that has metastasized [spread]. Young physicians get a very comprehensive understanding of the various disease stages in all their aspects, and this prepared me well for my work as a physician-scientist.

In 1995 I visited the University of California, San Francisco (UCSF), to get additional training in dermatopathology and learned about a new technology called comparative genomic hybridization (CGH), which had recently been developed by Dan Pinkel and colleagues at UCSF. The technique enables you to screen cancer cells for genetic changes. This held promise for solving some important clinical problems in dermatology, namely in distinguishing melanoma from benign lesions for cases in which overlapping features make this difficult using conventional methods. In 1997, I returned to UCSF from Germany to work for the next two years as a postdoctoral fellow in Dan’s laboratory, where I used CGH to analyze skin tumors. We subsequently identified distinct patterns of genetic changes in melanoma and benign moles that indicated that there are genetically distinguishable types of the disease.

Until then, the prevailing view had been that melanoma is a relatively homogeneous [uniform] disease. While some pathologists had noticed that melanomas could vary in their appearance depending on where on the body they arose, this wasn’t considered as important once the disease metastasized. Our genetic discoveries proved there are different melanoma subtypes and suggested that treatment and prevention strategies could be improved by tailoring approaches to the specific subtype.

As I continued studying melanomas, I was struck by how the appearance of a lesion on a patient’s skin or under the microscope was linked to the genetic alterations found in its cells. For example, we now know that a mutation in the gene BRAF is very common in melanomas originating on skin that is intermittently exposed to the sun but not sun damaged — and these same melanomas also have a distinct microscopic appearance. Targeting the effects of the BRAF mutation has proven effective in clinical trials at Memorial Sloan Kettering spearheaded by medical oncologist Paul Chapman. (Read more about Dr. Chapman’s work.)

This evidence of different disease types allowed us to form tentative categories and study them individually. For example, we found mutations in a gene called KIT in subtypes of melanoma originating on the hands, the soles of the feet, the nails, the mucosal membranes of the oral and nasal cavities, or the urogenital tract, as well as in the sun-damaged skin of older patients. As it happened, drugs that are active against the KIT protein, most notably imatinib (Gleevec®), were already being used against other types of cancers. These drugs had been tested against melanoma and proved ineffective before the heterogeneity of melanoma was widely understood.

Appreciating the association between the clinical and microscopic features of a tumor and its molecular alterations is the key to fully understanding all types of cancer.

Boris Bastian, Chair, Department of Pathology

I offered to test for the KIT mutation in any patient tumor sample sent to us at UCSF, and that was kind of the beginning of personalized medicine in melanoma. We also began collaborating with [Memorial Sloan Kettering medical oncologist] Richard Carvajal and [Chief of Memorial Sloan Kettering’s Melanoma and Sarcoma Service] Gary Schwartz in a clinical trial conducted at Memorial Sloan Kettering that confirmed the effectiveness of targeting the KIT protein in the subset of patients with the KIT mutation.

Looking back at the original, more general melanoma trials, it turned out that one patient had responded dramatically to this class of drug — someone whose melanoma had arisen on the sole of his foot, the only such case. That detail was overlooked. Nobody asked, “What’s different about this person’s melanoma?” Today, there is much more awareness about the different melanoma subtypes, the different genetic alterations of melanoma, and their relevance to therapy — similar to our understanding of other cancer types.

Appreciating the association between the clinical and microscopic features of a tumor and its molecular alterations is the key to fully understanding all types of cancer. Cancer pathology has traditionally relied on a subjective assessment of the microscopic features of a tumor. In a few years, we may be routinely analyzing a large number of genes or even a patient’s entire genome to make sense of alterations in his or her cancer. While the prospect of exploiting this genetic information for clinical benefit for the more common combinations of mutations is rapidly emerging, harnessing the entirety of the highly complex information of someone’s cancer to a level of truly personalized cancer medicine will take much longer.

The cancer pathologist of the future will need to integrate increasing amounts of genetic information into the tremendously rich knowledge on the pathology of cancer to make a diagnosis and assist in effectively guiding treatment. I plan to help promote that integration.

The prospect of using my experience in genetic and pathologic analysis of melanoma to collaborate with the clinical and scientific community at Memorial Sloan Kettering attracted me to the offer to head the Department of Pathology. I look forward to building an infrastructure that allows pathologists and disease management teams to use this approach broadly for other tumor types. I’m deeply impressed with the institution’s dedication to providing the best care for cancer patients and the ability to quickly translate laboratory findings into novel treatments. The extensive clinical experience, the large number of tumor samples in our tissue archives, the vibrant scientific community, and the commitment of leadership to promote discovery into the mechanism of disease and the development of new therapies create the synergy that makes this the premier cancer center.