Associate Member, (Genetics)
Department of Cancer Biology
Roswell Park Cancer Institute
Elm and Carlton Streets
Buffalo, NY 14263
Telephone: (716) 845-4529
Fax: (716) 845-8126
E-mail: Garth.Anderson@RoswellPark.org

General Research Interest

Genomic instability in tumor progression and malignancy

Current Program

• Quantitating genomic instabilities in human sporadic colorectal cancer progression
• Characterization of individual instability events, to clarify mechanisms involved
• Locating the genes which give rise to genomic instability in cancer
• Clinically exploiting genomic instability in colorectal, breast and thyroid cancer

Laboratory Personnel

Huferesh K. Darbary - Research Affiliate
Daniel Stoler PhD - Experimental Pathology RPCI

Description of Research

My research program is targeted at understanding the fundamental nature of cancer: what really is this disease? I have come to view cancer as a disease of genomic destabilization, arising out of the loss of genes designed to preserve genomic integrity, followed by many years of evolution and natural selection. Ultimately this leads to a highly heterogeneous population of proliferating, invading and continuously evolving cells, which we see as the malignant tumor. 

We have found that colorectal cancer cell genomes are horrendously damaged, with an average of about ten thousand events per cell.  Events in one area of a tumor generally can differ extensively from those in others. Early stage premalignant adenomas already contain over five thousand events, indicating genomic instability begins very early in tumor progression.

Why are we studying human cancers instead of yeast or mouse models of genomic instability? Tumor progression in any species requires a relatively narrow, unique range of instability in order to generate an evolutionary rate able to produce the number of critical mutations required to generate malignancy, without overwhelming the ability of proliferation and natural selection to generate a thriving, aggressive tumor cell population. For colorectal cancer, this process in humans typically requires ten to twenty years to complete. In mouse models, in contrast, tumor progression must be completed within its roughly two year life span, mandating much higher evolution rates. The short life span of the mouse further means the mouse demands far less stringent mechanisms for preserving its genomic integrity than does man, and this is even more the case for yeast. It is these integrity-preserving mechanisms which must be affected to enable the evolutionary process of tumor progression to occur. And obviously, yeast never develop cancer.

We developed the technique of inter-(simple sequence repeat) PCR as a simple, rapid and robust method of sampling the tumor genome to quantitate intrachromosomal genomic instability. We are also using microarray comparative genome hybridization, genome-wide loss of heterozygosity, and microsatellite instability to measure genomic instability in tissue DNAs. Each method preferentially detects different types of instability, and the instability detected by one method does not correlate with that detected by another.  At least four forms of intrachromosomal instability can exist in cancer, and all with different rates for each form. Add aneuploidy with whole chromosome losses or gains, let it simmer with years of evolutionary divergence, and cancer becomes a formidable problem. We all already knew it, but hadn't truly understood why.

On the clinical side, different evolutionary mechanisms (or combinations of mechanisms) generate different clinical outcomes. Ongoing studies indicate some forms of genomic instability lead to a better prognosis, while others yield a worse prognosis. This is of limited value in the treatment of colorectal cancer, where even adenomatous polyps are surgically resected once they are identified. But for other cancer types such as breast, thyroid and prostate where early stage disease is abundant but few aggressive malignancies result, the ability to identify the more rapidly evolving set which need aggressive treatment is of obvious value. Daniel Stoler and I are currently pursuing these clinical applications of the measurement of genomic instability, in collaboration with Thom Loree, MD (thyroid) and Nicholas Petrelli, MD (colorectal).

Publications

Click here for list of current publications