Compared with white men, African American men are 50 percent more likely to develop prostate cancer, and twice as likely to die from it. They also tend to develop it at younger ages.

Like other cancer genetics researchers, Matthew Freedman figured at least part of the explanation for this disparity must lie in the DNA. After all, defective genes that increased breast-cancer risk had been isolated in women, so it was reasonable to assume the equivalent was at work in prostate cancer.

It turns out, Freedman said last week, that genetic predisposition to prostate cancer is much more complicated.

Speaking at a Wistar Institute symposium on African American men and prostate cancer, Freedman explained how he and his colleagues at Dana Farber Cancer Institute in Boston discovered a cluster of infinitesimally small genetic variations that are associated with a higher risk of prostate cancer - variations that are far more common in African American than white men.

These variations are not genetic mistakes or mutations, unlike the BRCA gene mutations that are linked to breast cancer. Rather, the variants are simply differences in the individual chemical building blocks, called nucleotides, of the genetic code - like changing the spelling of gray to grey. These single nucleotide variants, called SNPs, account for 0.1 percent of the entire genome.

While these variants are acceptable, they somehow play a role in disease predisposition, presumably in combination with environmental factors, Freedman told his audience of several dozen researchers, physicians and cancer survivors.

The seven SNPs that Freedman's group has linked to prostate-cancer risk - some published last year and three more uncovered in the last few months - are far less common in European American men than African Americans.

"This is not an African American-specific piece of DNA," Freedman said. "It's present in all of us, just in different frequencies."

The impact may be huge. Freedman estimates that almost a third of prostate cancers that develop in white Americans and 68 percent in black Americans may be related to these genetic variants.

It is not yet clear precisely which gene, or genes, carries these tiny variations, but Freedman's team has narrowed the search to nine genes on Chromosome 8. They published that finding last August, three months after an Icelandic company called deCODE genetics reached the same conclusion in separate research.

The hope is that pinpointing the risky gene will provide an avenue for better detection and treatment of prostate cancer. It might even open the door to preventive measures.

"My personal goal is to impact clinical decision-making in prostate cancer," Freedman said.

That's a tall order. There are three billion pairs of nucleotides in the genome, and the crucial variants are somewhere in a region of four million pairs.

To home in on this region, his team used a new approach to gene hunting.

Usually, finding a disease-susceptibility gene involves tracing patterns of the disease in high-risk families, then mapping the unknown DNA in relation to known genes. DeCODE, which has capitalized on Iceland's unique store of genealogical and medical data, used this "linkage analysis" approach in its search for a prostate-cancer-risk gene.

But many African American prostate-cancer patients are not from clearly high-risk families. So Freedman's lab decided to scan the genes of mixed-ancestry populations to find regions in the code where the amount of DNA inherited from African ancestors was greater than the average in the human genome.

The key study - which scanned the DNA of 2,500 African American men, including 1,600 with prostate cancer - found that those who inherited the variant nucleotides from African rather than European ancestors had a higher prostate-cancer risk.

This herculean analytical feat was aided by the 2005 publication of much of the HapMap, a database that will include all 10 million gene variants that exist in the genetic codes of people from the United States, Nigeria, China and Japan.

In effect, Freedman explained, the HapMap shows the mosaic of genetic variations that have accumulated, generation by generation, since the human race spread out from its origins in Africa 100,000 years ago.

For Freedman's audience, most of them African Americans, better prostate-cancer detection and treatment cannot come soon enough. Oliver St. Clair Franklin, for example, has had three biopsies in five years - all negative - because the standard screening test, which measures a protein in the blood, suggested he might have cancer.

Franklin's father died of prostate cancer, and his younger brother has been treated for it. The fact that African Americans can do nothing to lower their risk - and did nothing to raise it - is a source of anguish.

"You're never secure," said Franklin, 61, president of International House, a nonprofit educational and cultural facility. "Finding this gene will be a way for us to screen people more accurately."

Contact staff writer Marie McCullough at 215-854-2720 or mmccullough@phillynews.com.