I have a strong childhood memory of seeing a little chart listing the survival rates for different types of cancer. Pancreatic cancer occupied the dismal spot at the bottom of the list, with just a 5 percent survival rate. A decade later when my uncle was diagnosed with the disease, I remembered that chart and instantly knew his chances weren’t good. I was away in graduate school and didn’t get a chance to see him before he died.
Sadly, nearly two decades after my uncle died of pancreatic cancer, the treatment options for the disease haven’t improved much, and the death rate remains tragically high: Each year in the US alone, the disease claims more than 30,000 victims. But recently pancreatic cancer research has been in the news, with what appear at first blush to be promising findings. Are we closer to finding a cure?
Sally Church, a pharmaceutical marketing consultant who keeps a watchful eye on new research, was intrigued by the results of a study released August 2. A vast team of researchers examined the genomes of thousands of people with pancreatic cancer and found that, compared to a population without the disease, many of these cancer patients shared specific genetic variations on one of their chromosomes. These variations were also associated with blood types A, B, and AB—and not the most common blood type, O.
Despite headlines proclaiming that “blood type may boost pancreatic cancer risk,” the blood-type/cancer-risk link, which has been known for decades, isn’t what’s new about these findings. Rather, by identifying the specific genetic variants that are associated with the disease, these findings may bring us a step closer to understanding the mechanism by which the cancer develops.
But as Church points out, the research still has a long way to go. We still don’t know why people with O blood types have a lower risk of pancreatic cancer. We don’t know how to put our knowledge of these differences to work. The pathway to a cure or preventative treatment remains unclear.
Church could be right to be skeptical. While we can question her motives as an industry consultant, the press releases issued by cancer institutes are also subject to bias: After all, they’re interested in demonstrating the value of their work as they seek additional funding.
So just how hard will it be for genome researchers to make headway against pancreatic cancer? Last March, Daniel MacArthur, a researcher of genetic variants that lead to disease, wrote an excellent post summing up the some of the difficulties of using DNA sequencing to identify disease risk. In a study led by Siân Jones, researchers discovered a mutation responsible for pancreatic cancer in a single patient, then confirmed the mutation’s presence in 3 of 96 other patients with pancreatic cancer. The mutation was entirely absent from more than 1,000 people without the disease. So, in principle, anyone with this mutation is likely to develop cancer and should be closely monitored for the disease, right? And if we could find other similar mutations, we might actually be able to eradicate it…right?
So the story goes. Unfortunately, as MacArthur said, the investigators also found “a whole stack of red herrings.” There were literally hundreds of abnormalities in the initial patient’s genome, and only through a quirk of the cancer (the fact that the normal, non-mutated gene is disrupted in cancer cells but not healthy ones) were they able to find the mutation responsible for the disease.
The problem is two-fold. Genomes aren’t orderly and neat; they’re exceedingly messy and complex, filled with “noise” from which subtle signals are difficult to filter. A disease can arise from one or two mutations, or from the cumulative action of hundreds. This means finding genome mutations responsible for diseases is both incredibly difficult and also often fruitless: The variation in individual genomes is so large that nearly every single potential disease-causing mutation typically turns out to be benign. And, of course, while we know most people who get pancreatic cancer have A, B, or AB blood types, the vast majority of people with these blood types don’t get pancreatic cancer.
This isn’t to say that you should ignore the next cancer “breakthrough” you see reported in the media—real strides have been made in fighting diseases like breast cancer and leukemia. But it’s never a bad idea to take a step back from many media reports to see what the experts in the blogosphere have to say about it.
Originally published August 19, 2009