Three years ago, pathologist Ashley Hill flew to Minnesota to meet a family wracked by one of the world's rarest cancers. 

Called pleuropulmonary blastoma, or PPB, this aggressive lung cancer affects just 25 to 30 children a year in the U.S., and most doctors have never heard of it. With such a small patient population, scientists told Hill that tracking down the genetic cause of the disorder—and getting the funding to study it—was a long shot.

Hill persevered because the cancer followed a stereotypical progression during early development, suggesting it was caused by one faulty genetic instruction. Still, she had her doubts. "Would there be enough families to participate?" she wondered, "If there were, would it be a single-gene disorder?"

But there she was in the basement of the Holy Rosary Catholic Church in Mankato, Minn., with 25-year-old Trisha (Schoettler) Anderson and nearly 70 members of her family. Anderson had the lower lobe of her left lung removed at age two before undergoing two years of intensive chemotherapy that nearly killed her. The smell of rubbing alcohol is enough to make her feel queasy again. Anderson's brother Adam died when PPB spread to his brain at age four, Peter had lung cysts at 16, David got them at 24, and her sister Katie developed a benign thyroid cancer as a teenager. "Cancer and tumors are part of the family," Anderson says. She also has a cousin with a rare kidney tumor and an uncle with a cancer that develops from muscles cells, both of which occur in families with PPB.

Many common diseases appear to stem from multiple genes and their interaction with the environment, yet the process remains too complex to be easily untangled with today's gene-sleuthing tools. The approach holds greater promise for rare diseases like PPB that are likely caused by a defect in a single gene.

So, armed only with a well-educated guess, Ashley Hill and her team from Washington University Medical Center in St. Louis traveled to Mankato to draw vials of blood from the outstretched arms of Anderson's family.

After mapping the disease on their family tree, Hill identified relatives who likely carry the faulty gene, such as Anderson's grandmother and great aunt, and then compared their genetic makeup with the rest of the family. It turned out that the suspected carriers and the afflicted had all inherited one particular segment on chromosome 14. The possibility of that happening solely by chance was less than one in 1,000. Hill expanded the study to three other families and found a similar result.

Unfortunately, the suspect stretch of chromosome 14 contained 72 different genes, and Hill felt she had a long way to go to find the genetic culprit—or culprits. She was a pathologist used to looking at cells and tissues under a microscope, not DNA sequences on a computer screen.  On a lark, she plugged the 10 most likely candidates into a public database of medical articles and noticed that scientists had recently found that a gene called DICER1 influenced lung development in mice.  She pulled up that 2006 paper and saw an image of a mouse that was missing both copies of the gene, and its lungs were dotted with cysts.  It looked exactly like the early stage of PPB.

The gene seems to encode a protein in one set of cells that helps slow down the growth of neighboring lung tissue and keeps cells from turning into tumors. When Hill later tested Anderson's genome for DICER1, she found that one of her two copies was cut short. "This is unbelievable," Hill says, "almost like it was meant to be found."  She and her team report their findings in Science this week. Hill, who is now chief of pathology at the Children's National Medical Center in Washington, D.C., says she is working to understand how Trisha Anderson and others lost the function of their second copy of DICER1, which is required for the disease to manifest itself.

Anderson, now 29, is an accountant and has a two-year-old son named Jackson* who has a 50 percent chance of coming down with the disease. If it is not caught early and treated properly, children with PPB have a 40 percent chance of survival. "He's in that age bracket when PPB shows up, so every little cough or cold gives me worries," she says. By knowing the gene involved, researchers can now develop a clinical test to help doctors diagnose PPB before its too late. And if Jackson proved to have two working copies of DICER1, then his mother would not have to keep taking him in for CT scans that expose him to dangerous radiation. "It's a glimmer of hope," she says.

*Correction (7/13/09): An earlier version of this story misidentified the name of Trisha Anderson's son.