FIFTEEN YEARS AGO, an up-and-coming oncology researcher in Boston named Catherine Wu had a hunch.

It went something like this: If Wu and her collaborators at the Dana-Farber Cancer Institute could identify some of the genetic mutations inside tumors, they could teach the body’s immune system to recognize tiny fragments of proteins associated with the mutations that sit on cell surfaces. And if the researchers could teach the immune system to identify a tumor based on those fragments, then they could teach the immune system to destroy it too.

Someday, the theory went, scientists might turn this capability into personalized cancer vaccines, genetically tailored to attack a patient’s existing cancer and prevent it from returning.

Science research is filled with these sorts of ambitious ideas. But a hunch is just a hunch. Wu and her team had no way to know if they were right.

And even if their hypothesis proved correct, developing these sorts of personalized cancer vaccines would require many more steps and many more discoveries. In asking the National Institutes of Health to fund their experiments, they were hoping the federal government would place a longshot bet—a $363,000 wager that insights into cancer biology from their studies might someday contribute toward the development of a workable medical treatment.¹

The NIH decided to make that bet, as one of about 5,300 standard grants it awarded that year. A decade and a half later, it looks like it has paid off. Big time.