Boosting the effectiveness of cancer therapies
Michael Hemann has an idea: instead of holding out for a series of breakthroughs by medical science and the drug industry in creating cancer therapies, let’s make existing drugs work better.
Sound like the musings of a certified non-expert? In fact, Hemann, an assistant professor of biology and MIT Center for Cancer Research member, thinks the concept will soon start becoming a reality.
The problem at issue is that most chemotherapy regimens work only for some. “Patients come in with a given type of tumor,” notes Hemann, “and the oncologist knows that the treatment selected is going to help some proportion of those patients — 50 percent, 80 percent — and not the others.”
Until recently, many scientists would have dismissed the idea that we could reliably identify the most likely beneficiaries. New biological tools, though, are opening ways to help doctors do just that.
“A lot of existing therapies would function more consistently if we could understand why, how, and when they work,” says Hemann. His overarching goal in seeking that knowledge, he adds, is to keep cancers from evolving such that they’re beyond medicine’s reach.
Surgeons, he notes, usually succeed at eliminating primary tumors — the main, and often the only, initial manifestations of a cancer. The real problems can arise when tiny remnants of a malignancy remain behind, typically well outside the primary tumor’s site.
“That’s when you may have a cancer that two or three years later turns out to have metastasized,” he says. “If that occurs, the chances of a cure are poor.” But if treatments can keep a cancer from relapsing, or can at least markedly delay that event, the patient benefits even if the disease isn’t necessarily cured. And Hemann says our improving ability to understand and manipulate genes is likely to enable that advance.
Hemann joined this quest in a roundabout way. He grew up in Cleveland, and though his family came from a line of farmers — “I chased a lot of pigs around as a kid,” he says — Hemann’s own dad is a university engineering professor.
It’s a background that often points bright youngsters toward science or engineering careers. But while Hemann was fascinated by his dad’s work, which included efforts to improve the fuselage tiles on space shuttles, it didn’t translate into a yen for the scientific life.
His intended major at Connecticut’s Wesleyan University was history — and though he eventually chose biology, he didn’t have in mind a scientific career. Still, when he later moved to Boston and had a chance to work as a lab technician, his choice of major made it seem like a reasonable fit.
In fact, it was much more. “From the first day you start work in a lab, you’re a scientist,” says Hemann. “You may not be an experienced scientist, but on any given day anybody can have a great idea.”
Hemann went on to graduate studies at Johns Hopkins and postdoctoral work at Cold Spring Harbor Laboratory on Long Island. He then came to MIT to work on cancer.
The question of why clinically identical tumors can respond so differently to specific treatment regimens has long been seen as related to genes: one patient’s genetic status differs in subtle but crucial ways from another’s. The problem has been figuring out for any given patient what the key gene defects are.
It’s a colossal challenge, because hundreds of genes can play a role in the development and spread of a tumor. “It’s like that kid’s game where you have a structure made up of a lot of small pieces,” says Hemann, “and you have to pull them out one by one until you find the one that’s holding the whole thing up.”
A technology generically known as RNA interference, though, has given researchers like Hemann the ability to shut down virtually any gene of interest in, say, a lab rodent. And that’s yielding key insights into issues like how cancer drugs interact with specific genes.
“It’s a level of technology that would have been out of the question five years ago,” notes Hemann. But in a recent study, Hemann’s group used it to turn up intriguing results about the workings of two popular cancer drugs, doxorubicin (trade name: Adriamycin) and camptothecin.
Doxorubicin works by blocking a specific DNA-modifying enzyme — let’s call it enzyme A — and its counterpart blocks another, enzyme B, that looks different but kills off tumor cells in a similar way.
Now, Hemann’s group has shown that many tumors seem to have either A or B, but not both. Importantly, they’ve also shown that both doxorubicin and camptothecin work much better when their “partner” enzyme is present in tumor cells while the other’s “partner” is missing.
In short, says Hemann, if tests show that a patient’s tumor cells are making, say, A but not B, “you can predict which therapy they’re most likely to respond to.”
The emergence of this kind of ability to uncover relevant genetic facts about a specific tumor is a big reason Hemann thinks the era of personalized cancer treatment may not be all that far off. “We have access to a tremendous amount of genetic information,” he says, “and as a result we’re really turning a corner in terms of our ability to introduce molecular logic into cancer treatment.”