Linda Griffith is among the engineering faculty increasingly focused on cancer work. A tissue engineer, her main focus at the outset was on growing up bone graft materials from clumps of actual bone cells. Over the last eight years, though, Griffith has targeted another challenge as well: creating a “liver on a chip.”

The idea reflects the fact that the liver’s a key player both in many diseases and in the process of testing new drugs. In patient trials of such drugs, she notes, “liver toxicity is the biggest reason for failure.”

Since rodent tests don’t reliably flag potential problems with drugs, says Griffith, a professor of biological engineering and mechanical engineering, she decided to create something very close to an actual human liver, but on a strikingly small scale.

The concept sounds relatively simple: seed a specially designed, porous “scaffolding” with human liver cells; place the structures in tiny wells on a plastic chip; ensure the cells have the environment and nutrients they need to thrive; introduce the drug you want to test; watch what happens.

But human organs, like human everything, are immensely complicated systems. “It was really an engineering problem,” says Griffith, “and it wasn’t any one thing that took time, it was a thousand things.” Drug tests using her chip are likely to begin soon. “We’ve heard from companies who think our system has a lot to offer,” says Griffith, who heads MIT’s Biotechnology Process Engineering Center. But drug testing’s not all she has in mind.

Several cancers metastasize to the liver. Working with a physician colleague interested in prostate cancer, Griffith’s already seeded her micro-livers with cells from that type of tumor. (“You can actually see the tumor in our wells;” she notes, “they look really nasty.”) Studies of this and other tumors should help shed vital light on how cancer spreads.