Why it’s so hard to develop a drug for cystic fibrosis

Why it’s so hard to develop a drug for cystic fibrosis



When the gene for cystic fibrosis (CF) was discovered 25 years ago, hopes ran high that new treatments were just around the corner. But novel drugs proved hard to design, and sometimes they don’t work as expected. Now, a pair of studies suggests why two breakthrough therapies might not help many patients as much as hoped.

CF is caused by mutations in the CFTR gene. Like many genetic diseases, a host of different mutations can trigger the condition, and median life expectancy for patients hovers in the late 30s. About 90% of patients carry at least one mutation called ΔF508—and it’s one that is especially complicated to repair. The mutation disables the CFTR protein in two different ways: It misfolds it, and it prevents it from reaching the cell surface, where it’s supposed to help control the balance of water and salt in cells. When CFTR can’t do its job, people develop a dangerous buildup of mucus in the lungs and elsewhere.

One of the new CF drugs, VX-770, was designed to tackle a much rarer CF-causing mutation. Called G551D, it’s behind about 4% of U.S. cases; the protein gets to the cell surface like it’s supposed to, but it’s closed up tight and can’t function. VX-770, designed by Vertex Pharmaceuticals, opens the protein into the right configuration so it can do its job. VX-770 was approved in 2012 for patients with this particular mutation, under the brand name Kalydeco.

Although this is great news for patients with G551D, the far bigger pool, with ΔF508, doesn’t benefit from VX-770 alone. Scientists had hoped that combining VX-770 with another still-experimental Vertex drug, VX-809, which helps move the CF protein to the cell surface, would do the trick. Two groups decided to explore this idea more thoroughly in lung cells.

What they found was surprising. A team led by cell biologist Martina Gentzsch at the University of North Carolina, Chapel Hill, looked at a type of human lung cell that’s often used in CF studies, with the ΔF508 mutation. They found that chronic exposure to VX-770, which is supposed to activate the mutated protein, made the other drug, VX-809, less effective, and this was especially true at high doses of VX-770. “At high concentrations” of VX-770, when combined with VX-809, “there is a very strong inhibition” of how well the protein functions, Gentzsch says. At lower doses, the mutated protein wasn’t always as affected. The work conflicts with earlier cell studies, in which VX-770 was given only briefly and showed success. Chronic exposure, it seemed, had a more complicated effect.

At McGill University in Montreal, Canada, cell biologist Gergely Lukacs and his team saw something similar. They found that exposing lung cells with the ΔF508 mutation to VX-770 long-term—mimicking what happens in patients who take the drug for extended periods—often made the CFTR protein function poorly when compared with short-term exposure. Lukacs’s hypothesis is that while in some respects VX-770 might be getting CFTR into the right functional shape, in other ways it’s destabilizing it. “It has this duality of effect,” he says.

What does this mean for patients? The papers, published today in Science Translational Medicine, coincidentally appear just weeks after Vertex announced results of a phase III trial, in which patients with the ΔF508 mutation got both drugs for a prolonged period of time. Their lung function did improve, albeit modestly, by 2.6% to 4%.

“At first blush, this might seem at odds” with the cell studies, says Philip Thomas, a biochemist at the University of Texas Southwestern Medical Center in Dallas, who studies CF and wasn’t involved in the current research. And it’s clear that in people with the most common mutation, the two drugs do work better together than either does on its own. But, he and others believe, the cell work is teaching the CF community something important: The drugs may interfere with each other, packing a less powerful punch than doctors and patients would like to see. The combo may work somewhat in patients because the dose of VX-770 is probably lower than in the lab work, he suspects, dampening the negative impact the cell studies report.

“It definitely makes me think there’s room for improvement,” says John Clancy, a pediatric pulmonologist at Cincinnati Children’s Hospital Medical Center in Ohio who wasn’t involved in the research. Back in his lab, Lukacs has also worked with an experimental compound similar to VX-770 that didn’t have the drug’s downsides. In the long run, he suggests, that compound might be a better option for some patients.