Vertex Pharmaceuticals shot itself in the foot last month when it reported on some encouraging partial results from an ongoing clinical trial, then later had to correct some faulty statistics. Now it’s had time to run the final analysis from that study, and while the numbers have changed a little, the conclusion hasn’t: It still looks like it has an important new combo treatment in the works for cystic fibrosis.
The Cambridge, MA-based biotech company (NASDAQ: VRTX), which has significant operations in San Diego, is reporting today that its combination of VX-809 and ivacaftor (Kalydeco) reached its goal of improving breathing ability in a mid-stage study of 109 cystic fibrosis patients. The combination of the drugs helped patients with two copies of the F508del gene mutation, who are known as homozygous patients. The effect wasn’t as clear for patients with just one copy of the mutant gene (heterozygous patients).
The company plans to release the detailed findings at an upcoming medical meeting, and it hopes to be ready in early 2013 to start the third and final phase of clinical trials normally required for FDA approval. If Vertex can nail that next big trial, it will be able to build on the momentum it has gained this year with ivacaftor, a groundbreaking treatment the FDA has approved for about 4 percent of the cystic fibrosis population that has a mutation to the G551D gene. That drug made history as the first of its kind to work against the underlying biological problem, rather than just treat the symptoms. Now Vertex is hoping that by creating a combo-drug regimen, it can extend some of the benefit across a wider group of patients. About 30,000 people in the U.S. suffer from cystic fibrosis, and about half have two faulty copies of the F508del gene, making them potential candidates for the new combo therapy.
“When we came out in May with the interim analysis, I stressed that the data was encouraging but that it was still just an interim analysis. It made me hesitant to draw too many conclusions,” says Michael Boyle, the director of the Johns Hopkins Adult Cystic Fibrosis Center, and the study’s lead investigator. “But with these final data, we can say conclusively that we are seeing a promising effect on lung function in the homozygous population.”
Before I proceed to the deep-dive into the biostats on this final analysis, here’s a little background. Cystic fibrosis, the result of mutations to a gene called CFTR, impairs the transfer of water and salt across cell membranes, which leads to the buildup of thick, sticky mucus in the lungs. It causes people to suffocate gradually, and patients usually die by their late 30s or early 40s. Doctors have made significant progress over the years through improved treatment of symptoms, but Vertex’s ivacaftor is different because it’s the first drug that works by targeting the CFTR protein directly; helping to prop open these cellular gateways to allow better transport of ions across cell membranes.
The problem for most people with cystic fibrosis, though, is that their genetic form of the disease prevents CFTR proteins from getting close enough to the cell surface to benefit from the effect of ivacaftor. So Vertex, which got $75 million in financing for its effort from the Cystic Fibrosis Foundation, went a step further by developing VX-809. The second drug is supposed to work by helping shuttle the CFTR protein toward the cell surface, where ivacaftor is doing its thing, opening the doors to effective transport.
The study was designed to randomly assign patients to get a 200 milligram, 400 milligram, or 600 milligram daily dose of VX-809 or a placebo for the first 28 days. Researchers did that because they wanted to see how well the drug worked on its own after 28 days, to provide more understanding about how the drugs work together. So patients stayed on VX-809 and then also started getting ivacaftor along with it from days 28 to 56. Researchers were keenly interested in seeing whether the two-drug combo could help improve lung function after 56 days.
Patients with cystic fibrosis gradually lose lung function, and are typically expected to lose about 1 to 2 percent every year on a score that measures how much air they can force out of their lungs in one second—a measurement known as FEV1. Sure enough, patients on the placebo in this study lost 0.9 percentage point on that score after 28 days, and those on VX-809 also declined in that period, which wasn’t surprising because of the way it works, Boyle says. What came next is what impressed researchers. From Day 28 to 56, the placebo patients continued their steady decline, while the patients with two mutant copies of the F508del gene who got the high dose of VX-809 and ivacaftor rebounded. Instead of declining, those on the drug combo saw their lung function improve by an average of 6.1 percentage points from days 28 to 56. The difference in lung function was 8.6 percentage points, on an absolute basis after 56 days, for the patients on the drug combo, compared to those on the placebo.
“As soon as you start the combo therapy, there’s a clear distinction,” Boyle says. “The placebo patients continue to have some decline, and the treatment groups immediately start improving. It’s both clinically meaningful and statistically significant,” he added, meaning that it’s unlikely the benefit is a result of random chance.
Those statistics cited above are averages, but physicians in cystic fibrosis also want to know how many patients are getting to the standard benchmarks of improvement in lung function. In this study—again among the F508del patients with two mutant copies, and those getting the highest dose—the two-drug combo excelled. Eleven of 20 patients in that group (55 percent) reached a threshold in which they had at least a five percentage point improvement in lung function, compared with just two of the 21 placebo patients (9.5 percent). Five of the 20 patients (25 percent) reported a 10 percentage point gain in lung function, while none of the placebo patients did that well.
“Five percent is the magic number, that’s what everybody wants to get to. But 10 percent is a huge effect,” Boyle says. When I interviewed Boyle last month about what those numbers mean, he said the outcome can vary depending on how sick a patient is, but it always counts for something important. A person who improves from 20 percent lung function to 30 percent goes from the verge of a lung transplant to having a longer life expectancy. An improvement from 40 percent lung function to 50 percent clearly enables a patient to exercise more. A person going from 50 percent to 60 percent can feel a significant difference when climbing stairs, he said.
Vertex’s stock soared in May when it reported the first interim results from the first 48 patients. It lost some of those gains—and some credibility—a few weeks later when it had to correct its earlier press release. The company said it had mistakenly reported on relative improvements in patients when it had meant to report on absolute improvement.
It was a significant error, and although it made Vertex’s combo drug benefit appear more modest, Boyle says it didn’t change his conclusion about the drug combo. The benefit is still strong, and it was reinforced in this final analysis, which he says clears the way for a pivotal study. He says there is still enthusiasm in the CF community, which should help researchers recruit patients for the Phase III trial coming up in 2013.
The biggest fallout from last month’s correction was some even more thorough than usual double-checking of the final numbers this week.
“It wasn’t just me, but everybody wanted absolutely no chance of a repeat correction,” Boyle says. “We looked at the data upside down, back and forth, sideways, you name it.” While he says investors will focus on the magnitude of benefit in this study, he said it’s probably best not to read too much into the data at this stage. “This is still Phase II, so what we’re trying to do is decide if we are confident that we have a clinical effect at a dose and a level of safety that allows us to move to Phase III,” Boyle says. “The answer to that is yes.”