Beall rode the emotional roller coaster of the CF gene discovery, and, like others, he expected new therapies to emerge. When everybody got swept up in the gene therapy craze of the early ’90s, he explains, “we were the same.”
But with CF gene therapy efforts failing, one after the next, Beall knew he had to find another way. “We had discovered the CF gene and knew the root cause of the disease,” he says, “but the pharmaceutical companies were still not getting involved.”
Beall told parents it was time to forge another path. He was scouring the scientific literature when he hit upon an article describing high-throughput screening, a new technique that used robots to test the therapeutic properties of thousands of chemical compounds a day in cells in laboratory dishes. Reflecting on the impasse, Beall thought this could provide an answer for CF: Instead of giving patients healthy CFTR genes, he would launch a massive search for chemicals to fix the mutant proteins in the patients’ cells.
Without a large government grant, Beall knew that no academic team could take on this challenge. And no company would embark on such an expensive drug search because it would never recoup its investment with such a rare disease. Instead, the CFF would need to front the effort, protecting companies from risk.
At the time it was unheard of for a nonprofit to take such a gamble. After all, Beall admits, he didn’t know much about drug discovery. “This is a risky thing that we are about to do,” Beall told his board at the time. “We are going to invest big. But the biggest risk for us would be not to do it.”
With initial funding of $3.2 million in hand, Beall called the five technology leaders specializing in the high-throughput screening mentioned in the article; two returned his calls. One was Aurora Biosciences, a San Diego-based start-up specializing in screening drug candidates. They agreed that it was possible to find a molecule to interact with the defective protein and correct it.
Aurora came with a perk no money could buy: Paul Negulescu, a cell physiologist who studied CF while a graduate student at Berkeley. As Negulescu well knew, the task was complex: Each of almost 1,900 mutations causing the CFTR protein to malfunction required its own unique fix.
But it made sense to start with the most common mutation, Collins and Tsui’s Delta F508, the one afflicting most of the CF population. (Of those with CF, some 50 percent carry two copies of the gene with the Delta F508 mutation, and another 40 percent carry one copy; those with just one copy of Delta F508 carry a second bad copy of the CF gene, with an alternate mutation that must be fixed as well.) Some 10 percent have rare mutations that don’t involve Delta F508 at all.
Negulescu knew that patients with the Delta F508 mutation produced a CFTR protein that couldn’t fold properly, like a crumpled origami sculpture, foiling its ability to even reach the surface of the cell, where it was supposed to be. To relocate the protein required one drug — dubbed a “corrector” — to tweak its shape so that it could be trafficked to the cell’s outer surface.
But once this defective protein was lodged in place, there was a second glitch: The protein still wouldn’t allow chloride to pass in and out of the cell. It was as if a door had been jammed shut. To wedge it open would require a second, “doorman” drug.
Patients with the Delta F508 mutation would need two drugs. But those like Laura and Cate — who have an even rarer mutation, called G551D — might be easier to treat. Unlike the common Delta F508 mutation, the G551D mutation yielded a protein that could reach the cell surface and wedge itself into the membrane, but it suffered from the door-jamming problem: Chloride still could not flow in and out. The Cheevers needed only the doorman to remove the jam that stopped chloride from flowing back and forth.
Hunt for Drugs
To accelerate drug discovery, Negulescu and the Aurora team searched for both corrector and doorman simultaneously. To execute their high-volume search starting in 1997, they grew cells carrying the malfunctioning CFTR protein in plastic trays, each containing 384 tiny wells in a 16-by-24 grid.
To identify correctors, a unique candidate drug was added to each well and allowed to steep overnight at body temperature, 98.6 degrees Fahrenheit. Next, a fluorescent dye was added to the mix. Then they added a chemical called genistein, a known door-opening drug that, unfortunately, was so weak it worked only in the test tube. Finally, a robotic eye scanned each mixture.
If cells were unaffected, the dye caused them to glow orange. But if the mutant CF protein had been elevated to the cell surface by a corrector candidate drug, then genistein, as the doorman, would open the channel and allow chloride in and out, making the cells glow blue.
To search for a doorman drug, Negulescu’s team followed almost the identical strategy, but they incubated cells with candidate drugs overnight at a much cooler 80.6 degrees. Conveniently, this cooler temperature acts as a corrector helping more proteins with the Delta F508 mutation to reach the surface, where they encounter the candidate doorman drug. If the chemical had no impact, the cells glowed orange. If the molecule could open the channel, the cells glowed blue.
The assay that Negulescu’s team developed worked. Some of the candidate drugs were definitely boosting the mutant CFTR to the cell surface while others seemed like they could open the door. It was proof that this new automated screening system could find something transformative. Yet after testing 122,000 chemicals in all, the researchers found that even those that showed early potential failed in later trials. Some were toxic, some too weak, and some, for whatever reason, couldn’t activate the CF protein on a second try.
Aurora scientists were confident that it was just a matter of auditioning more molecules to find ones that worked. And in 2001, when Vertex Pharmaceuticals acquired Aurora, it decided to continue the quest as long as Beall could muster the money to keep projects moving and the dream alive.
Parent Power
That’s when Beall recruited Joe O’Donnell, a CF parent and fundraiser extraordinaire. Joe and Kathy O’Donnell, Massachusetts natives, teamed up with the CFF the moment their son Joey was diagnosed with the disease in 1974. It was a brutal beginning. Tests were botched, the pediatrician hadn’t seen CF babies before, and Joey couldn’t eat, choking every time he tried sucking milk from a bottle.
After three harrowing months, another pediatrician familiar with CF finally made the diagnosis. The news was devastating; at the time, CF children rarely lived beyond 5. But after the correct diagnosis, Joey had a feeding tube inserted into his stomach to deliver predigested food and bypass the coughing and gagging, and he began to improve.
“Compared to the first eight months, the next five years were a party, almost,” says his father. “He got better, and he was a magnificent kid.”
By age 12, Joey was president of his seventh-grade class, a prankster, a grade-A student and a pinball wizard. “He loved girls,” his mother adds with a smile, “and he loved baseball.”
Yet the hospitalizations had become more frequent, and more serious. Many times they were told, “He wouldn’t make the night,” says his father, “but he always came through.” Except when he didn’t. He died in 1986. He was 12 and a half and barely 50 pounds.
Six months later, the O’Donnells, together with close friends, launched the Joey Fund to raise money for CF research. By 2001, it raised almost $50 million. To continue the drug search, Beall needed a lot more — $175 million, to be exact — and he asked Joe, a former board member, to bring it in.
O’Donnell admits it was an odd campaign. “That’s how I billed it. We’re not naming anything, endowing anything. We’re taking every dollar you give us and putting it into research,” he explains. “And guess what, we could end up with nothing. But for sure, we’re going to end up with nothing unless we do this. So that was the whole speech.”
O’Donnell insisted upon pure venture philanthropy, not venture capital. If there were royalties and other profits, he wanted that money rolled into more research, not someone’s pocket. No other foundation had successfully bankrolled a cure in this fashion. But as everyone who knows him says, he has a gift, and he raised the money. That was crucial because Vertex was making progress.
In 2002 and 2003 Vertex’s San Diego facility, which does the drug screening, tested another 200,000 compounds, and a couple of them looked promising. The top candidates — dubbed VX-770 and VX-809 — doorman and corrector drugs, respectively, made the mutant CF cells in Negulescu’s assay glow fluorescent blue, a sign that chloride was on the move. “That’s when we got excited,” says Negulescu, who was absorbed into Vertex to run the San Diego screens.
To test VX-770, the doorman drug, researchers used lung cells from a CF patient with the G551D mutation — the same one sickening the Cheevers girls, who required only a doorman drug to function in healthy mode. VX-770 made the cells glow blue, proving that the chloride channels were open. Scrutinizing the cells, Negulescu could see why: Lung cells are covered in fine hairlike structures called cilia. In healthy folks, these move back and forth and sweep mucus out of the lungs. On CF cells, however, the cilia are matted down with mucus, like a shag carpet covered in glue.
When Negulescu peered through a microscope at the sick G551D lung cells growing in dishes in the lab, they resembled a mat of small, still, gray spheres. But when his colleagues dosed these sick cells with VX-770, the tiny hairs sprang to life. Under the microscope, he could see cilia swaying back and forth, like a crowd at a stadium doing the wave. As the cilia swayed, the cells started to vibrate as if caffeinated. With an active chloride channel, the mucus would be watery, he thought, like in healthy people, and the revived cilia could sweep it away.
“That gave us so much optimism, some people were crying it was so beautiful,” says Negulescu. Perhaps VX-770 could do the same thing in live patients.
By 2007, the drug had been tested on cells and in a phase I clinical safety trial in healthy volunteers. VX-770 was on its way to becoming the drug Kalydeco.
The phase II trials for those with the G551D mutation were small, just 39 patients, but they were cleverly done to squeeze out as much data as possible. In spring 2008, Vertex’s chief physician showed Olson some data; a few numbers he saw were worth a thousand words. “It was remarkable,” says Olson. After just two weeks, concentration of salt in the sweat had plummeted from around 100 millimolar — a typical value when the CFTR protein is dysfunctional — to about 50 to 60 millimolar, a bit higher than average but below the diagnostic bar for CF.
Then, in 2010, as part of phase III trials, VX-770 was given to patients with the G551D mutation, including the Cheevers sisters, Laura and Cate. “This is a once in a lifetime for a pharmaceutical scientist,” says Olson, the project lead. “We are not just treating symptoms. We are fixing the protein that actually causes this disease.”
Laura began the trial unaware whether she was taking the drug or a placebo. She continued to cough, couldn’t gain weight and ultimately developed a severe lung infection requiring heavy-duty antibiotics.
For Cate, things were different. “I could just feel like I was getting better, like growing more, and I could see the difference between me and Laura,” says Cate, who coughed less, slept better, gained weight, ate like a horse and was bursting with energy. Later, Laura, who had been taking a placebo, was switched to the drug, and she, too, got well. The FDA approved the drug in 2012.
Experts agree the treatment is transformative for patients with the Cheevers’ form of CF. “We have just started using it in practice,” says Henry Dorkin, a pediatric pulmonary specialist and director of the Cystic Fibrosis Center at Children’s Hospital in Boston. “While it’s still early, the results are very encouraging.”
Dorkin began treating patients more than 35 years ago, and the window ledge of his office is crammed with pictures of kids — patients he’s treated, many of whom died from the disease. Patients typically lose about 1 percent or 2 percent of lung function each year. If the decline slows, or stops, and they continue to gain weight, then, Dorkin says, “I would have to say that it’s a game changer.”
Laura and Cate’s daily regimen of two pills of Kalydeco costs $841 per day; that’s $307,000 each year, making it one of the world’s most expensive drugs. In most cases, private insurance picks up the bill. Medicare or Medicaid may pay as well. For patients themselves, the cost is about $15 for one month’s supply of the drug. For those who lack insurance, Vertex offers financial assistance so they can access the drug.
The company can afford the largesse: Kalydeco saw windfall profits of $172 million in 2012, boosting the company’s stock price and visibility. Although many have questioned the ethics of that profit and the burden of the drug price on the health-care system, Beall says that without Vertex, there would be no drug. And, for the CFF to negotiate a drug price before there was even a drug would have been a deal breaker.
CFF has also profited from the discovery and sales of Kalydeco. It just sold a portion of the royalty rights for the drug, bringing in $150 million. As per the business plan that O’Donnell vehemently supported, that entire amount will be reinvested to fund more CF research and drug development. That’s important because Beall is not complacent that Kalydeco is a cure.
“We need to be careful about using the word ‘cure’ when talking about Kalydeco, although the drug is clearly a game changer and has fueled incredible optimism in the CF community — and for me personally,” says Beall, “but we have to be cautious. We only have two years of data on how patients are doing on the drug, and it’s premature to say whether it will be a cure for them.”
Indeed, even with Kalydeco, Cate and Laura still must take 20 to 30 pills a day to digest their food. They also require 30 minutes of physical therapy — clapping and beating their sides and back — to help dislodge the mucus. But they are staying free of lung infections and gaining weight.
Completing the Cure
Those with two copies of the Delta F508 mutation — half the CF population — are watching the Cheevers to see if the treatment keeps working its magic. They are waiting as Vertex conducts phase III trials of the second drug, the corrector, VX-809, in combination with Kalydeco, the doorman, to see whether the defective proteins can reach the cell surface and open the door to get chloride flowing again. If it works, this drug combo could halt the disease in its tracks for the majority of patients — as Kalydeco seems to have done for Laura and Cate.
There’s reason for hope. Phase II of Vertex’s combo trial showed that the VX-809 plus Kalydeco improved lung function in patients with two copies of the Delta F508 mutation. The larger and longer phase III study of 1,000 patients will continue for 24 weeks and should yield an answer in 2014.
Beall isn’t placing all his bets on Vertex. O’Donnell has embarked on another campaign for $75 million to give to other pharmaceutical companies. The Foundation has already invested $58 million with Pfizer to develop a second generation of similar but more potent drugs to treat those with two copies of the Delta F508 mutation.
While Kalydeco and the corrector should work, Vertex or another company may ultimately be able to engineer more effective molecules. And Beall is still concerned about the 40 percent of patients who have only one copy of the Delta F508 mutation and one copy of another mutation: How effective will Kalydeco and corrector combo be in this group?
So the search goes on. “We’re not going to settle for less than 100 percent of patients,” he says.
To this end, Vertex is expanding clinical trials to encompass other rare mutations. The R117H mutation, carried by about 3 percent the CF population, creates a protein that reaches its destination on the cell surface, but then malfunctions. While the impediment is unclear, Olson guesses it may be the door-jamming problem.
If so, as with the Cheevers’ G551D mutation, Kalydeco might fix the defect. Olson adds that there are also two more corrector drugs in the pipeline, VX-661 and VX-983. “You want to fill your nest with lots of molecules, each of which has slightly different properties,” he says. People carrying different mutations may require specific correctors, or more than one corrector, or a complex combination of these drugs, which is why Vertex continues it search.
“Robert Beall deserves a lot of credit for placing a huge and expensive bet on an enterprise that could well have failed,” says Collins, the current NIH director. The CFF’s strategy is a promising model for attacking other genetic diseases, Collins adds, and other groups are trying to embrace the innovative drug development model as well.
CFF may soon succeed in creating a long-sought cure, but for Olson and Negulescu, the journey has been bittersweet. Over the past 15 years, the scientists have embraced the CF community. They’ve participated in fundraising walks and bake sales. They’ve become acquainted with CF families who have visited Vertex to share their stories and participate in research. Along the way, they’ve experienced many losses. Olson describes one family who lost three children within the past three years. “We just weren’t fast enough for that family.”
Every year the CFF holds its annual meeting in a different city, and over the past decade, Joe O’Donnell has gotten to know many members quite well — volunteers, mothers who have lost a child, others who are on the cusp. In October 2012 in Orlando, Fla., there were nearly 4,000 who refused to quit working toward a cure, all the time wondering whether it was really going to happen for them.