The Antidote: Inside the World of New Pharma (51 page)

Keith Johnson got a phone call that Monday from the codirector of the cystic fibrosis center in the department of pediatrics and pulmonary medicine at Beth Israel hospital, Dr. Maria Berdella. He was at the W hotel in Union Square on business. “She said, ‘They’ve terminated the study,’ ” he recalls.

“She said, ‘Look, I know you’ve done really well on this med, but they say it’s not meeting their end points. They’re just terminating the study.’ I said, ‘What do you mean, terminating the study? What do you mean, not hitting their end points? Why take it away from
me
?’ I thought I was on it forever. It would be FDA approved, and somehow we would find a way to get it. I thought somebody somewhere would look at this and say, ‘Okay, this guy gets the drug.’ Somebody would be my advocate. There would be a mechanism in place, because why wouldn’t you want this for a patient?”

Ken had anticipated the disappointment facing Johnson and others
who had rolled over in the study, and he was “really ripped about the way we’re talking about it” inside the company. Smith wanted to disclose that almost a third of the patients rolled over, suggesting a robust response to the drug. But Ken worried that the criteria used in determining who got to stay on the drug after the initial evaluation period were badly designed and thus the data would ignite false hopes. Subjects in the first part of the trial qualified for the open label extension based on either a 10 percent actual rise in FEV1 or a 10 percent drop in sweat chloride—
at any single point in the study
. “You could have been twenty points below at the end of the study, but if you had a spike up ten points during the study you got rolled over,” Ken observed. Sweat chloride can’t be influenced by other factors. But FEV1 “is not exactly a precise measurement,” as he explained. “You have these spikes on a regular basis.

“As proof of that, we had a significant number of placebo group people who qualified for the rollover—because of an FEV1 spike. We have a whole bunch of people in that group and I had to fight to make sure that we were not going to disclose any of that, because I thought the disclosure was misleading. It turns out that we have now disclosed that we’ve terminated that rollover group—because it’s not having any effect. In other words, exactly what I was worried about becoming the reality. So I want to get that information out. I pushed to have that released because I want doctors to know that at least that little bit of data that we have
says it doesn’t work
.”

Ken feared that Vertex faced a dangerous dilemma if there was any ambiguity about whether 770 alone could benefit patients with two copies of the DelF508 defect. Despite persuasive preclinical evidence that the drug could be effective in people with other gating mutations, the FDA, moving cautiously, had indicated it was likely to approve it only for the 4 percent with G551D. With such a narrow label, the looming question of what physicians and payers would do regarding the other 96 percent once VX-770 was approved weighed heavily on the company.

“Because what’s going to happen here,” he says, “and I’ve already been told this by a bunch of doctors in the CF area, they’re gonna give it to people with delta-F508 mutations. They say they’re gonna do that, but what if the insurers don’t pay for it?—which in my humble opinion
they shouldn’t. I don’t believe that my tax dollars should go to funding Medicare and Medicaid drug payments for CF patients who have homozygous delta-F508 structures because it doesn’t work. But they’re gonna do it anyway. And they’re going to come to us and want us to do an expanded access. My answer, though it’s not controlling on the company—but if it were—is, ‘No.’ There’s no data that suggests that this works so we’re not gonna give it to you. I think it would be irresponsible to do that. I might as well give you apricot pits.”

Johnson was sure there was no way he had been taking a placebo. He knew his body. Since 2004, his chief goal each year was to go twelve months without needing IV antibiotics. Every year he’d failed, going six months, maybe seven, before a pulmonary exacerbation. The last time he was hospitalized was eighteen months ago. VX-770 had turned his life upside down and now, abruptly, he couldn’t get it—“a putting Keith on an island type of thing,” as he described his jagged emotions: loss, desertion, separation, exile, solitude, hopelessness, rage. He was told he had to return the drug, and Adrienne joined him at Beth Israel. “When we’re done with the visit,” he recalls, “I just leave. It’s just unraveling, and every hour that slips by it’s getting further and further away from me and I’m coming to the realization that we’re gonna have to live the way it was before. I’m not saying I wish I never had this experience. I’m glad I did. But every hour it’s further and further away in the rearview mirror.”

Vertex had devised a policy to enable selected patients, based on what used to be called “compassionate use,” to receive VX-770 before approval. In May, the company put in place an expanded access program that was reviewing about one hundred patients in the United States who had FEV1s below 40 percent but weren’t “in groups where we think the risks are too great,” Ken said. It wouldn’t, for instance, provide the drug to a patient on a respirator. “It’s unlikely they would benefit. Bad things happen to those people on a regular basis and you wouldn’t be able to prove that it wasn’t because of your drug.” Ken:

In CF you have some pretty compelling stories about people who are in rapid decline, and there’s a drug that can help them, and it’s not approved, so what do you do? Our position has been in the past
and continues to be that if we focus on a single patient, it’s almost always compelling. What you have to do is focus on the greatest benefit for the greatest number of people. That’s not what patient advocates want to hear, but that’s probably the most important thing a company can do. Our view is, we need enough data about a compound so that we believe we have a pretty good handle on its side effect profile and its toxicities.

The company’s interest has got to be to the larger group that might be harmed if the drug is delayed, or not approved. We need affirmative data that the drug has a certain profile that we have a handle on so we can assess whether to give it to a particular patient. The second criteria is that we need to be certain when we open up expanded access we’re not allowing expanded access for people who otherwise might be in a clinical trial we are either running or planning to run, because that will gut the trial process. If we allow people the certainty of getting the drug in an expanded access program versus the uncertainty—because there’s always a placebo—of getting the drug in a clinical trial, they’ll always opt for the expanded access, and the clinical trial program will die. The FDA understands that, and they take the same position.

Selling advanced data storage technologies gave Johnson a keen appreciation of both the scientific process and the perils of information-sharing. He was avid for privacy, spurning social media because of their invasive mining practices. He read carefully the endless releases he had to sign as a patient, and respected Vertex’s position, particularly the clauses in its release form designed to keep patients from broadcasting their experiences and feelings. Though he was too healthy to qualify for expanded access, and staggered by his re-reversal of fortune, he wouldn’t have disagreed with Ken’s logic or his arguments. “I just think about how irresponsible and reckless it would have been for me if I would have gone onto a forum or something like that and said, ‘I’m on 770 with double delta-F508, and it works. Everybody, you’re cured,’ ” he says. “You have to be careful about the expectations you set in a situation that is seemingly hopeless to begin with. So I have loads and loads and loads of
respect for that because I actually think it could interfere with the study and the science.”

At Beth Israel, after surrendering his meds, Johnson blew a 52 percent—near his original baseline. Indeed, ever since his peak FEV1s of 60 percent a year ago and 62 percent in November, his numbers had been declining. He dismissed the newest drop-off as part of the usual variability. “That was a bad day at the office, really, which will happen sometimes.” After his initial feelings subsided, he resolved grimly to fight through his abrupt new impasse, although he had no idea how. What he did know—the same thing that troubled Ken—was that FEV1 spiked and dipped, and there was much else to consider in deciding if a drug for cystic fibrosis was effective. “Who’s to determine what my baseline baseline is?” he asks.

“What does five percent mean to me? It means everything. It means getting up the stairs from the subway with no problem. It means being able to play with my kids, most of the day. Even five percent means life. It’s a different category of life.”

JULY 28, 2011

“When I see the dailies,” Partridge said merrily, “I think, ‘Okay, who’s guarding Wilt? You may want to put your hand up in his face a little more.” For the last couple of days, the Incivek-Victrelis sales figures dribbling out of IMS Health were stunningly lopsided: 92–8. Partridge was riffing on the all-time National Basketball Association single-game scoring record of 100 points set in 1962 by Wilt Chamberlain in Hershey, Pennsylvania, against a hapless and vertically challenged New York Knicks squad. If this was an indication of the ultimate outcome of Vertex versus Merck in hepatitis C, the fight could be over before it started.

Partridge posted himself at the whiteboard for the Q2 earnings call. After two decades of reporting on pipeline progress and clinical data, Vertex at last had product earnings to talk about. The Wall Street consensus for Incivek was around $30 million in sales. Porges, at the low end, projected $20 million. A “whisper number” of $40 million made
the rounds, but it included inventory, and most company-watchers dismissed it as hype.

Smith entered tapping on his BlackBerry, grinning. He had kept Vertex flush with other people’s money for a decade by raising their hopes; now he wielded results. Before reporting the figures and restating the company’s guidance on expenses, he told the analysts: “The opportunity that exists for Incivek is very large, and with this new opportunity we are giving careful consideration to managing reinvestment and R&D for future growth, yet achieving earnings and cash flow both for shareholder value.” Smith then announced that total revenues for the quarter had increased to $114 million, from $32 million a year ago, based on net product revenues from Incivek of $75 million—crushing the consensus figure. About half the total represented inventory build, “channel-filling,” not prescriptions. Vertex’s cash cushion had gone below $500 million but it finished the quarter with more than $600 million in the bank. Based on the sharp uptake of the drug, Smith forecast conservatively, the company would be “significantly earnings positive in 2012.”

In the twenty years that Wall Street had been trying to determine what Vertex might be worth based on models and projections, here was the first shred of real data, representing less than a month and a half of sales of its flagship drug. The absurdity was lost on no one involved in the call, but that didn’t stop an urgent spike in speculation. Citigroup’s Yaron Werber asked Smith to help him extrapolate.

“So we’re talking about maybe six weeks, and you sold seventy-five million dollars,” Werber started. He did a few back-of-the-envelope calculations. “You said half was retail demand, and then you said that retail stocking on the order of around thirty-seven, thirty-eight million is already out of the channel. So that’s within three weeks. And then there would have been inventory on top of that. It almost seems like I can back into a doubling of the rate of sales within the first three or four weeks of July over the last quarter. Am I thinking about this correctly?”

“Directionally, Yaron, I think you’re thinking about it correctly,” Smith said.

Ken, listening in on the call, worried that Smith’s comment was “on the verge of a projection, a forecast,” in which case Vertex might have
to issue a press release confirming it, in order to conform with Reg FD. He concluded no harm was done, but it was part of his job to keep watch over Smith’s close, ambiguous relationship with the Street. Ken:

I think Ian has actually pulled back a bit. The guy who’s really out on the edge right now is Michael, which is okay because that’s his job. He’s the guy that wants to answer every question that Wall Street has, and if we don’t have the answer we can speculate with them about it. I think Ian’s big task is coming up. I think he’s gonna be
very
hard pressed, emotionally, not to speculate on earnings and sales.

Ian absolutely talks the right way about not being driven by Wall Street on a quarterly earnings basis. What I haven’t seen yet is what’s actually the only thing that’s important, which is how you act, not how you talk. He presented something to the ET and Matt jumped all over it. I had gone crazy over one slide. The slide was four charts that showed what Wall Street’s projected earnings per share were for 2012 and 2013. And the headline of the chart was: How Can We Meet Wall Street Projections?

That is exactly what I’m talking about when I say walk the walk. Talk all you want about not being driven by Wall Street, but as soon as you start
thinking
about how you’re going to meet Wall Street projections, you’re off the reservation. I was really gonna go after it, and then, before I could say anything, we get to that page and Matt says, “Let me say something before we get any further. I don’t ever want to see a slide like this again.”