In the expanded-access study, a widespread collection of individual clinical trials for patients who were still in need of the drug after the phase II trials had completed enrollment, the responses followed the same patterns. The results of these three studies were without precedent in the history of cancer treatment. No drug had done better.
The side effects were the same as in the phase I study. There was some nausea and some diarrhea, but those subsided after patients acclimated to the medication. The stomach cramps let up, and the leg pain ran its course. Puffy eyes abounded. There were some rashes here and there. One patient died from liver toxicity, most likely because the presence of large amounts of acetaminophen in that patient’s body had exacerbated an otherwise minor problem, an unforeseen complication. Between 10 and 20 percent of patients did experience a severe drop in blood counts, a serious condition that may have been a sign that the drug was working, since the mutant Philadelphia chromosome had been triggering the growth of too many white blood cells.
By the end of 2000, Novartis had enough data to submit the drug for FDA review.
THE REPORT SUBMITTED to the FDA included all of the toxicology studies, the phase I trial, and the phase II trials. In addition, the FDA was given all the information about the composition and manufacture of the drug: what chemicals were in it, the step-by-step process of how it was prepared, all other particulars about the manufacturing process, its inactive ingredients, and its form.
In early February 2001, a truck carrying boxes of binders filled with pages upon pages of data and information left Novartis. On February 27, 2001, the drug was officially filed for review by the FDA. Forty years had passed since Nowell and Hungerford spotted the abnormally small chromosome, and thirty years since Janet Rowley discovered the translocation at the heart of the Philadelphia chromosome. The binders sent to the FDA were the culmination of decades of research—the discovery of the Abelson virus at an NIH laboratory; the piecing together of the Bcr/Abl fusion protein and its connection to CML by Owen Witte, Naomi Rosenberg, and many others; the gradual shedding of light on kinases and phosphorylated tyrosine; the cellular origin of oncogenes; and proof that the mutant gene encoded a mutant kinase that was the sole cause of CML. Eighteen years had passed since Ciba-Geigy gave Matter the resources for his under-the-radar kinase inhibitor program. Years of chemistry work to create a molecule that could block that kinase had eventually generated an experimental molecule that blocked the kinase in question, and years of struggle to turn that molecule into a drug had followed. For long stretches, no one involved in the work felt certain this moment would ever come, and yet they’d all held out hope that it would.
plate 19
Finally, the principle had been proved. The kinase had been killed, and so had the cancer. Now the future of this drug was in the hands of the FDA. So much rested on the approval of this drug—the lives of CML patients, and possibly, if the principle applied to cancer generally, the lives of so many other sufferers. Proving the principle of kinase inhibition had come to be about much more than treating a rare leukemia. The drug was changing the way people thought about cancer. This deadly illness, so impervious to decades of medical efforts, was finally being wrestled into a new era by virtue of the growing clarity that cancer is, at its root, a genetic disease. With that knowledge, the possibilities for bettering treatment were not just exciting—they were real and concrete, backed by evidence, grounded in logic and science.
When a new drug application is filed with the FDA, the application doesn’t arrive as a surprise. The agency is well aware of the progress of various experimental drugs and the status of submissions for its review. In this case, the FDA was especially prepared. Agency officials knew about the study results so far and the media coverage of the drug. The fast track and accelerated approval designations meant the agency had been involved more closely than it usually was with new drugs.
Novartis had also secured another FDA designation for STI-571 that garnered it special attention. Just days before Novartis shipped off the binders full of clinical trial data, the FDA granted STI-571 orphan drug status. Orphan drugs treat rare diseases, those occurring in fewer than 200,000 people per year in the United States. Orphan drug status can also be granted to drugs that treat more widespread diseases but that, for one reason or another, are unlikely to recoup their development costs. For the bulk of pharmaceutical industry history, rare diseases were largely ignored because companies knew that the money required to create a drug was unlikely to be recovered and that profits were even more unlikely. In addition, conducting clinical trials for rare diseases was thought to be exceedingly difficult because there wouldn’t be enough patients for a trial to yield statistically meaningful data.
The Orphan Drug Act, or ODA, was passed in 1983 to change this dire circumstance. The law provides incentives for the pharmaceutical industry to turn its attention to orphan ailments: federal funding for clinical trials, a 50 percent tax credit on trial costs, and, most important, seven years of market exclusivity. That last provision guarantees that no competing drugs will be sold for the same illness for seven years, even if the patent expires. Patents are obtained much earlier in the development process, often before an experimental compound has been made into a drug. This protection generally lasts for thirteen years, but by the time clinical trials are completed and the FDA approves the drug, those years may be more than half over. Market exclusivity, which is not granted to drugs for more common diseases, means that a company with a successful new drug for a rare disease will own that market for a full seven years.
The ODA has dramatically influenced the pharmaceutical industry’s willingness to add rare diseases to its scope of interest. Before the act, there were ten drugs approved for the treatment of rare diseases in the United States. By 2010, 367 orphan drugs were on the market, with more than 2,000 in the pipeline. The federal financial incentives, coupled with the high prices typically set for these drugs (some cost $500,000 or more per year) turned rare diseases into a goldmine for the pharmaceutical industry. The orphan drug market was worth more than $58 billion by 2006, with a predicted annual growth rate of 6 percent, bringing that amount to more than $112 billion by 2014. Even if an orphan drug later turns out to be a blockbuster (as happened with Botox, originally approved for the treatment of a rare muscle disease and later approved as a wrinkle smoother), the drug maker is assured that the ODA benefits will not be rescinded.
Because CML struck only about 5,000 people per year in the United States and treatment options were so inadequate, STI-571 was a perfect candidate for orphan drug status. The designation didn’t ensure profits, but it did take a significant edge off the investment. The tax credit and market exclusivity would certainly go a long way toward paying back the company’s investment. And its standing as an orphan drug made reviewing STI-571 even more of a priority for the FDA.
In March 2001, with the data already under review, Novartis received the news that STI-571 had been granted priority review. This status completed the process of special consideration that had begun with fast track designation, which gave Novartis privileged access to the FDA, followed by accelerated approval status, which enabled the drug to be reviewed based on the surrogate markers of benefit measured in the phase II studies. Priority review simply meant that the FDA would read through all the data more quickly than it might for a drug that was not offering such a significant improvement over currently available treatment options. The goal was to have an answer by six months after the submission date for the new drug application. The Novartis executives, the investigators, and the patients were anticipating that the drug would be approved by around September 2001.
With the submission, Novartis was required to provide a name for the drug. The generic name the team had selected was imatinib mesylate. The brand name chosen by the company was Glivec, pronounced “GLEE-vek.”
Under fast track and accelerated approval, Novartis had worked with the FDA to craft the phase II studies in a way
that satisfied the agency’s criteria for reviewing new drugs: what kind of response could be considered a true surrogate for clinical benefit, what defined an interferon failure, how much follow-up time was needed for the patients in the trial.
The review period was fairly uneventful for a while. The agency had some questions and so requested more information. Some wording in the proposed package insert needed changing, and the company had to clarify its definitions for certain aspects of the disease or response to the drug. Then more wording changes for the package insert. Mostly the issues were minor and easily resolved.
Of all the information included in those binders shipped to the FDA, the name proved to be the greatest sticking point. Names draw heavy consideration in FDA new drug reviews. A proprietary name is not allowed to sound like the disease it treats, and it can’t be too similar to other drugs on the market, even if those medicines address vastly different problems. The choice of brand name follows no obvious metric. Generic names tend to be garbles of mysterious syllables, occasionally referring to the type of molecule at hand (the “inib” in imatinib, for example, is short for “inhibitor”). By comparison, brand names are easier to remember—Glivec, instead of imatinib mesylate, certainly fit that bill. But the goal of creating a memorable brand name doesn’t explain where the names actually come from. Often, pharmaceutical marketing teams select from lists of potential names, plucking one at random that seems to fit. Novartis’s choice had nothing to do with happiness. It was just a name.
In mid-April, after an extensive review of the proposed proprietary name, the agency told Novartis that its choice, Glivec, sounded too similar to a drug named Glyset. Novartis fought the decision. The medication was for a serious, rare disease, and therefore would be distributed in a highly controlled manner to a very small population. It wouldn’t be crossing paths with Glyset, so mixing up the drugs wasn’t an issue. The Glivec pills in no way resembled Glyset, a diabetes drug. Plus the diabetes drug name was pronounced with a long i, to rhyme with cry, not with a long e, as in glee.
The FDA was not swayed. That more than one hundred cases of mix-ups had been reported among the drugs Celebrex, Celexa, and Cerebyx, each for different uses, was proof that the sound-alike concern was well founded. The proprietary name study and ensuing debate runs fifteen pages long in the FDA’s review document. After the prolonged debate, the solution was simple. The pharmacist conducting the name review suggested that the company change the name to Gleevec, to match the pronunciation of Glivec. Novartis relented.
Many trial patients, especially those who’d been there from the start, took an immediate dislike to the name. They aired their complaints in the STI Gazette. To them, the name “Gleevec” sounded totally strange. “Seems a bit odd for such a killer drug,” one patient wrote in, “but who’s complaining.” Some wanted it to be named Drukercillin. To all of them, it remained STI-571. Referring to it by the pipeline name, or even just STI, like a nickname, became a point of pride and sentiment. Those who were truly in the know and whose lives had been saved by this little pill would always refer to it as STI-571.
On April 30, 2001, the FDA sent a fax to Novartis requesting that the proposed package insert—the folded paper of fine print that would be included with the bottle of pills, should the drug be approved—include all of the side effects that had been observed in patients during the studies, not only those thought to be caused by the drug. Novartis knew that, per the accelerated approval designations, they were guaranteed an answer from the FDA regarding approval by September, six months after the data had been submitted. Communications from the FDA were only a sign that the review was continuing, not that approval was imminent. Most drug reviews, those that were not accelerated, took twelve to fifteen months on average. The goal of a six-month review seemed fitting for a home-run drug like STI-571. But was it possible for the agency to move that fast?
On May 10, 2001, Novartis received a letter from the FDA. “We have completed review of this application, as amended, according to the regulation for accelerated approval,” it read, “and have concluded that adequate information has been presented to approve Gleevec (imatinib mesylate) 50 mg and 100 mg capsules for use as recommended.”
It was the letter that patients, investigators, and various teams at Novartis had been waiting for, some for many years. From a genetic mutation to its haywire fusion protein, from the fusion protein to the leukemia, a path had been traced for the rational design of this drug. From CGP-57148B to STI-571 and finally to imatinib mesylate and Gleevec, the world’s first drug targeted against a specific mutant protein that stopped cancer in its tracks was ready for wide release. The drug was approved.
The FDA approval regulated only how the drug could be marketed. But marketing the drug was exactly what the company had been waiting to do. Now it could sell the drug rather than only pay for its use in clinical trials. Insurers would add it to their formularies, the lists of medications covered by their policies, the final step needed to earn money from prescriptions written to CML patients. Novartis still had to complete the phase III clinical trial; per the accelerated approval program, the FDA’s nod was still conditional as the data from the large, randomized study matured and the actual survival benefit emerged. But there was no reason to doubt that anything would change with this drug. Doctors were free to prescribe it. Patients across the country could get it. The drug had been set free from clinical trials.
With the approval, the fears about toxicity shriveled. The FDA confirmed that although liver and kidney problems had been observed during the trials, these were temporary, minor, and reversible. The reviewers concluded that the toxicology studies supported approval of the drug. The package insert did include a warning that women should not become pregnant or breast-feed while on the medication, in light of evidence that the chemical had seeped into the milk of lactating rats. But all of the concerns about what harm the drug might cause—the worries that had spurred tests in rats, rabbits, mice, and monkeys—had been allayed. Even the blood clots in dogs, the problem that had screeched the drug’s development to a halt, fortuitously enabling the oral formulation to take center stage, had turned out to be a false alarm. Some time after those canine tests it had become clear that the catheters, not the drug, had been the problem. The last sighs of relief over potential problems had been breathed long ago. The drug was powerfully effective, and it was safe.
33
_______
THE FATHERS OF VICTORY
The approval was announced at a news conference held in Washington, DC, by the National Cancer Institute and the FDA. On May 10, 2001, several agency officials stood in front of members of the press and invited guests to praise those responsible for the breakthrough, and to highlight the rapidity with which this life saving drug had been brought to the people who needed it. The speakers included Tommy Thompson, secretary of the Department of Health and Human Services; Richard Klausner, director of the National Cancer Institute; Richard Pazdur, who directed the oncology drug products division at the FDA; Suzanne Dreger, a patient from Falls Church, Virginia, who’d been part of the clinical trials at OHSU; and Daniel Vasella.
Although Gleevec had already made the pages and websites of every major news outlet, the press conference was the official coming-out party for molecularly targeted medicine. “This drug has been engineered in the laboratory to target a single, cancer-causing protein, and like a light switch, turn off its signal to produce leukemia cells,” said Thompson, reading from his prepared notes. “We believe such targeting is the wave of the future.” Klausner, the most expert oncologist among the speakers, offered a similar view of the significance of the approval. “This new drug, we believe, is the picture of the future of cancer treatment, and a vindication of the scientific approach to disease,” he said.
Everyone acknowledged that the long-term benefit of Gleevec was still unknown. Not enough time had passed to confirm whether the drug actually prolonged the lives of people with CML without eventuall
y causing intolerable side effects. People newly diagnosed with the disease had four to six years to live, on average, and some much more than that. The phase I patients had been on the drug for less than three years, and the phase III study, in which newly diagnosed patients were receiving Gleevec as their first treatment, was still ongoing. So the answer to that ultimate question—did Gleevec prolong life?—was still years away. But, insisted Klausner, molecularly targeted therapies were the key to “a long but hopefully more successful war on cancer.”
Klausner differentiated between other drugs that target specific, known substances in cancer cells. He was well aware that the breast cancer drug tamoxifen was really the first targeted drug because it works by specifically blocking the production of estrogen. The difference, said Klausner, responding to a question from a journalist, is that the estrogen receptor targeted by tamoxifen is not the root cause of cancer. CML is a one-gene disease. A single mutant gene encodes a single mutant enzyme, and this enzyme alone causes the cancer. “This was a target that is not just in a cancer but is responsible for the cancer,” he said.
His optimism about molecularly targeted drugs extended far beyond CML. “There are scores of drugs now that fit into this class,” he said. As the clinical trials of STI-571 had been churning along, the oncology community had been digging for potential molecular targets in other types of cancer. Researchers probed for genetic mutations hidden inside breast tumors, lung tumors, kidney tumors—any malignancy at all. They measured the amounts of different enzymes and other proteins to see if any were occurring in excess, a potential sign that the substance was part of the ecology enabling the lesion to grow, if not the root cause. In academic laboratories, researchers continued to unpack the signaling pathways triggered by a particular gene that may be responsible for causing cancer. By the time the drug was approved, more than sixty potential targets had been identified in breast cancer. Molecular biology had transformed cancer research, and this drug’s power was a testament to the promise of this new direction. “It’s very difficult to describe what a different world this is than just five to ten years ago,” said Klausner. In fact, he emphasized, this new, targeted drug paradigm applied to all diseases, not just cancer.
The Philadelphia Chromosome Page 25
