The Philadelphia Chromosome

by Jessica Wapner

  In October 1999, it was time to send the message to Novartis. Meanwhile, Druker was preparing for that year’s ASH meeting, at which he would present the results thus far from the phase I study. The first few months’ worth of data had been presented at the 1998 ASH meeting, but now the trial had been going on for more than a year. When the investigators submitted the study for consideration back in August, they had six months of data at the 300-milligram dose level. The responses were so impressive that the ASH advisory board invited Druker to present the study at the meeting’s plenary session. It was the highest honor the organization could give. Four years earlier, Druker had presented the preclinical study to an audience of about fifty, hoping someone would take interest. Now he would be speaking to an audience of around 20,000 clinicians from around the world. Every major news outlet was going to cover his report.

  The CML patients clamoring for access to STI-571 also knew about the meeting. They knew it would be the perfect place to call out Novartis on how long it was taking to make this lifesaving drug more widely available. If they put out a press release during the ASH meeting, Novartis would come under harsh scrutiny and would have no choice but to provide the drug as fast as possible. A petition sent only to Novartis, without a press release, might not raise the stakes enough. They asked Druker, whom they loved and to whom they would always defer, to weigh in.

  Druker was conflicted. He knew that if the petition was made public, the message surrounding STI-571 at the ASH meeting would be about the drug shortage. But this meeting would be the first time that any results from the drug trial would officially be made public. Shouldn’t the message be about this unprecedented breakthrough? He advised the patients against the press release.

  His counsel was also strategic. The petition gave him some much-needed leverage with the company. He alerted the executives about the letter they’d be receiving from patients. He also emphasized that he needed an announcement about the phase II trial from Novartis. “If I’m going to give a presentation, there has to be a clinical trial these patients can get into,” he explained. He turned their attention to the media nightmare that would ensue if a trial were not scheduled.

  Novartis got the message. On November 2, 1999—McNamara’s thirty-third birthday—she received a call from Druker. “Suzan, I just got off the phone with Novartis,” he told her. “Your petition did a world of good.” The company had agreed to speed up production and open a trial in the next month or so. A week later, she received a letter from James Shannon, a representative for the company, confirming the start of the next study. “An international multicenter phase II study is planned that will open for enrollment in January 2000, if not sooner,” Shannon stated. “Currently limited by availability of drug supply, Novartis has devoted substantial attention to making sufficient quantities of the agent available as soon as possible.” People with CML had rescued a drug that would, they hoped, rescue them.

  According to Vasella, the company had already been ramping up production of STI-571 for the phase II trial of chronic CML. There was no way the company would not move ahead with a drug that was potentially valuable. “Would you stand in front [of people] and say, ‘We have a potentially lifesaving drug, and I’m not going forward?’” he asked, the notion obviously ridiculous. He knew there were patients who might benefit from the drug. He also knew that in the business of developing new drugs, “more are failures than successes,” he said; risk of failure is part and parcel with working in the pharmaceutical industry.

  As Vasella told the story some years later, the company had always planned to produce STI-571. McNamara’s petition had only given the company a nudge to speed things up a bit. The availability lagged only because the trials were moving so quickly, not because the company was moving slowly or shelving STI-571 for good.

  “Dr. Druker will stand by the fact that it was file 13, never to be seen again,” said McNamara. “For me, I’ll never know.”

  29

  _______

  A RESPONSE RATE OF ONE HUNDRED PERCENT

  A few weeks later, Druker stood on a podium at the ASH plenary session. Having seen the preliminary data a year earlier and caught wind of the results that Druker would now present, the news media had demanded that ASH release the data from its embargo so as to allow newspapers and television to report on the study on a Friday, two days before Druker’s presentation. By the time that session came, the drug was already making headlines in major US newspapers, and all the participants were talking about tyrosine kinase inhibition. Under the spotlights of the New Orleans Convention Center’s cavernous hall, row upon row of chairs filled by cancer and blood doctors and researchers, Druker presented the slides he’d prepared for the occasion, his midwestern drawl adding to his understated manner. He didn’t need to say much; the numbers spoke for themselves.

  Druker took the audience through the study. All of the enrolled patients were in the chronic stage of CML, with fewer than 15 percent blast cells. All had stopped responding to interferon, which meant they’d had no significant change in their blood after three months of treatment, had no changes in the number of cells containing the Philadelphia chromosome after a year, had lost whatever response they’d been having, or were intolerant. All of the enrolled patients had ceased treatment for at least one week before starting STI-571.

  The trial, reported Druker, was a standard dose-escalation study. Eleven dose levels had been tried, starting at 25 milligrams and reaching 600 milligrams by the time of the ASH meeting. The average age of the sixty-one patients who had entered the study was 57 years. They’d been taking STI-571 for an average of 190 days. The duration of their disease was variable. Some had received the diagnosis less than a year before entering the study. At least one had had CML for more than thirteen years.

  Druker addressed the toxicities first. “No dose-limiting toxicity has been encountered,” his slides read. A few patients had mild myelosuppression, a decrease in activity inside the bone marrow that resulted in diminished production of red and white blood cells. Forty percent had experienced low-grade nausea, a problem that faded as the body grew accustomed to the foreign substance. There had been a bit of muscle cramping and a bit of swelling—mostly puffy eyes—but that was it, he told the crowd.

  Druker pressed the button on the remote, and the slide projected on the screen behind him changed. Hematologic Responses. The chart showed how many such responses had occurred at each dose level. At the 25- and 50-milligram levels, no changes had been observed. At 85 milligrams, one out of four patients had a response. At 140 milligrams and higher, 100 percent of the patients—all of the fifty-one patients who had come through the entire study so far—had experienced a hematologic response. The audience gasped.

  Druker pressed the button again. Complete Hematologic Responses. Everyone’s eyes floated to the bottom of the giant screen behind him. Among the patients who’d received doses of 300 milligrams or higher, every single one had had a complete hematologic response. Their blood had reverted to a completely normal state.

  Graphs charting the changes in white blood cell counts showed how the levels flatlined as the study progressed, plummeting from 20,000 or higher to well below 10,000. One patient whose counts had been nearly 100,000 at the start of treatment was now at about 6,000. The significance was obvious. Everyone in the audience knew that anything below 10,000 was considered normal.

  He pressed the remote again: Cytogenetic Responses. At five months, 45 percent of patients—nine out of twenty receiving 300 milligrams or higher—had a cytogenetic response, ranging from minor to major to complete. The number of white blood cells with the Philadelphia chromosome had vastly diminished, in some cases disappearing entirely.

  Druker showed a slide summarizing the data, and a last one to acknowledge many of the people behind the work, those who’d brought the drug to this very moment: Elisabeth Buchdunger, Jürg Zimmermann, Alex Matter, Nick Lydon, the members of his lab at OHSU, Grover Bagby, John Goldman, Charles Sawyers
, Moshe Talpaz, Tom Roberts, and Jim Griffin. Many of them were among the transfixed crowd.

  Druker thanked the audience for listening, and the talk ended. The room erupted in applause.

  30

  _______

  GOOD STRESSFUL

  Once Novartis made the decision to launch phase II, the mood at the company surrounding STI-571 changed completely. The stalling had been replaced with a blank check and a cleared schedule. “When Novartis got behind this project, [it] was a high-speed train and you just got out of the way,” said Druker, who credits Vasella with the company’s turnaround. No expense was spared to produce sufficient amounts of the drug as quickly as possible. Production of the drug was moved from Basel, where only small batches of STI-571 could be made at a time, to its manufacturing plant in Ringaskiddy, a village in County Cork, in southern Ireland. Ringaskiddy had only been used to produce already approved drugs, never anything in the pipeline. To manufacture hundreds of kilograms of a completely new compound as quickly as possible, the plant was pushed to its maximum capacity. The team there was working twenty-four hours a day, seven days a week.

  As Druker had predicted, the presentation of the phase I results at the 1999 ASH meeting had created an enormous expectation. Suddenly, “everything has to be done by tomorrow,” recalled Renaud Capdeville, a pediatric hematologist from France who’d joined Novartis in 1997 and taken over command of the STI-571 clinical trials in 2000.

  The decision to expedite production of STI-571 was driven by the need to open the trials to as many patients as possible. But once the company had committed itself to developing the drug, another pressing need quickly followed: getting it reviewed by the FDA as soon as possible. Although it does not regulate how doctors use the medications it approves, the FDA does regulate how products are labeled and sold. An FDA approval grants a pharmaceutical company permission to advertise a drug for the indication specified in that approval. Pharmaceutical companies cannot legally market their products until they are approved by the agency. All of the data from the clinical trials of STI-571 would have to be sent to the FDA’s Center for Drug Evaluation and Research for review. In the late 1990s and early 2000s, the FDA had been accused of taking too long to review beneficial new medications, and Novartis knew that the road to approval could be a long one. Clinical trials were all about investment, and these studies were costing the company hundreds of millions of dollars. Not a single dollar could be earned from STI-571 until that approval was granted.

  Despite the company’s commitment to developing STI-571, the rarity of CML was still an issue. STI-571 would be a patented drug, and eventually that patent would expire. When the patent expired, so would the profits. “When you lose patent, basically 90 percent of the volume is gone within a few days,” said Vasella. “It just evaporates.” Once the patent expired, generic manufacturers could use the drug formula to make their own capsules. With STI-571, the patent pressure was intensified because the number of prescriptions during its proprietary years would be fewer than usual. The rarity of CML meant the volume of sales would be far lower than for the average medication coming through the clinical trial system. STI-571 wasn’t even approved yet, but there was no time to wait on strategizing how to wring the most money out of it. The first part of the plan was to get the drug approved as soon as possible, so that Novartis could maximize sales while it had exclusive rights to the drug formula.

  STI-571 wasn’t the first such drug to come through the trials system, and the FDA had mechanisms in place for drugs that addressed an unmet need, as STI-571 did. Before the phase II trial had launched, the FDA had granted STI-571 fast-track designation and accelerated approval status. Normally, after the small phase I and larger phase II studies, a new drug must go through a phase III trial in which it is directly compared to the current best treatment option before the FDA would consider approval. For STI-571, that would have meant a large study in which patients were randomly assigned to treatment with either the experimental drug or interferon plus ara-C, and their responses would be compared after several years. The data were too promising to wait that long. The current best treatment was not all that great, and holding up approval of STI-571 until completion of a phase III study would have been unethical.

  The fast-track designation gave the company enhanced access to the FDA—more meetings, more correspondence, more assistance—to expedite the clinical trials process. Because STI-571 gave CML patients a new and necessary treatment option, the FDA was willing to take the time to give Novartis careful guidance about exactly what would be needed to prepare for approval review.

  Accelerated approval meant the company could submit its phase II data for FDA review, with the agreement that a phase III study would be completed if the drug was approved. Because the phase I and II trials are short, single-arm studies—that is, all patients received the same treatment, rather than being randomized to one of two arms, as they would be in phase III—the results don’t hold as much clout. But the advantage of submitting phase II data as opposed to phase III data was quite significant. It shaved years off of the development time. With accelerated approval status, the FDA would review STI-571 based on the surrogate end points of hematologic and cytogenetic response. Changes in blood counts and the amount of cells with the Philadelphia chromosome didn’t prove that the drug worked, but they did point strongly in that direction. Accelerated approval status meant those measures, alongside the data on side effects, were sufficient for an FDA review. Conditional approval granted all the same marketing freedoms as a standard approval.

  Even with this special status, Novartis still needed a year’s worth of phase II data before submitting the drug for approval. Novartis—in particular, Jörg Reinhardt, the global head of development—was pushing to file the data with the FDA by December 2000, just two and a half years after Bud Romine took his first 25-milligram pill. That timetable was at least a year less than average for drug development. The pills would need to be manufactured quickly.

  STI-571 is created through an intensive, multistep process. Following the intricate cascade of chemical reactions that Jürg Zimmermann had plotted out when he had created the compound years earlier, the raw materials are mixed in solvent, heated, evaporated, condensed, filtered, and put through several other exacting chemical reactions out of which the powder eventually precipitates. Then the powder has to be milled into pills, which also takes time. As with any serious medication, STI-571 had to be created under the strictest quality control, with extra precautions—protective suits and a plantwide containment system—to prevent exposure to hazardous chemicals used to make the drug.

  With instructions for making the pills translated from German, the manufacturing team at the 60,000-square-foot plant in Ringaskiddy was ready to begin production. By November, the first seven of the twelve steps required to produce STI-571 were complete. For the remaining five steps, which would normally take a year, Novartis gave the plant seven months. The company wanted to have a metric ton of material by the summer of 2000. To produce that volume, the Ringaskiddy plant had to be operating twenty-four hours a day, seven days a week, its entire staff put on the job.

  The phase II trials for blast crisis patients opened in August 1999, and continued accruing patients until March 2000. The phase II study for patients with accelerated disease, those with slightly longer to live, opened the same month and was full by April 2000. The phase II trial for chronic-stage CML patients began enrolling in December 1999.

  By the end of 1999, trial sites had opened at Hammersmith Hospital, in London, with John Goldman as the lead investigator; at Druker’s former workplace, Dana-Farber Cancer Institute; and at cancer centers in Detroit, Miami, and New York. By early 2000, nineteen sites were up and running. In early January, Michael O’Dwyer, an Irish doctor who’d come to OHSU to study hematology, joined Druker’s group to help with the ongoing clinical trials there and Druker’s continued research of CML and tyrosine kinase inhibition.

  On January 1,
2000, Suzan McNamara, the woman behind the Internet petition, flew to Portland, Oregon. As soon as she arrived, she went to OHSU for preliminary testing. She could barely walk up the stairs. She was 5’ 8” tall and weighed 110 pounds. After reviewing her test results, the staff told her that they thought she was just at the borderline of accelerated stage, no longer in the early, chronic phase of the disease. The thought panicked her because it meant that, even though she could enroll in the trial, the disease might already be too advanced for the drug to have any long-lasting benefit. When she returned to the hotel, her phone was blinking with a message. It was Druker. “Suzan, I don’t want you to worry,” he told her. “Your tests came out fine. We’re getting you on the trial.”

  Within a week, after suffering through a brief but sharp bout of leg pain, McNamara was already feeling better. She stayed in Portland for five weeks with her boyfriend. “We had the time of our lives,” she says. “I felt reborn.” Among her best memories of those weeks is all the eating she did. For the first time in two years, she could eat whatever she wanted. By the time she left Portland, she had gained twenty pounds.

  But even with the company’s full commitment, the drug supply still fell short. For the phase II study of chronic-phase patients, which had opened in December 1999, the shortage was all too real. Each trial site was allowed to enroll ten eligible patients per month. But Druker already had a waiting list of 200 patients when the study first opened. How could he, or any of the other investigators, who also had waiting lists, choose which ten to enroll in a given month? “I was having to prioritize patients and make decisions about who should get this drug and who shouldn’t,” said Druker. People were showing up at OHSU to demand access. He never turned anyone away from the study; he just added names to the list and tried to not have anyone wait longer than three or four months. For some, that wait would be too long, and they would die of the disease. As usual, to relieve himself from the physical and mental stress, Druker ran. In August 1999, it was the Hood to Coast relay, stretching from Mount Hood to the Pacific coast beach. In October 1999, it was the Portland Marathon. The rest of the time, it was the steep hill that led to his clinic.