Still, for the patients and clinicians, the outcomes had already far surpassed all reasonable hopes. Blood counts were being restored. The bad gene was disappearing. Patients were feeling better. It was time to open the phase II trial and to tell the world about this drug.
27
_______
BUZZ IN THE CHAT ROOMS
The phase I patients were increasingly aware that they were part of something special. In Oregon, they became friends and, in the summer of 1999, started a monthly gathering. They’d share a meal and hash over the details of their illness, how they had arrived at Druker’s doorstep, how they were feeling. They spoke a secret language, where words like “FISH” and “blood counts” carried the weight of their survival. They felt charged with responsibility, as if knowing the science were somehow their job now. They were ordinary people who’d been diagnosed with a blood cancer and had found themselves in a rare circumstance: the first initiates in an entirely new kind of medicine.
The chatter migrated online. In the late 1990s, Internet chat rooms were in their fledgling years. This kind of forum was perfect for the CML community, allowing its members to exchange information about the disease, about its treatment, about how to handle the various side effects. More experienced patients enjoyed being able to pass on their hard-earned wisdom. For many people, local cancer support groups included people battling all types of malignancies. There were rarely enough CML patients in a single county to form a local support group. Online, however, the conversation was confined to their disease only, and so the talk proved much more appealing and useful. People got to know one another through message-board postings, and they opened up about the emotional challenges of their disease and the rest of their lives, too. They were a support group of people who each knew exactly what the others were going through.
A message board started at Egroups.com by a man named Robert Neil, whose mother had been diagnosed with CML, was the earliest of the many incarnations of online CML forums. There were a few dozen active members, and many silent readers who soaked up the information. Among the group were a few people who were enrolled in the phase I trial. Naturally, they posted their test results and recounted conversations they’d had with their doctors. They talked about how they were feeling. Nearly everything they wrote was positive.
The message boards were open to anyone. The trial investigators, who were aware of the forums, suddenly found themselves in the middle of a new phenomenon: Internet sharing.
Each day, the investigators would return home after a long stretch at the clinic and find confidential conversations recounted verbatim online. The information gave hope to people with CML who were struggling with chemotherapy and interferon and had not known about the trial or been able to reach a study site. Because of the World Wide Web, many patients with CML now knew more than their doctors about this experimental new drug.
The positive reviews stirred up demand among those not enrolled. It wasn’t long before Druker, Sawyers, and Talpaz were hearing from hundreds of people from around the world who’d been diagnosed with CML. They’d read about the trial on the Internet, seen the incredible data, and they wanted in.
Having had minimal, if any, exposure to web-educated patients, the investigators had no plan in place for how to respond. Should they water down their messages in the clinic, knowing that whatever they said to their patients was going to be broadcast far and wide? That might be wise. But the drug was working. Why shouldn’t everyone know about it? Plus, censoring the conversations in the clinic wasn’t their style. All along, Druker had made himself available for any questions, and that openness was inherent to the intimate nature that his phase I group had developed.
They also wondered if the publicity might serve another purpose. It was time to ready everyone for the next clinical trial, and Novartis was not moving quickly enough. Maybe the freewheeling chatter about this new drug that was melting cancer away would give the company a kick. “It was important . . . for Novartis to feel the heat, to get their shit together, and start really ramping up the project,” said Sawyers.
The plan, as Sawyers knew, was to open three separate phase II studies: one for chronic stage, one for accelerated stage, and one for blast stage. As with the phase I study, these trials were restricted to patients who had already tried interferon, the medication either proving intolerable or ceasing to work. Cancer drugs are rarely approved for all stages of cancer at once; rather, new medicines are approved for each specific population in which the drug was studied. When the FDA first approved the drug—if the FDA approved it—the first indication would be for patients for whom the current best treatment option failed. These three phase II trials, which would enroll a thousand patients altogether, would provide the evidence for those approvals. After that would come phase III, a large and lengthy study in which newly diagnosed patients would be randomly assigned to treatment with either STI-571 or interferon plus ara-C. Only at phase III would the drug be tested on patients who’d never been on any other CML medication.
But there was a problem. According to Novartis executives, there wasn’t yet enough of the drug to start a phase II study. And there was no telling when there would be enough. Sawyers wondered if the message board chatter might stir the publicity-fearing pot at the corporation.
Every day, there were new posts from the members lucky enough to be in the phase I trial. Every day, there were calls to OHSU, UCLA, and MD Anderson from patients interested in enrolling. And every day, there was nothing new to tell them. The trials were full. There was no more drug. They would have to wait.
28
_______
SAVING THEIR OWN LIVES
Because the transition from the phase I study to the phase II study was so quick, Novartis had not had time to produce enough STI-571 by the time the investigators wanted to start enrollment. A few kilograms of STI-571 had been manufactured for the phase I study. Hundreds would be needed just to launch phase II. The clinicians and the industry representatives had talked about the phase II study since February 1999, but it wasn’t until the summer that the company realized the shortage it faced. “They hadn’t planned for success, they had planned for failure,” said Druker. That mind-set had left the company unprepared for the next steps.
It wasn’t just that the phase I study had been designed to give a quick readout about whether the drug worked; many people had assumed it wouldn’t work. The assumption was understandable. Fewer than 10 percent of experimental compounds tested in phase I studies are ultimately approved by the FDA. This compound had killed dogs in the toxicity studies. Targeting a kinase was an untested strategy. While some people at the company pushed to move ahead, others dampened the enthusiasm by pointing again to the reality of the market. The sales projections were dim. Betting on failure was much more reasonable than betting on success. By the summer of 1999, those miscalculations had caught up with the company and were delaying the delivery of a breakthrough treatment to patients who needed it.
At a meeting in Bordeaux, France, in the summer of 1999, Druker discussed some of the data from the phase I test so far. John Goldman, the British leukemia expert who’d been excluded from the first trial, was in the audience. Stunned by what he heard, Goldman asked John Ford when he could get the compound to start a trial with his patients. “Well,” Goldman later recalled being told, “Novartis isn’t planning to develop it. Maybe in a couple of years.” Goldman wrote a letter to Daniel Vasella, then the CEO of Novartis, urging him to move forward, but he never heard back.
Druker was doing all he could to speed up the process. His only contact at Novartis was Ford, who was bound by what the company was telling him.
“Who can we talk to, to try to accelerate this?” Druker asked him.
“Me,” said Ford.
“Who do you work for? Can we talk to them?” Druker pushed back.
“No,” he recalled Ford replying.
“But you’re not fixing this drug supply shortage.”
> There was nothing Druker could do. Part of him understood. “Coming into this new way of treating cancer, a new category of drug, a small market projection, who in their right mind would put a lot of time and effort into that project?” he said later. But with the phase I data now in hand, it was time to change that thinking. “We were moving at warp speed, and we needed a champion inside Novartis.”
Druker was entirely frustrated. There were no more phone calls he could make. There was no one he could talk to who could influence Novartis to get moving. All he could do was stare at the growing list of patients who wanted the drug.
And run. He joined marathons. He ran to work, up the steep hill to OHSU. After long hours at the clinic, he ran home. It was as if the speed he longed to see in the clinic was demanding another outlet. So he ran.
Vasella tells the tale differently. “There was no drug shortage,” he said. And the situation was not unusual for an experimental drug. “You never have that,” Vasella said of the need for an ample quantity at the very start of phase II. The only difference with STI-571 was that the need had arisen much sooner than usual, leaving the company with an “unexpectedly high demand.” The phase II trial was being set in motion unusually fast, before the phase I had even been completed. That timetable was unforeseen and hadn’t left the company with enough time to make the necessary amount of the drug.
Vasella acknowledges the continued resistance within the company, but also how important it was to him to take the risk. Jörg Reinhardt, who was the head of the development team at the time, was insisting on proceeding with caution, telling Vasella about not having this or that bit of data or the manufacturing capacity. But having seen the phase I results, Vasella wanted to plow on. “Come on now,” he told Reinhardt, “move ahead.” When Reinhardt expressed concern about the cost, Vasella dismissed it as unimportant. “The problem is somebody has to be willing to take a risk and accept that he might get the blame for failing,” Vasella said later. As he tells it, after his initial intervention that moved the drug from the preclinical phase into clinical trials, Vasella continued to be its champion. Though uninvolved in managing the details of what it took to manufacture sufficient amounts of the drug, his willingness to take the risk of developing STI-571 provided cover to those upon whom those responsibilities fell.
In August 1999, the phase II trials for accelerated and blast crisis patients opened. Novartis had managed to produce enough drug for that population. It would have been considered highly unethical to withhold a drug from those patients if it clearly could be made in some quantity. Patients with accelerated or blast crisis CML could enroll in the phase II trials without restriction.
But the next trial for chronic-stage disease remained on hold. Patients in this early phase of the disease worried that the leukemia would progress before they began treatment. Even though drug access for later-stage patients was guaranteed, no one wanted to wait until they got to that stage because it might be too late. They wanted the medication when it could do the most good. “Their view,” Druker said by e-mail, “[was] that Novartis had essentially crafted a death-bed strategy.” That the company had launched phase II studies for blast crisis and accelerated disease was evidence of moving forward with development, but chronic-stage patients stuck on interferon wanted what they knew was a superior treatment as soon as possible.
ONE OF THE patients on Druker’s list of those wanting access to STI-571 was Suzan McNamara. In March 1998, when she was 31 years old, she began experiencing bone pain and rapidly lost 15 pounds. A blood test revealed her white blood cell count to be 380,000. She had CML. Petrified by the prospect of a bone marrow transplant, she opted for interferon. The treatment worked for a while, lowering her white blood cell count, but eventually she had to quit her job because she felt too sick to work. Each day, she injected herself with interferon and ara-C, and each day she felt worse. She was depressed, she couldn’t sleep, and her hair was coming out. She could barely eat, and she continued losing weight.
McNamara had stumbled across the Egroup.com forum during an online search for information about CML. She read the posts by the few patients on the phase I trial for STI-571. Their test results were encouraging, and the patients reported that they felt good. McNamara, whose life by then revolved around her bed and the computer two feet away, seized on those words.
The trial patients often talked about their progress in meticulous detail. People on the forum who’d attended meetings at which the investigators discussed the drug reported back what they’d heard. Everyone spoke positively. McNamara asked her doctors about enrolling in the trial, but they discouraged her. A million drugs went through phase I tests, they told her. Why would this one be any different? Still, she could not ignore what she was reading. All of the non-trial CML patients reading those posts wanted in, and they all knew they had to wait until they relapsed. Some got creative, finding ways to expedite the process—say, insisting that they could not stand interferon’s side effects anymore—even before their white blood cell counts began their inevitable ascent. They knew the criteria that would make them eligible for the trial, so they made themselves eligible.
For McNamara, relapse was all too real. In October 1999, her treatment stopped working. About six months after starting treatment, her platelet level fell dangerously low. A lower dose of interferon and ara-C didn’t work. She knew that it was only a matter of time before she entered the accelerated stage of CML. Now approaching 33, she could be dead in a year.
She took printouts of the online reports from the phase I trial patients to her doctors. Finally, they agreed to call. There was no good news. The trial was full and closed to further enrollment.
Out of sheer desperation, McNamara sent an e-mail to Druker telling him her story. He replied within eight hours. He was sympathetic, but he told her there was not enough drug. It would be four or five months before she could enroll in the next trial, but Druker wasn’t even certain that would happen. “We are fearing it might close down eventually,” McNamara recalled him as saying. “Maybe as a patient you can do something.”
The moment sent a shaft of light down to McNamara, stuck in the depths of her illness. His words “lit a fire in me,” she said. “There is no way that there is a drug out there that can help me that’s not going to be given to me.”
She lay in her bed trying to think of what she could do, something big that had a real shot at working. She remembered something she’d seen on the Internet and wondered if it might work: an online petition.
At the time, the idea was still fairly novel. Although accessing the web from home was not uncommon in the late 1990s, many of the innovations that people now take for granted had not yet evolved. Not every advertisement included URLs. Not everything could be found online, and websites certainly had not become the vehicle for patient advocacy that they are today. The Internet had not taken off as a social medium, and its potential power to generate social change was unforeseen. Yes, the trial patients were posting all the minutiae of their experiences with STI-571, disclosing intimate details about their health and personal lives. But their readers were not mobilizing online to get themselves the drug. Forging a new kind of online alliance wasn’t their intention. Although most of them had witnessed the AIDS crisis during the 1980s and seen the power that patients could have when they banded together, this cancer was different. It was a rare disease, not an epidemic, and there was no social charge behind the cause. They weren’t angry. They just wanted the drug.
The online CML community readily embraced McNamara’s idea. The petition, which called for the manufacture of STI-571 to be sped up, was posted online on September 21, 1999. Within three weeks, 3,300 people had signed it, and the signatures continued to amass. Patients, caregivers, friends, and friends of caregivers had all logged on to this call to Novartis to make more STI-571 and get it to the patients who needed it. Members of the Egroup message board wrote a letter to accompany the petition and addressed it to Daniel Vasella, then the ch
airman and CEO of Novartis. The letter read:
Through extensive interaction on the internet, a large number of CML (Chronic Myelogenous Leukemia) patients are well-informed about the results of the Phase I and Phase II trials of the new drug STI-571 which is produced by Novartis. The sources of information are varied and include published papers on the drug, papers given at professional meetings, and the direct experience and knowledge of a significant number of patients who are taking part in the trials. . . .
As you know, the drug has shown no significant toxicity to date, and the results, in terms of both hematological responses and cytogenetic responses, have been more than impressive at this early stage. . . . There is also an acute awareness that the drug is in clinical trials and a realization that there is a need to be cautious at this relatively early stage. . . .
Nevertheless, many of us who have signed the petition believe that the theory behind this new drug is fundamentally sound. . . . It is not impossible, based on results to date, that the new drug will prove to be a functional cure for some patients. . . .
Because of the particularly good prospects for this new drug, we have viewed with growing concern our belief (based on information from various sources) that the supply of the drug has not been sufficient to expand the trials as fast as the evidence to date would warrant. . . . There are many CML patients undergoing difficult treatments, in some cases with considerable suffering, and where the outcomes are not at all assured. . . .
We therefore ask for your assurance that everything will be done to produce a sufficient supply of STI-571 to ensure that the trial investigators are not held up in any way at all in trialing this new drug, and in advancing to the certification that we anticipate.
The Philadelphia Chromosome Page 22