The Pill That Could · First in Human

First in Human, part 3

But it was already, by mid-1999, an open secret in the small world of CML research. Investigators at the trial sites were telling colleagues. Patients on the trial, hearing what kind of responses they were having, were telling other patients. Hematologists not on the trial were calling Druker and Sawyers and Talpaz to ask whether they had any open slots and how their patients might be enrolled.

By the summer of 1999 there was, in the CML community, a wave of demand for the drug that the company was struggling to keep up with. Manufacturing had been scaled up to support a phase 1 trial of perhaps fifty patients; it was not anywhere near what would be needed if even a small fraction of the CML patients in the United States wanted access. The company was also, reasonably, hesitant to expand access ahead of trial results that hadn't yet been formally analyzed. The position was the conventional one for an early-stage compound: trust the trial process, finish phase 1, plan phase 2 carefully, expand access in the standard way.

From the outside, this looked unbearable. Patients with CML, watching their counts rise on inadequate treatment, hearing about a drug that was producing complete responses in people on a trial they couldn't enroll in, started to push back.

* * *

There was, in 1999, a CML patient living in Quebec named Suzan McNamara. She was thirty years old. She had been diagnosed two years earlier, had failed interferon, was running out of options, and had heard about the trial. She wanted in. She was, like many CML patients of her generation, an active member of online patient communities, where word of the trial's results had been spreading rapidly. She had, after several attempts to enroll directly, been told that the trial was not currently accepting patients with her particular profile and that she should wait for phase 2.

Wait for phase 2 was, in McNamara's case and in many similar cases, an answer that did not engage with what waiting actually meant. CML in chronic phase, untreated or inadequately treated, progresses. The chronic phase ends. The accelerated phase begins. After that, blast crisis. Once a patient is in blast crisis, the drug, if it ever becomes available, is much less likely to work, because the disease has acquired additional genetic changes that are harder to reverse. From the patient's perspective, every month of waiting is a month closer to the point where the drug, if it were given, might no longer help.

McNamara wrote a letter. The letter was directed to Daniel Vasella, the CEO of Novartis. It explained, in plain and unadorned terms, what waiting meant. It asked him to expand access to the drug. She also organized, with other patients she had connected with online, a petition asking Novartis to manufacture more of the drug and make it available to patients outside the trial. The petition collected, by various accounts, several thousand signatures.

Vasella has said in subsequent interviews that he read the letter, and that the patient pressure mattered to him. Other Novartis executives have said similar things. Whether it was the deciding factor in any particular decision is hard to know from the outside, but the company did, over the course of 1999 and 2000, dramatically expand its manufacturing of STI-571 and its plans for further trials. Phase 2 trials were designed and launched in patients across the spectrum of CML disease stages: chronic phase, accelerated phase, blast crisis. Production capacity was scaled up to support a level of demand that, even at a year or two earlier, would not have been considered.

Some of this was patient pressure. Some of it was the trial data itself, which was getting harder to ignore as it kept coming in. Some of it was the people inside Novartis who had been pushing for the program all along. The combination, by late 1999 and through 2000, had moved STI-571 from a small-scale phase 1 study into something that looked very much like a serious development effort.

* * *

There was a moment in this period — worth noting because it captures something about how the trial felt at the time, not just to investigators but to the patients in it — that became, in retrospect, a small fixture of the story.

Druker had a patient named Bud Romine, an Oregon truck driver who had enrolled in the trial in 1998 with end-stage CML. Romine's CML had progressed, by the time he entered the trial, past chronic phase. He had been told, before the trial, that he probably had months. He started on STI-571 at one of the higher dose cohorts. His blood counts came down. His bone marrow improved. He felt, after a few months on the drug, well enough to go back to driving. He did. He drove a truck across the western United States for years on STI-571. He brought, periodically, postcards back to Druker's office from places he had been able to go.

Romine was one patient. The trial had dozens. Most of them did not write postcards, but most of them, at the higher doses, were having responses that fell somewhere on the spectrum from very good to remarkable. By the end of 1999, the senior investigators were sitting on a phase 1 data set with a complete hematologic response rate, in chronic-phase CML patients at therapeutic doses, of essentially one hundred percent.

All 13 chapters of The Pill That Could · Library · All courses