The Philadelphia Chromosome

by Jessica Wapner

  He was baffled that Ciba-Geigy, upon seeing the results, was not plowing ahead. The company knew that the next step was to prepare for clinical trials by testing the drug on animals. It would need sufficient amounts of toxicology data from animals if it was to ask the FDA’s permission to study the drug in humans. That was the routine process for so-called investigational new drugs, one in which the company was well versed.

  “Ciba was being odd about all this,” Druker said. He knew that Lydon shared the same frustration, and he tried to be patient. He recognized that some progress was being made, but he also knew there was no time to waste. “It just seemed like, can’t you just do this right now? Can’t we start our clinical trial next month?” Druker remembered thinking.

  The company also knew the signs of a promising drug candidate. CGP-57148B had them all—almost. There was one irreconcilable problem: the limited market. There were too few patients with CML to make developing this new drug worthwhile for the company. To make the investment required to move the drug from animal testing through clinical trials and then into distribution and marketing if the drug was approved by the FDA (a costly process in itself), a pharmaceutical company needed to be assured of profits typically in the hundreds of millions. No way would a drug indicated for the rarely occuring CML bring in that level of sales. There was no way around that fact. The potential efficacy of the drug was intriguing, but it wasn’t enough to sway the company away from its concern about the bottom line.

  Plus, the mission of the company was to provide medical advancements to the people who needed them. If the company was going to invest its resources in cancer, didn’t it make sense to focus on the most widespread types with the highest incidence so that the most people possible could benefit from the research?

  “There was a lot of resistance in the Ciba-Geigy marketing department that CML was too small an indication [for] trials,” said Lydon. His group was being told to go for PDGFR as the first clinical entry point. “[CGP-57148B] was always the stepchild,” says Druker. “It was not something that [the] company was pushing fast to get into the clinic.”

  Fortunately, another insight from Druker’s preclinical study added ammunition to the plea for a clinical trial. Early on, when he and others were still just hypothesizing and daydreaming about targeting kinases, Druker had thought that such a drug might ultimately be given by treating the bone marrow. The concept was that the bone marrow of a person with CML would be removed and incubated with the drug. The cancer cells in the marrow would die, and the normal cells could then expand. This cleaned marrow would then be transplanted back into the patient.

  For that reason, Ciba-Geigy had sponsored Druker’s lab to test out a second approach: mixing the compound right into the marrow as a way to stop the leukemia. This strategy was why he’d obtained marrow samples to test along with the mouse and human cell lines. Druker had tested both samples—those free of CML and thus the Bcr/Abl kinase, and those from CML patients known to have the Philadelphia chromosome—with the experimental compounds sent by Lydon.

  plate 13

  For these tests, the cells were brought to the cytogenetics lab for analysis. A new microscopy technique known as FISH, short for fluorescence in situ hybridization, had just become available. By coaxing specific areas of DNA into glowing one color or another when viewed through a fluorescence microscope, FISH allowed geneticists to obtain a highly accurate count of cells with a particular gene sequence, or rearrangement of a gene sequence. With FISH, bcr glows green and abl glow red. In a normal cell, the red and green dots will be spaced far apart, signifying their location on separate chromosomes. In CML cells, the red and green dots of the rearranged chromosomes are adjacent, brought together through the translocation of material between chromosomes 9 and 22 that defined the Philadelphia chromosome abnormality. When next to each other, the red and green dots often appear yellow, a result of the properties of the fluorescent light being emitted. Looking through a microscope, a researcher could count exactly how many cells in a particular sample were positive for the Philadelphia chromosome by finding those yellow spots. The invention was incredibly well timed for studying CGP-57148B because it enabled Druker to know the exact extent to which CGP-57148B was affecting the cancer cells. If the compound was doing anything, then the count of cells with red and green dots lumped together would be lower.

  Without divulging too much information, Druker would carry vials down to the cytogenetics lab in a circular Styrofoam cooler, their plastic pop-top caps concealing an undisclosed solution mixed with thick clusters of bone marrow. There, Helen Lawce, the resident chromosome expert who’d worked briefly with Joe Hin Tjio (the Indonesian scientist who’d provided the first accurate count of human chromosomes), exposed the cells to fluorescing DNA probes and examined them under the microscope. Her white hair gathered in a ponytail flowing down her back, Lawce sat in a darkened room adjusting the focus on the giant fluorescence microscope, the folk music she preferred playing quietly in the background.

  In the non-CML samples, the compound did nothing. But when Lawce looked at the samples that had come from CML patients, she was perplexed. She called Dr. Druker to explain her confusion. “I can’t tell you how many of these cells are Philadelphia positive because they’re all dead,” she told him. Druker was elated. Lawce didn’t know what to make of his response. “What’s that about?” she thought to herself. “I can kill cells with Clorox.” Druker was intent on keeping his work under wraps, so for Lawce, his delight remained a mystery. She couldn’t understand why it was a good thing that all the cells had died.

  But Druker knew their death meant the compound had worked. In this set of experiments, up to 80 percent of the bone marrow cells from CML patients were destroyed by exposure to CGP-57148B. As the remaining cells repopulated, those that did so in the absence of the compound contained the bcr/abl gene. Among those remaining exposed to CGP-57148B, fewer than 20 percent contained the mutant gene. Druker also noted that in the marrow sample that came from a CML patient who did not have the Philadelphia chromosome—a rare but possible occurrence—the compound did not impede the growth of the cells.

  The marrow data showed the team at Ciba-Geigy two things. First, it revealed that the marrow-cleaning approach, known as purging, wouldn’t work because the normal cells would not expand back to a safe quantity. But also, the work proved without a doubt that the compound killed CML cells and spared normal cells. All of the other test tubes provided solid evidence of the compound’s potential. But bone marrow from real CML patients was the closest thing to actual patients. If the compound worked with marrow, odds were it would work with people.

  “For us, this was a very convincing piece of data that pushed CML to the forefront,” said Lydon. He, too, had been frustrated by Ciba-Geigy’s hesitation about moving forward to the next phase of development. He was sympathetic, at least to an extent. “It was a commercial organization, and they had to make a drug that could sell,” he said. But he also knew that when it came time to pioneer an entirely new type of medicine, the smartest approach would be to test the drug in the area where it was most certain to succeed. “CML was that area,” said Lydon. With the marrow data, Lydon, Matter, and the rest of the research team had irrefutable proof that this compound deserved a clinical study. “[This] was the piece of data that we used to convince the organization that CML was the area to test this drug.”

  Finally, in 1995, the company agreed. That June, Ciba-Geigy held its first meeting to plan the clinical development of CGP-57148B. The meeting was attended by Lydon, Buchdunger, Druker, John Ford—the central medical adviser in the clinical development department at Ciba-Geigy—and a man named Alois Gratwohl, an outside consultant to John Ford, there to advise on the best course of clinical trial action. CGP-57148B was cleared to enter toxicology studies. A phase I clinical trial—the first stage of human testing for any new drug—was tentatively scheduled for November 1996, twelve years after Matter had first started the kinase inhibitor pr
ogram.

  22

  _______

  GAIN AND LOSS

  There were two tasks at hand once CGP-57148B was cleared for development. First, the compound had to be turned into a drug. Second, the drug had to be proved safe. Safety didn’t mean the drug would have no side effects, but the company had to be sure it wouldn’t kill people.

  Moving from the laboratory studies to this early development stage was a huge undertaking. It was one thing to have a molecule that worked inside a cell. It was quite another thing to turn that molecule into a palatable medication that worked inside a living, breathing human, particularly one who was sick with leukemia. The immune system isn’t the body’s only guardian against foreign substances. The digestive system is also geared to reject anything that has no benefit to the body. “The stomach is made to destroy chemicals,” noted Zimmermann.

  Finding a way to turn the compound into a drug that would not be immediately rejected or disintegrated by the body would not be easy. To enter the body, the molecule had to be stable in water without dissolving. To leave the body, the drug, having done its work, had to penetrate the portal vein, which goes through the liver, where the remnants could then be broken down for excretion. And these qualities had to be instilled in the molecule without disrupting its selectivity for the Abl kinase, the strength with which it bound its target, and its ability to kill cancer cells.

  Although the formulating work could be done in a laboratory, the next steps required giving the drug to animals. This was the domain of pharmacokinetics and pharmacodynamics. The development team had to know how it was absorbed into the bloodstream and distributed around the body. They had to know how the chemical would be metabolized by the body and eventually excreted. All of these factors were crucial to the compound’s viability, just as important as its ability to block the kinase. If the chemical could not be safely eliminated from the body, then that was a problem because it could accrue in the kidneys or the liver and become toxic. Charting all of the biochemical affects on the body, the good ones and any bad ones, was also essential. Did the drug raise blood pressure? Did it cause diarrhea? Did it cause excessive thirst? Such measurements could only be obtained by giving the drug to animals.

  Within Ciba-Geigy, each department had its responsibilities and its boundaries. The wall separating the discovery team, which created new drug candidates, and the development team, which took over once the lead candidate had emerged, was tall and thick. “Once you have given your candidate compound to development . . . you have no say and no power over this compound anymore,” said Matter. “You could talk and talk and talk, but they will do whatever they believe has to be done.” As soon as the candidate compound was handed over for toxicology testing, Matter and his group—Lydon, Zimmermann, Buchdunger, and the rest—had no control over what happened.

  From Matter’s point of view, that separation was problematic. The development team, he felt, considered the early researchers to be “just fools and clowns,” whereas they, the developers, had the real know-how about making new drugs. They also tended to be very conservative, understandably averse to taking risks with an unknown chemical. These two traits seemed to enhance one another, creating an endless loop of caution.

  That caution surfaced very soon after the drug was passed from discovery to development. Medications get into the body through a few different delivery systems. A drug can be made into an intravenous formulation, injected through a needle into a vein; a subcutaneous formulation, injected under the skin; or an oral formulation, a tablet or liquid given by mouth.

  Matter, Lydon, Druker, and the rest of the discovery team had been hoping to turn CGP-57148B into a pill. The preclinical observations of how the drug worked in cells indicated that the drug would need to be given fairly frequently to continuously shut down the cancer-driving kinase. In Druker’s earliest experiments, the kinase had to be shut down for sixteen hours at a time for the cancer cell to be killed, hinting that a daily dose might be required to treat CML in people. A pill would be far more convenient than an injection if the medicine had to be given every day. An infusion wasn’t out of the question—plenty of cancer drugs were given as daily injections for a week or two, and sometimes the injections took hours—but having a bottle of capsules that could be taken at home was far preferable.

  But Matter was quickly told by the development team that the pill approach wouldn’t work. “We got our first spurious data . . . telling us this compound could not be given orally [because] it didn’t have bioavailability,” he said. In other words, the formulation required to turn the compound into a pill would render the drug inaccessible by the body. An oral formulation, Matter was informed, was out of the question.

  Matter wasn’t sure that conclusion was accurate. In his estimation, the development team wasn’t giving the pill form a fair shot. He was pretty sure that the friction between the teams had somehow biased their result, but the tension also meant he had no way to contest it. He, Lydon, and Druker settled on an intravenous formulation instead. CGP-57148B would have to be turned into a substance that could be shot through a needle into the vein of a patient with CML. The drug would then circulate through the blood, entering cells and killing any that contained the Bcr/Abl tyrosine kinase. It wasn’t the best case scenario, but at least the drug was in development. That was what mattered.

  WHILE THE DEVELOPERS worked on the intravenous formulation, Druker prepared his first public presentation on the compound. He had just submitted the paper on the preclinical studies to a third journal, Nature Medicine, in mid-November, and was getting ready to speak at that year’s annual meeting of the American Society of Hematology (ASH), held in late 1995 in Seattle, Washington. It is the largest professional meeting in the world for doctors who treat diseases of the blood, including the three liquid cancers (leukemia, lymphoma, and myeloma), and conference presentations are often pivotal moments in the careers of cancer researchers, notches on the wall of accomplishment.

  “We have demonstrated specific killing of the Bcr-Abl expressing cells by CGP-57148B,” read the study abstract. Druker talked the small audience through a set of slides chronicling the dramatic effects the compound had had on Bcr/Abl-positive cells, and the lack of change among unexposed cells containing the mutant kinase. “This compound may be useful in the treatment of CML and other BCR-ABL positive leukemias,” the slides read. Druker was lead author on the presentation. Buchdunger, Lydon, and Grover Bagby, who’d recruited Druker to OHSU, were also credited.

  The presentation was attended by about fifty people and didn’t cause much of a stir. That was no surprise; cell-line studies are so preliminary that they are rarely the stuff of press releases and rapt crowds. A man named John Goldman was one of the few who approached Druker afterward. Goldman, an English oncologist who was famous for pioneering the use of bone marrow transplants for leukemia in Europe, had been alerted to Druker’s work by Ciba-Geigy. Intrigued, Goldman followed Druker back to Portland after the meeting, sleeping in the ballroom of a Holiday Inn because all the hotels were full. The next day, Druker agreed to give Goldman some of the compound to test back at Hammersmith Hospital in London. When experiments in his own lab reproduced Druker’s results exactly, Goldman knew this compound was special.

  Yet the attention of this highly regarded hematologist did little to calm Druker’s nerves. He knew the intravenous formulation had been made and that toxicology tests were finally beginning, and he was anxious for any news. He was also unsettled about the rumblings he was hearing from Ciba-Geigy about the clinical trial. Now, after the first encouraging planning meetings, he was being told that the company was considering conducting the initial human study solely at MD Anderson Cancer Center, in Houston, Texas. “They weren’t sure they wanted me to be involved,” says Druker. In part, he understood. Interferon, the standard treatment at the time for CML, had been developed largely at MD Anderson, with a leukemia expert named Moshe Talpaz at the helm, and patients with CML flocked there because it
was reputed to provide the best care. The top leukemia doctors in the world were there, and Ciba-Geigy wanted seasoned investigators for the phase I study. Druker had to admit that he was not the ideal person to lead a clinical trial. “How many CML patients do I have in my clinic? I have three.” Yet he had been part of this compound’s creation for years now, and he was one of the few medical oncologists in the world who had championed the idea of kinase inhibitors. “I had to fight for this, to make sure if we’re going to run a clinical trial, I was going to be part of it,” he said.

  On this point, Druker wrestled as much with his own reaction as he did with the company. After all, wasn’t the real importance that the drug be made? “If a drug works for people, should I care [about my role]?” Druker asked himself. “No, I shouldn’t.” But he couldn’t let go. “This is my baby,” said Druker. “This is what . . . I’ve staked my career on, and I want to be part of this clinical trial.” He refused to give up, but knowing the company’s lackluster commitment to developing the drug left him unsteady, uncertain of Ciba-Geigy’s intentions when it came to his involvement. In the end, it was Lydon who brought Druker the reassurance he needed. “I always had confidence with Nick there, and with his guidance, that I’d be involved, and more importantly, that the compound was going to make it into clinical trials.”

  IN FEBRUARY 1996, Ciba-Geigy held a second meeting to discuss the clinical development of CGP-57148B. Even though the drug wasn’t ready yet, preparations needed to begin long before the clinical trial. The team—those spearheading the development of new drugs from within the company together with the physicians who would conduct the tests—had numerous matters to settle. The company representatives included Lydon, Matter, and Ford, who was continuing to consult with Gratwohl. Druker was invited, bolstering his confidence that, at least for now, he was still in the running to be a so-called principal investigator, a clinical trial leader. John Goldman returned for the second planning meeting, too.