A Cancer Drug with Fan Mail

Copyright (c)2000 by the American Society for Clinical Investigation The crystal structure of AMP-PNP bound to Lck was used to make a model of the Abl kinase domain. Left, the general architecture of the Abl kinase domain; top right, topology of the ATP-binding cleft of the Abl protein kinase; bottom right, amino acid side chains lining the APT-binding cleft. "I feel so good that sometimes I forget I have leukemia," writes Gay Bratton of her experience with STI571, Novartis Pharma's experimental drug for treatment of chronic myelogenous leukemia (CML). Bratton, a high school math teacher and grandmother, started taking STI571 at Houston's M.D. Anderson Cancer Center last December as a participant in a Phase I trial of the drug. With STI571, her white blood cell counts are under control and her anemia is gone, giving her the energy to function normally. Judy Orem was the ninth person to take STI571 for CML at Oregon Health Sciences University (OHSU) in Portland. After more than a year on the drug, she has some puffiness around the eyes and has learned to eat before taking her pill to avoid an upset stomach. Her real side effects, she writes, are "getting my taste back, being able to be active again, having energy, having my mind back, and enjoying life." Jerry Mayfield is not taking STI571. Unlike Bratton and Orem, his CML can still be controlled with interferon-* (IFN-*), the current front-line treatment; for that reason, he was not eligible to enroll in the Phase I trial. But Mayfield, a retired Louisiana state trooper, is so passionate about STI571 that he operates, at his own expense, a Web site to spread the word (www.newcmldrug.com). "It's just so amazing to hear things like that," says Brian Druker, the OHSU physician who is principal investigator of the STI571 clinical trial. "If you heard that from one patient, that would be one thing," he says enthusiastically, "but those are the sorts of things I hear from every single one of my patients." A New Class of Drugs Although STI571 is only now in Phase II trials, its Phase I results--positive responses from every patient, each of whom had exhausted other treatment options--make it difficult to believe the Food and Drug Administration won't eventually approve it. And that will be a milestone for pharmacology, for STI571 is the first in a new drug class: protein kinase inhibitors. Thanks to the dedication of Druker and scientists at Novartis, STI571 has dramatically disproved what once seemed a very reasonable argument against kinase inhibitors--that they inevitably would have side effects too severe for patients to tolerate. Protein kinases transfer phosphates from adenosine triphosphate (ATP) to other proteins. They are ubiquitous players in signal transduction--humans have hundreds of them--and central to proper control of myriad biological processes, including cell growth and differentiation. Their dysfunction is important in cancer. However, enthusiasm for developing drugs to inhibit cancer-related protein kinases was tempered by an observation that goes back to the 1980s: It was evident that the logical binding sites for inhibitors, the kinases' ATP binding domains, had a high degree of primary sequence similarity. That led easily to the belief that finding an inhibitor selective for a single ATP binding domain would be extremely difficult. Therefore, because of unintended inhibition of related protein kinases, side effects of inhibitors were bound to be severe, and probably intolerable. But in the early 1990s, new research encouraged Ciba-Geigy (today, merged into Novartis) to set aside that argument and launch a protein kinase inhibitor program. Ciba and other companies found extremely selective epidermal growth factor (EGF) receptor inhibitors that targeted the receptor's ATP binding site. Around the same time, the first protein kinase crystal structure was solved; it showed a large active site with amino acid residues not used by ATP that could be exploited for drug binding. "At that point, it became clear the dogma [about selectivity being unachievable] was wrong," recalls Nick Lydon, who led the team that discovered STI571. When Ciba started its program, the most attractive drug targets were the EGF receptor, which was highly overexpressed in some human tumors, especially squamous cell carcinoma, and Bcr-Abl of CML. The Philadelphia Chromosome CML results from a single genetic accident, the translocation of chromosomes 9 and 22 to produce a distinct hybrid known as the Philadelphia chromosome, detectable cytogenetically in 90 percent of patients. At the junction point of the Philadelphia chromosome, translocation joins part of the Bcr gene from chromosome 22 to part of c-Abl from chromosome 9, creating a hybrid gene, Bcr-Abl. The tyrosine kinase activity of the Bcr-Abl fusion protein derives entirely from c-Abl, but translocation deletes the c-Abl domain regulating that activity. Deregulation, by a process poorly understood, initiates overproduction of myeloid blood cells. This begins the chronic stage of CML, which can endure several years. Common symptoms are fatigue due to anemia, pain, weight loss, and enlargement of the spleen. Bone marrow transplantation is the only cure, but most patients are either too old for it to be effective or unable to get marrow donors. IFN-* and other drugs can make CML manageable, but almost inevitably, further mutations drive CML to its final stage, blast crisis, characterized by severe anemia, immune system failure, and platelet deficiency. Few survive blast crisis beyond six months. Lydon's team at Ciba laboratories in Basel, Switzerland, including Juerg Zimmermann, Elizabeth Buchdunger, Peter Traxler, and Helmut Mett, screened compound libraries at random, looking for inhibitors of cancer-associated kinases. Eventually they found a 2-phenylaminopyrimidine that was a weak inhibitor of the kinase activity of the platelet-derived growth factor (PDGF) receptor (PDGF-R). Their interest turned to CML when they discovered derivatives that were potent dual inhibitors of PDGF-R and Bcr-Abl; their compounds killed Philadelphia chromosome-positive (Ph+) cells in vitro and in vivo when implanted into nude mice. At that point, around 1995, they began collaborating with Druker, a CML specialist. But it wasn't a logical next step to start a CML drug development program. CML represented a very small market--only about 4,500 Americans are annually diagnosed with CML, according to the Leukemia and Lymphoma Society of America. To a big drug company, numbers like that often mean a drug isn't going to be profitable. But Druker and Lydon argued that their best drug, STI571, should be developed to validate the concept of protein kinase inhibitors. As they point out in a recent paper,1 CML made a good test case for a new drug class. The heart of their argument was that if a protein kinase inhibitor could have a big effect on a cancer, it might be more likely with CML, caused by just one mutation, than with other cancers, which nearly always arise after damage to multiple genes. Eventually, Novartis committed to clinical trials, and the preclinical team, including Peter Graff, Ulrika Pfaar, and Hans Michael Buerger, formulated STI571 into pill form and got it ready. Response: 31/31 Phase I began in June 1998, with Druker coordinating tests in Portland, Houston, and Los Angeles. Enrolling patients were essentially in the latter part of the chronic phase; they either had rising Ph+ cell counts in spite of interferon, or were unable to tolerate interferon's side effects. "It became evident very early on," says Lydon, "that there were some spectacular effects. Once we reached the target dose that in animals and cell-based studies killed off the leukemic cells, we saw normalization of blood cell counts in all the patients." As Druker recently reported at a meeting of the American Association for Cancer Research, at 300 mg/day, 31 of 31 patients responded, with minimal side effects. "We are doing better than with previous treatments," Druker says. "All the people in the study were people who had failed interferon. And we now have them back with normal blood counts, and half of them are having a decrease in the percentage of cells that have the Philadelphia chromosome. They are tolerating their therapy very well; not only do they feel better than when they were on interferon, but their disease is under better control than when they were on interferon." Some of the patients have been on STI571 daily for a year with no apparent side effects. It is, however, too early to say the drug is delaying progression to blast crisis, "although there are some hints of that," Druker says. He adds that he does not consider anyone cured. Although 10-20 percent of the patients show no Philadelphia chromosomes in cell counts, RT-PCR assays show they have low levels of Bcr-Abl mRNA transcripts. If they get down to none and stay that way for a couple of years, Druker might consider taking them off STI571. The mildness of STI571 side effects is immensely surprising, since indirect data suggest it is also inhibiting c-Abl, the c-kit tyrosine kinase, and PDGF-R. While imperfect selectivity was worrisome prior to clinical trials, Druker argued that the plight of CML patients made the risk worth taking. The apparent short-term safety of STI571 will be encouraging if it is now used in clinical trials for glioblastoma, where PDGF-R may play a role, and small-cell lung cancer, where c-kit may be important. STI571 isn't as successful for patients in blast crisis; Druker has observed a remission rate of 25 percent, with about 10 percent having remissions lasting more than six months. Nevertheless, STI571 is the best treatment yet; with addition of other medications, perhaps the remission rate will be improved. His results are similar when treating Ph+ acute lymphocytic leukemia (ALL) patients. Bcr-Abl fusion proteins are slightly different between ALL and CML, but the Abl tyrosine kinase domain STI571 targets is identical. The lower effectiveness against ALL and blast crisis CML is probably due to the presence of other mutations. Many kinase inhibitors are now in development. Tyrosine kinase inhibitors for the EGF receptor are in clinical trials; others are being developed to treat prostate cancer and control angiogenesis. Some new drugs may come from Lydon's company, Kinetix Pharmaceuticals Inc., of Medford, Mass. Lydon started Kinetix in part out of frustration with the slow pace of Novartis' development of STI571. Kinetix specializes in kinase inhibitors, using data from protein kinase crystal structures to design selective compounds. Some of the company's drugs are for cancer, others for immune system disorders. Druker hopes the FDA will approve STI571 in a couple of years, after Phase II is completed. He plans Phase III as a survival study lasting five years or more, measuring the durability of benefits and studying long-term side effects. Given that results with advanced-stage patients are so spectacular, he predicts STI571 will show especially pleasing data when tested for newly diagnosed CML. With the excitement and hope it stirs, the new drug has captured the hearts of CML patients. Bratton's words undoubtedly speak for many: "We are hopeful that STI571 will continue to work miracles, but if we learned today that it would only work for a short time, it would be worth taking for the improved quality of life." S Tom Hollon (thollon@starpower.net) is a freelance writer in Rockville, Md. 1. B.J. Druker, N.B. Lydon, "Lessons learned from the development of an Abl tyrosine kinase inhibitor for chronic myelogenous leukemia," Journal of Clinical Investigation, 105:3-7, January 2000.

A Cancer Drug with Fan Mail

Interested in reading more?

Become a Member of