《药物化学》课程文献资料（Medicinal Chemistry）GLIVEC（STI571, IMATINIB）,A RATIONALLY DEVELOPED,TARGETED ANTICANCER DRUG，Nat Rev Drug Disc 2002.pdf_大学文库

© 2002 Nature Publishing Group REVIEWS Until the early 1980s, drug discovery programmes for cancer were focused almost exclusively on DNA synthesis and cell division, and resulted in agents such as antimetabolites, alkylating agents and microtubule destabilizers. These drugs showed efficacy, but at the price of high toxicity due to lack of selectivity. Also, resistance was frequently observed after initial stabilization or regression of the disease. The discovery of cancer-causing genes, later called oncogenes, represented a radical departure — all of a sudden, genes were identified that were uniquely associated with cancerous cells. The molecular epidemiology of these genes was established over many years of studying clinical tumour samples, but as described below, it was clear at the outset that chronic myelogenous LEUKAEMIA (CML) — a haematological stem-cell disorder that is characterized by excessive proliferation of cells of the myeloid lineage — represented a particularly interesting case. Target selection: BCR–ABL CML is characterized by a reciprocal translocation between chromosomes 9 and 22 (REF. 1). The shortened version of chromosome 22, which is known as the Philadelphia chromosome, was discovered by Nowell and Hungerford2 , and provided the first evidence of a specific genetic change associated with human cancer. The molecular consequence of this inter-chromosomal exchange is the creation of the BCR–ABL gene, which encodes a protein with elevated tyrosine-kinase activity. The demonstration that Bcr–Abl as the sole oncogenic event could induce leukaemias in mice3–5 has established BCR–ABL as the molecular pathogenic event in CML. As the tyrosine-kinase activity of BCR–ABL is crucial for its transforming activity 6 , the enzymatic activity of this deregulated gene could plausibly be defined as an attractive drug target for addressing BCR–ABL-positive leukaemias. For the first time, a drug target was identified that clearly differed in its activity between normal and leukaemic cells. It was conceivable that this enzyme could be approached with classical tools of pharmacology, as its activity — the transfer of phosphate from ATP to tyrosine residues of protein substrates — could clearly be described and measured in biochemical as well as cellular assays. Furthermore, cell lines that were derived from human leukaemic cells with the same chromosomal abnormality were available. Such cell lines were instrumental for in vitro and animal studies, which laid the groundwork for the clinical trials. So, the essential tools were assembled to go forward with the aim of identifying potent and selective inhibitors of the ABL tyrosine kinase. GLIVEC (STI571, IMATINIB), A RATIONALLY DEVELOPED, TARGETED ANTICANCER DRUG Renaud Capdeville, Elisabeth Buchdunger, Juerg Zimmermann and Alex Matter In the early 1980s, it became apparent that the work of pioneers such as Robert Weinberg, Mariano Barbacid and many others in identifying cancer-causing genes in humans was opening the door to a new era in anticancer research. Motivated by this, and by dissatisfaction with the limited efficacy and tolerability of available anticancer modalities, a drug discovery programme was initiated with the aim of rationally developing targeted anticancer therapies. Here, we describe how this programme led to the discovery and continuing development of Glivec (Gleevec in the United States), the first selective tyrosine-kinase inhibitor to be approved for the treatment of a cancer. NATURE REVIEWS | DRUG DISCOVERY VOLUME 1 | JULY 2002 | 493 Novartis Oncology, Novartis Pharma AG, S-27 2.033, CH-4002 Basel, Switzerland. Correspondence to R.C. e-mail: renaud.capdeville@ pharma.novartis.com doi:10.1038/nrd839 LEUKAEMIA Leukaemia is an uncontrolled proliferation of one type of white blood cell (leukocyte)

© 2002 Nature Publishing Group APOPTOSIS Programmed cell death. SYNGENEIC MODEL An animal model in which the injected tumour cells are derived from the same animal species as the host animal. 494 | JULY 2002 | VOLUME 1 www.nature.com/reviews/drugdisc REVIEWS and low solubility in water. The attachment of a highly polar side chain (an N-methylpiperazine) was found to improve markedly both solubility and oral bioavailability. To avoid the mutagenic potential of aniline moieties, a spacer was introduced between the phenyl ring and the nitrogen atom. The best compound from this series was a methylpiperazine derivative that was originally named STI571 (imatinib, now known as Glivec or Gleevec), which was selected as the most promising candidate for clinical development9,10 (FIG. 1d). Docking studies11 and X-ray crystallography12 showed that binding of Glivec occurs at the ATP-binding site. Analysis of the crystal structure12 showed that Glivec inhibits the ABL kinase by binding with high specificity to an inactive form of the kinase. The need for the kinase to adopt this unusual conformation, which favours binding, might contribute to the high selectivity of the compound. Unexpectedly, these analyses indicated that the N-methylpiperazine group (added to increase drug solubility) also interacted strongly with ABL by means of hydrogen bonds to the backbone carbonyl group of isoleucine (Ile)360 and histidine (His)361. In an in vitro screen against a panel of protein kinases, the compound was found to inhibit the autophosphorylation of essentially three kinases: BCR–ABL, c-KIT and the platelet-derived growth factor (PDGF) receptor (TABLE 1). More recently, activity against ARG kinase has also been reported13. Pharmacological profile In collaboration with Brian Druker, the selective inhibitory activity of Glivec was shown at the cellular level on the constitutively active p210BCR–ABL tyrosine kinase14. Subsequently, a similar inhibitory activity was also shown on other ABL fusion proteins, such as p185BCR–ABL (REFS 15,16) and TEL (ETV6)–ABL15. The inhibition of autophosphorylation of BCR–ABL was closely related to the antiproliferative activity of Glivec. Incubation with submicromolar concentrations of Glivec selectively induced APOPTOSIS in BCR–ABL-positive cell lines, and induced cell killing in primary leukaemia cells from patients with Philadelphia-chromosome-positive (Ph+) CML and acute lymphoblastic leukaemia14,16–20. In in vivo experiments, once daily intraperitoneal treatment with 2.5–50 mg kg–1 of Glivec, started one week after injecting BCR–ABL-transformed 32D cells into SYNGENEIC mice, caused dose-dependent inhibition of tumour growth14. In nude mice implanted with KU812 cells, oral treatment with 160 mg kg–1 daily in three divided doses for 11 consecutive days was associated with continuous blockage of p210BCR–ABL tyrosine phosphorylation, and resulted in tumour-free survival of the animals20. The antitumour effect of Glivec was specific for BCR–ABL-expressing cells, as no growth inhibition occurred in mice that were given injections of U937, a BCR–ABL-negative myeloid cell line. Recently, Glivec was shown to be orally active in a mouse model of CML, based on retroviral p210BCR–ABL Medicinal chemistry The starting point for every medicinal-chemistry project is a lead compound with a given pharmacological activity. However, the biological activity of a molecule must be complemented by other properties that make the molecule a good drug — it is estimated that a large proportion of molecules fails in late stages of drug development due to drug–drug interactions or poor ADME (absorption, distribution, metabolism and excretion) features. Not detecting these liabilities early in the drug discovery process can be extremely costly and time consuming. On the basis of physical and calculated properties for known drugs, criteria for ‘drug-likeness’ have been established7 . In the case of Glivec, a lead compound was identified in a screen for inhibitors of protein kinase C (PKC). This compound — a phenylaminopyrimidine derivative — had promising ‘lead-like’ properties8 and a high potential for diversity, allowing simple chemistry to be applied to produce compounds with more potent activity or selectivity. Strong PKC inhibition in cells was obtained with derivatives bearing a 3′-pyridyl group at the 3′-position of the pyrimidine (FIG. 1a). During the optimization of this structural class, it was observed that the presence of an amide group on the phenyl ring provided inhibitory activity against tyrosine kinases, such as the BCR–ABL kinase (FIG. 1b). At this point, a key observation from analysis of structure–activity relationships was that a substitution at position 6 of the diaminophenyl ring abolished PKC inhibitory activity completely. Indeed, although the introduction of a simple ‘flag-methyl’ led to loss of activity against PKC, the activity against protein tyrosine kinases was retained or even enhanced (FIG. 1c). However, the first series of selective inhibitors that was prepared originally showed poor oral bioavailability N N N N H N N N N N R1 H H 6 O N N N N N R1 H H O H3C N N N N H H N O N N a b c d Figure 1 | Summary of the chemical optimization. The core structure of the lead compound, a phenylamino derivative, is indicated in black. a | The addition of a 3′-pyridyl group (blue) at the 3′-position of the pyrimidine enhanced the cellular activity. b | An amide group (red) attached to the phenyl ring provided activity against tyrosine kinases. c | A ‘flag methyl’ (purple) attached to the diaminophenyl ring abolished the undesirable protein-kinase-C inhibitory activity. d | The final attachment of an N-methyl piperazine moiety (green) markedly increased the solubility and oral bioavailability

© 2002 Nature Publishing Group NATURE REVIEWS | DRUG DISCOVERY VOLUME 1 | JULY 2002 | 495 REVIEWS Fundamental phenotypic features in BCR–ABLpositive cells involve resistance to apoptosis, enhanced proliferation and altered adhesion properties. The impact of Glivec on some known downstream signalling molecules of BCR–ABL has been examined. A link between constitutive activation of STAT5 (signal transducer and activator of transcription 5) and enhanced viability of BCR–ABL-transformed cells has been shown22,23. Glivec had a profound inhibitory effect on STAT5 activation in vitro and in vivo21–23. Furthermore, inhibition of the BCR–ABL kinase activity by Glivec in BCR–ABL-expressing cell lines and fresh leukaemic cells from CML patients induced apoptosis by suppressing the capacity of STAT5 to activate the expression of the anti-apoptotic protein BCL-XL 23. The adaptor molecule CRKL is a prominent target of BCR–ABL, and its tyrosine phosphorylation has been a useful marker of BCR–ABL kinase activity24. As expected, a decrease in tyrosine phosphorylation of CRKL has been observed in Glivec-treated cell lines, and has also served as an indicator of BCR–ABL kinase activity in patients (see below). There is increasing evidence that cell-cycle regulation is disturbed in BCR–ABL-positive cells; however, the underlying molecular mechanisms are poorly understood. Recently, BCR–ABL has been shown to promote cell-cycle progression and activate cyclindependent kinases by interfering with the regulation of the cell-cycle inhibitory protein p27 (REF. 25). Glivec prevented downregulation of p27 levels in BCR–ABLexpressing cells25,26. The effects of Glivec on cytoskeletal changes and adhesion have been investigated using BCR–ABLtransfected fibroblasts27. Glivec was shown to restore normal architecture and to increase adhesion in this model of BCR–ABL expression. Clinical development in CML Because of the three known targets of Glivec, many potential cancers can be speculated to be good candidates for clinical testing of this new drug. However, in most cancers, tumorigenesis is complex and involves the disruption of multiple genes and signalling pathways. By contrast, CML can be considered as one of the few examples of a malignancy in which a single signallingpathway defect is thought to cause the disease. In addition, in contrast to most of the solid tumours, for which the measurement of tumour response is complex, pharmacodynamic response in CML can be measured easily using blood leukocyte count as the end point. For these reasons, CML was selected as the first indication for Phase I clinical testing. Clinically, CML is a chronic disease that evolves through three successive stages, from the chronic phase to the end stage of blast crisis that resembles acute leukaemia (FIG. 2). Overall, the median survival time of patients with newly diagnosed CML is approximately 5–6 years with an interferon-based treatment regimen. The first trial with Glivec was a Phase I study in patients with chronic-phase, and subsequently also with blastphase, CML. In this trial, patients were treated with doses transduction of transplanted bone marrow. Survival of animals was significantly prolonged, together with a marked improvement in peripheral-white-blood-cell counts and splenomegaly21. Table 1 | Cellular profile of Glivec Assay IC50 (µM) Inhibition of autophosphorylation v-ABL 0.1–0.3 p210BCR–ABL 0.25 p185BCR–ABL 0.25 TEL–ABL 0.35 TEL–ARG 0.5 PDGF receptor 0.1 TEL–PDGF receptor 0.15 c-KIT 0.1 FLT3 > 10 c-FMS and v-fms > 10 EGF receptor > 100 c-ERBB2 > 100 Insulin receptor > 100 IGF-1 receptor > 100 v-SRC > 10 JAK2 > 100 Inhibition of MAPK activation PDGF dependent 0.1–1 SCF dependent 0.1–1 Inhibition of AKT activation SCF dependent 0.1–1 Inhibition of IP release PDGF induced 0.25 Inhibition of c-FOS mRNA expression PDGF induced 0.3 –1 EGF, FGF or PMA induced > 100 Antiproliferative activity* 32D, MO-7e, BaF3 cells > 10 BCR–ABL-transfected 32D, MO-7e, BaF3 cells < 1 BCR–ABL-positive human leukaemia lines|| 0.1–1 BaF3 TEL-ARG 0.5 BALB/c 3T3 v-SIS (PDGF autocrine) 0.3 BaF3 TEL–PDGF receptor < 1 U-87 human glioma‡ ~1.5 U-343 human glioma‡ ~1.5 MO-7e, SCF stimulated ~0.1 H526 human SCLC, SCF stimulated§ 0.8 Human GIST882 line¶ < 1 Human mast-cell leukaemia line HMC-1# 0.01–0.1 Glivec concentrations that cause 50% inhibition (IC50) are given13–20,47,48,53,54,61,63,66. EGF, epidermal growth factor; FGF, fibroblast growth factor; FLT3, fms-related tyrosine kinase 3; IGF-1, insulin-like growth factor-1; IP, inositol phosphate; MAPK, mitogen-activated protein kinase; PDGF, plateletderived growth factor; PMA, phorbol 12-myristate 13-acetate; SCF, stem-cell factor; SCLC, small-cell lung cancer. *Antiproliferative experiments were carried out in 10% fetal calf serum, except for those that were carried out in ‡ 5% human-platelet poor plasma or under § serum-free conditions. ||K562, KU812, MC-3, MBA-1, KBM-5, Z-33, Z-119, Z-181. ¶Expresses the activating KIT mutation K642E (lysine 642 to glutamic acid). # Expresses the activating KIT mutation V560G (valine 560 to glycine)

© 2002 Nature Publishing Group 496 | JULY 2002 | VOLUME 1 www.nature.com/reviews/drugdisc REVIEWS Subsequently, three large multinational studies have been carried out in 532 patients with late chronic-phase CML in whom previous interferon therapy had failed31, in 235 patients with accelerated-phase CML32, and in 260 patients with myeloid blast crisis33. Treatment was given at a dose of 400 mg in the chronic-phase trial and 600 mg in the two other studies. The results of these three studies indicated that the rate of both haematological and cytogenetic response increased as the treatment was started earlier in the course of the disease (FIG. 4). Importantly, the achievement of a haematological and/or cytogenetic response was associated with improved survival and progression-free survival31–33. In the chronic-phase study, in which patients started treatment within a median of 32 months after their initial diagnosis, the estimated probability of being free of progression at 18 months was 89.2%31. The most frequently reported adverse events were mild nausea, vomiting, oedema and muscle cramps. However, rare but serious adverse events, such as liver ranging from 25 to 1,000 mg per day, and no maximal tolerated dose was identified, despite a trend for a higher frequency of GRADE III–IV ADVERSE EVENTS at doses of 750 mg or higher. On the other hand, a clear dose–response relationship with respect to efficacy was described in patients with chronic-phase CML.At doses of 300 mg or higher, 98% of the patients achieved a complete haematological response, and trough serum levels were above the concentrations required for in vitro activity28,29. Subsequently, a mathematical modelling of the relationship between dose and response, as measured by leukocyte counts after four weeks of therapy, confirmed that doses of 400 mg and higher were optimal in inducing a haematological response30 (FIG. 3). In addition, effective inhibition of the BCR–ABL kinase was documented in patient samples by inhibition of the phosphorylation status of the downstream target CRKL27. From this study, doses ranging from 400 mg (for chronic-phase patients) to 600 mg (for advanced-phase CML) were recommended for subsequent studies. GRADE III–IV ADVERSE EVENTS For each adverse event that is associated with a specific treatment, grades are assigned and defined using a scale from 0 to V. Grade III, severe and undesirable adverse event; grade IV, life-threatening or disabling adverse event. 1,000 0 200 400 600 Dose (mg) Model-fitted line WBC counts (× 109 –l 1) 800 1,000 100 10 1 Chronic patients Upper limit of normal range of WBCs Figure 3 | Dose–response relationship of Glivec in CML (Phase I study). Using the leukocyte (white blood cell; WBC) count after 28 days of treatment as a pharmacodynamic marker, the relationship between dose and response was modelled using an Emax model , which makes the assumption that once the maximal effect is achieved (Emax), increasing the dose further does not translate into additional benefit. The data indicate that at doses of 400 mg per day or higher, all the patients are predicted to achieve a reduction of their leukocyte counts within normal range below 10 x 109 l –1 . Adapted with permission from REF. 30 © (2001) American Society of Clinical Oncology. CML, chronic myelogenous leukaemia. Chronic phase Median 4–6 years stabilization Accelerated phase Advanced phases Median duration up to 1 year Blastic phase (blast crisis) Median survival 3–6 months Figure 2 | Clinical course of chronic myelogenous leukaemia

© 2002 Nature Publishing Group NATURE REVIEWS | DRUG DISCOVERY VOLUME 1 | JULY 2002 | 497 REVIEWS observed by Kuriyan and co-workers12 in a complex between mouse c-Abl and a Glivec analogue, and cannot bind ATP. The knowledge of the crystal structure allows a better understanding of the decreased sensitivity of mutated BCR–ABL to Glivec, and can be a powerful tool in the design of new BCR–ABL inhibitors that maintain inhibitory activity against these mutated kinases. Resistance to Glivec might also be related to pharmacokinetic factors. Glivec is a substrate of the multidrug-resistance-associated P-glycoprotein (PgP). toxicity or fluid-retention syndromes, were also reported. Neuropaenias and thrombopaenias were more common in patients with advanced disease, which indicates that haematological toxicity might be related more to an underlying compromised bonemarrow reserve than to toxicity of the drug itself through inhibition of c-KIT-driven haematopoiesis. Taken together, these findings have established Glivec as a safe and effective therapy in all stages of CML, and were the basis for marketing approval by the FDA on 10 May 2001 — less than three years after the start of the first Phase I study (FIG. 5). Resistance. In CML blast crisis, even though the rate of haematological responses with Glivec is high, these responses are usually short lived, and most patients will ultimately develop resistance and undergo disease progression. A prerequisite to optimally develop strategies to prevent or overcome this resistance is to get a good understanding of the potential mechanisms of resistance in these patients. On the basis of preclinical and clinical data that are available at present, several potential mechanisms of resistance have been described, which are summarized in BOX 1. They can be categorized into two main groups: mechanisms whereby BCR–ABL is reactivated and cell proliferation remains dependent on BCR–ABL signalling, and mechanisms whereby the BCR–ABL protein remains inhibited by Glivec, but alternative signalling pathways become activated. BCR–ABL overexpression and BCR–ABL gene amplification has been shown in p210BCR–ABL-transformed mouse haematopoietic Ba/F3 cells that are resistant to Glivec34,35, as well as in human BCR–ABL-positive leukaemia lines LAMA84 and AR230 (REFS 35,36). In treated patients, there is now increasing evidence that amplification of the BCR–ABL gene and mutations in the BCR–ABL kinase domain are two common mechanisms of resistance to Glivec. The occurrence of these mechanisms was first reported by Sawyers’ group37. In a study of 11 patients with blast crisis and overt clinical resistance when treated with Glivec, 3 had amplification of the BCR–ABL gene and 6 had a point mutation in the ABL kinase domain, which resulted in a T315I (threonine 315 to isoleucine) amino-acid substitution. Following this initial report, the T315I mutation as well as further mutations in the ABL kinase domain have been reported by various investigators38–41. Even though these mutations vary in their type and frequency, it is speculated that they might all lead to a reactivation of BCR–ABL-driven signal transduction. To understand the molecular mechanism by which such mutations might cause resistance to Glivec, current studies are using X-ray crystallography to analyse the three-dimensional structure of a complex between the drug and the human c-ABL kinase domain. Glivec binds to an unusual, inactive conformation of ABL with the amino terminus of the activation loop, which contains the highly conserved DFG (asparagine-phenylalanine-glycine) motif, folded into the ATP-binding site42. This conformation has been Haematological response Major cytogenic response % of patients 100 80 60 40 20 0 Late chronic phase, IFN failure (n = 524) Accelerated phase (n = 181) Blast crisis (n = 229) 95 69 30.6 60 24 16.2 Figure 4 | Haematological and cytogenetic response in CML: Phase II data. In all studies, results are expressed as the percentage of responding patients among the patients for whom the diagnosis of the correct phase of chronic myelogenous leukaemia (CML) was confirmed after a central review of the data. A major cytogenetic response combines both complete (0% Ph+ metaphases) and partial (1–35% Ph+) responses. Haematological response was defined as complete haematological response (CHR) in the chronicphase study, and as either a CHR, a marrow response or a return to chronic phase (RTC) in the advanced-phase studies, all to be confirmed after at least four weeks. In the chronic-phase study, CHR was defined as white blood cells <10 x 109 l –1, platelets <450 x 109 l –1, myelocytes and metamyelocytes <5% in blood, no blasts and promyelocytes in blood, basophils <20% and no extramedullary involvement. In advanced-phase studies, CHR was defined as neutrophils = 1.5 x 109 l –1, platelets = 100 x 109 l –1, no blood blasts, marrow blasts <5% and no extramedullary disease. A marrow response was defined by the same criteria as for CHR, but with neutrophils = 1 x 109 l –1 and platelets = 20 x 109 l –1 . An RTC was defined as <15% blasts in marrow and blood, <30% blasts and promyelocytes in marrow and blood, <20% basophils in blood and no extramedullary disease. IFN, interferon; Ph+, Philadelphia chromosome positive

© 2002 Nature Publishing Group 498 | JULY 2002 | VOLUME 1 www.nature.com/reviews/drugdisc REVIEWS that are unrelated to BCR–ABL signalling, and further studies will be necessary to fully assess the significance of these findings and their relevance to CML patients. Current and future development in CML The activity of Glivec in patients with newly diagnosed CML is being further investigated by a large randomized Phase III study to compare first-line therapy with Glivec against standard interferon in combination with low-dose cytarabine. This study, known as the ‘IRIS’ study (International Randomized study of Interferon versus STI571), has enrolled 1,106 patients. The results of an interim analysis with a median follow-up of 14 months indicate a better tolerability and a superior efficacy of first-line Glivec compared with interferon and low-dose cytarabine in terms of cytogenetic response, haematological response and, more importantly, time to progression to accelerated phase or blast crisis46. Preclinical studies have shown that the combination of Glivec with various anticancer agents might have synergistic effects. Consequently, several Phase I/II studies are evaluating the feasibility of combining Glivec with interferon, polyethylene glycol (PEG)ylated interferon, cytarabine and other single-agent or combination chemotherapy regimens, in patients with either chronicphase or advanced CML. Accordingly, the uptake of Glivec was reduced in Glivec-resistant LAMA84 cells in association with an overexpression of the PgP protein. Sensitivity to Glivec was recovered when cells were treated with the PgP inhibitor verapamil35. At clinically relevant concentrations of Glivec, binding to plasma proteins is approximately 95%, mostly to albumin and α1-acid glycoprotein (AGP). It has been suggested that a potential mechanism of resistance might relate to this high binding to increased levels of AGP, which would lead to insufficient availability of free drug for antileukaemic activity 43. However, the clinical significance of this hypothesis is uncertain, in particular in view of the finding that AGP purified from CML patients failed to block the effect of Glivec on the proliferation of leukaemic cells44. Recently, Hofmann et al.45 studied a small group of patients with Ph+ acute lymphoblastic leukaemia who were resistant to Glivec by using DNA-microarray expression profiling. They described an association between the occurrence of resistance to Glivec and upregulation of genes encoding proteins such as Bruton tyrosine kinase and two ATP synthetases (ATP5A1 and ATP5C1), and downregulation of other genes, such as the pro-apoptotic gene BAK1 and the cell-cycle-control gene INK4B45(also known as p15). This is the first report to identify dysregulation of genes June 1998 – First patient with CML treated. 1990 – Lead compound identified in a screen for inhibitors of PKC. 1996 – In vivo activity shown in BCR–ABL-transformed cells in syngeneic mice. June 2000 – Phase III trials initiated. November 2001 – Approved in Europe and Japan for CML. May 2001 – Approved by the FDA for CML. June 1999 – Phase II trials initiated. 1992 – First batch of Glivec synthesized. February 2001 – NDA submitted to FDA for CML. February 2002 – Approved by the FDA for GIST. Discovery Clinical development N N N N H H N O Discovery Clinical development Typically ~8 years Typically ~7 years Typical development timeline Glivec development timeline Figure 5 | Key points in the discovery and development of Glivec. The clinical development was particularly rapid, as can be seen by comparison with the typical drug discovery and development times shown in the inset. An NDA for Glivec was submitted just two years and nine months after treatment of the first patient with CML, and FDA approval was given less than three months after application. CML, chronic myelogenous leukaemia; GIST, gastrointestinal stromal tumour; NDA, new drug application; PKC, protein kinase C

© 2002 Nature Publishing Group NATURE REVIEWS | DRUG DISCOVERY VOLUME 1 | JULY 2002 | 499 REVIEWS Other c-KIT-expressing tumours. In human systemic mastocytosis, most cases show a point mutation in codon 17 of c-KIT, which results in a D816V (aspartic acid 816 to valine) amino-acid substitution in the kinase-2 domain of c-KIT. Interestingly, this mutated c-KIT is resistant to inhibition by Glivec53,58. Expression of c-KIT and SCF has been reported in a retrospective small-cell lung cancer (SCLC) series, indicating that SCLC growth might involve an AUTOCRINE loop. Inhibition of c-KIT activation by transfection of a dominant-negative c-KIT gene results in loss of growth-factor independence59,60. Furthermore, the c-KIT/PDGF-receptor inhibitor AG1296 inhibits growth of SCLC cells in serum-containing medium60. In H526 SCLC cells, pretreatment with Glivec inhibited SCF-mediated c-KIT activation with an IC50 (half-maximal inhibitory concentration) of 0.1 µM (REF. 61). The compound also blocked downstream signal transduction, as evidenced by inhibition of SCF-mediated activation of MAPK and AKT, and potently inhibited SCF-mediated growth in serum-free medium, with a marked increase in apoptosis. Glivec also inhibited the growth of SCLC cell lines in a dose-dependent fashion when grown in serum-containing medium; however, the average IC50 was in the range of 5 µM (REFS 61,62). Although c-KIT expression has been documented in various other human tumours, including acute myelogenous leukaemia, ovarian and testicular cancer, it will be important to determine the activation status of the receptor and its importance in the pathogenesis (for a review, see REF. 58). Furthermore, it needs to be explored whether pharmacological inhibition of PARACRINE or autocrine activation of this kinase will be successful therapeutically. Exploratory clinical studies are continuing at present in patients with c-KIT-expressing SCLC and acute myelogenous leukaemia. PDGF receptor as a target The third target of Glivec is the PDGF-receptor tyrosine kinase. Cellular studies have shown potent inhibition of the two structurally similar PDGF-α and PDGF-β receptors (PDGFR-α and PDGFR-β), as well as blockade of PDGF-mediated cellular events47,63. PDGF is a connective-tissue-cell mitogen with in vivo functions that include embryonal development, wound healing and control of interstitial-fluid pressure in soft connective tissue. There is increasing evidence that the PDGF ligand–receptor system also has an important role in tumorigenesis64. Paracrine and/or autocrine activation of the PDGFR kinase has been postulated in numerous malignancies, and the presence of PDGF autocrine loops is most well documented in gliomas65. Glivec inhibited the in vitro and in vivo growth of cells with autocrine PDGF signalling, including the formation of tumours by the human glioblastoma lines U343 and U87, which had been injected into the brains of nude mice66. The inhibitory effects were mediated predominantly through promotion of growth arrest rather than apoptosis. c-KIT is another target In addition to various oncogenic forms of the BCR–ABL tyrosine kinase, Glivec also inhibits the receptor for stem-cell factor (SCF) — c-KIT, a member of the type III group of receptor kinases. Preclinical studies have established that the drug blocks c-KIT autophosphorylation, as well as SCF-stimulated downstream signalling events, such as activation of the mitogen-activated protein kinases (MAPKs) ERK1 and ERK2, and AKT (also known as protein kinase B)47,48. Development in c-KIT-positive GISTs. Gastrointestinal stromal tumours (GISTs) represent a rare subset of softtissue sarcomas that involve the gastrointestinal tract and are thought to be derived from the interstitial cells of Cajal. Scientific rationale for the use of Glivec in the treatment of these tumours comes from the landmark work of Hirota et al.49, who first identified somatic gainof-function mutations in the c-KIT gene in patients with GIST. Oncogenic c-KIT mutations in GISTs have been localized to the extracellular domain, kinase domains 1 and 2 and predominantly in the juxtamembrane domain of the c-KIT protein50–52.As c-KIT serves as a phenotypic marker of GISTs and has a key role in their pathogenesis, it provides an ideal target for molecular-based therapy. The first evidence that Glivec might inhibit GIST cells that express mutated c-KITwas obtained from studies in a mast-cell leukaemia line expressing a mutated c-KIT similar to that found in GISTs48,53. Furthermore, Glivec rapidly and completely abolished constitutive phosphorylation of c-KIT in the human cell line GIST882, which expresses an activating c-KIT mutation in the first part of the cytoplasmic split-tyrosine-kinase domain, and inhibited proliferation in this GIST line54. Similarly, a primary GIST cell culture that expressed a c-KIT exon 11 juxtamembrane mutation was also inhibited by Glivec54. As reported recently, a pronounced tumour response was first observed in a single patient with progressing GIST55. Following this case report, the high level of efficacy of Glivec in GIST has been shown in two subsequent Phase I (REF. 56) and Phase II studies (REF. 57). Two large Phase III studies are being carried out at present to compare the effectiveness of two doses of Glivec (400 mg or 800 mg daily). On the basis of the Phase II data, the FDA approved the use of Glivec for GISTs on 1 February 2002. AUTOCRINE Describes an agent secreted from a cell that acts on the cell in which it is produced. PARACRINE Describes an agent secreted from a cell that acts on other cells in the local environment. Box 1 | Mechanisms of resistance to Glivec in CML BCR–ABL-dependent mechanisms (cells remain dependent on BCR–ABL signalling) • Amplification of BCR–ABL gene • Mutations in BCR–ABL kinase domain prevent correct binding of Glivec • Efflux of Glivec (for example, by PgP-associated MDR protein) • Protein binding of Glivec (for example, to circulating AGP) BCR–ABL-independent mechanisms (BCR–ABL is inactivated) • Activation of signalling pathways downstream of BCR–ABL • Activation of leukaemogenic pathways unrelated to BCR–ABL AGP,α1-acid glycoprotein; CML, chronic myelogenous leukaemia; MDR, multidrug resistant; PgP, P-glycoprotein

© 2002 Nature Publishing Group 500 | JULY 2002 | VOLUME 1 www.nature.com/reviews/drugdisc REVIEWS cells79, but might also be influenced indirectly through paracrine action on PDGF-responsive stromal and perivascular cells, which are a principal source of vascular endothelial growth factor (VEGF)80. PDGF has also been shown to induce the expression of VEGF in endothelial cells, which in turn causes an autocrine VEGF loop81.Anti-angiogenic activity of Glivec has been shown in vitro through inhibition of serum-stimulated capillary sprouting from rat aorta, and in vivo in a subcutaneous implant model in which the drug inhibited PDGF- and also VEGF- and basic fibroblast growth factor (bFGF)-stimulated vascularization82. Blockade of PDGFR signalling by Glivec has also been shown to inhibit angiogenesis and tumour growth in an experimental model of bone metastasis83. Glivec treatment of nude mice injected with PC-3MM human prostate-cancer cells into the tibia inhibited tumour-cell growth and induced apoptosis, both in tumour cells and tumour-associated endothelial cells. The effects were pronounced when mice were treated with the combination of Glivec and taxol. Interestingly, immunohistochemical studies showed that tumour cells growing in the bone (but not those in surrounding musculature) expressed high levels of PDGF-α, PDGF-β, PDGFR-α and PDGFR-β. Tumourassociated endothelial cells within the bone also expressed PDGFR-α and PDGFR-β. These data indicate that inhibition of the PDGFR in combination with chemotherapy might provide a new approach for the treatment of bone metastasis. Conclusion The discovery and development of Glivec has shown that is possible to produce rationally designed, molecular-targeted drugs for the treatment of a specific cancer. The research programme has also clearly shown that it is possible to define in vitro and animal models with high predictive quality, as the results of the subsequent clinical studies have largely corroborated the preclinical findings. The predictive quality was achieved in this particular case by using models with identical genetic abnormalities as those found in man. The case of Glivec also shows that compounds that do not only affect one, but two or more targets (which is frequently the case), can be beneficial in allowing several diseases with differing molecular abnormalities to be addressed, without paying too high a price in terms of toxicity. The clinical data available so far in CML, GIST and chronic myeloproliferative disorders that involve rearrangement of the PDGFR gene indicate that the inhibition of BCR–ABL, c-KIT and PDGFRs can be achieved with Glivec in humans, and translated into clinically meaningful patient benefit. Providing clinical ‘proof of concept’, these data validate the initial hypothesis of this programme, and underscore the importance of rationally selecting the target diseases to be considered in the early phases of development of a molecule such as Glivec. Beyond these reasonably well-understood malignancies, Glivec could have potential in the treatment of other malignancies that involve any of these signalling Autocrine PDGFR activation is also well documented in tumour cells of dermatofibrosarcoma protuberans (DFSP), a highly recurrent, infiltrative skin tumour that is characterized by a chromosomal rearrangement involving chromosomes 17 and 22. The resulting fusion-gene product collagen I, α1 polypeptide (COL1A1)–PDGF-β triggers the autocrine stimulation of the PDGFR67. COL1A1–PDGFβ-transformed fibroblasts, as well as primary DFSP and giant-cell fibrosarcoma cell cultures, were inhibited by Glivec in vitro and in vivo 67–69. The main mechanism by which Glivec affected DFSP tumour growth was through induction of apoptosis69. Preliminary data indicate that Glivec might also be active in patients with DFSP70. Relatively little is known about the ligand-independent activation of PDGFR. However, rearrangement of PDGFRβ has been described in chronic myeloproliferative diseases. The best known of these is the t(5;12) chromosomal translocation in chronic myelomonocytic leukaemia (CMML), in which PDGFRβ, which is located on chromosome 5, is fused to the TEL gene on chromosome 12. Transformation of haematopoietic cells occurs through oligomerization of the TEL– PDGFR-β fusion protein, which causes ligand-independent constitutive activation of the PDGFR kinase71. Glivec inhibited the growth of cells expressing TEL–PDGFRβ15, and in transgenic mice that expressed the TEL–PDGFRβ, treatment with Glivec inhibited tumour formation and prolonged survival of the animals72. A remarkable haematological and complete cytogenetic response has been observed in two patients with chronic myeloproliferative disorders associated with a t(5;12) translocation — one of them with a well-characterized TEL–PDGFR fusion gene and the second with a rearranged PDGFR gene with an as yet unidentified partner gene73. Other exploratory clinical trials are being carried out in gliomas and in prostate cancer. Targeting the tumour microenvironment An alternative strategy to influence tumour growth is to interfere with the tumour stroma and microvasculature. Paracrine PDGF signalling in the connective-tissue tumour stroma has been described in various types of solid tumour64. Several lines of evidence indicate a role for PDGF in the regulation of interstitial fluid pressure (IFP)74–76. As most solid tumours have an increased IFP, pharmacological reduction might be a way to increase the uptake of anticancer drugs into tumours77. Recent experiments have shown that Glivec significantly reduced tumour IFP in subcutaneously growing PROb rat-colon carcinomas, and a concomitant increase in trans-capillary transport of a radiolabelled tracer compound into the tumour interstitium was observed78. These effects were mediated by inhibition of the expression of PDGFR on blood vessels and stromal cells, as tumour epithelial cells in this tumour model do not express PDGFRs. The angiogenic activity that has been described for PDGF might not only be explained by its direct effects on capillary endothelial cells, pericytes and smooth-muscle

© 2002 Nature Publishing Group NATURE REVIEWS | DRUG DISCOVERY VOLUME 1 | JULY 2002 | 501 REVIEWS far as is technically feasible. This point has been crucial in the successful outcome of the CML, GIST and CMML trials. The activity of Glivec in more common cancers with multiple and more complex molecular abnormalities remains to be determined, and is the objective of continuing research in diseases such as SCLC, prostate cancer and gliomas. The potential activity of the combination of Glivec with other signal-transduction inhibitors or anticancer agents is also being investigated. pathways, or through targeting of the tumour microenvironment. However, most human cancers are likely to be heterogeneous with regard to molecular abnormalities, such as oncogene activation, and involve multiple signalling pathways in addition to either cKIT and/or the PDGFR. Consequently, careful attention will have to be paid in designing clinical trials in these more complex indications as to how patients should be selected on the basis of the expression or activation of the molecular target in their tumour, as 1. Rowley, J. D. A new consistent abnormality in chronic myelogenous leukaemia identified by quinacrine fluorescence and giemsa staining. Nature 243, 290–293 (1973). 2. Nowell, P. C. & Hungerford, D. A. A minute chromosome in human chronic granulocytic leukemia. Science 132, 1497 (1960). 3. Daley, G. Q., Van Etten, R. A. & Baltimore, D. Induction of chronic myelogenous leukemia in mice by the p210Bcr/Abl gene of the Philadelphia chromosome. Science 247, 824–830 (1990). 4. Kelliher, M. A. et al. Induction of chronic myelogenous leukemia in mice by the v-Abl and Bcr/Abl. Proc. Natl Acad. Sci. USA 87, 6649–6653 (1990). 5. Heisterkamp, N. et al. Acute leukaemia in Bcr/Abl transgenic mice. Nature 344, 251–253 (1990). 6. Lugo, T. G. et al. Tyrosine kinase activity and transformation potency of Bcr–Abl oncogene products. Science 247, 1079–1082 (1990). 7. Lipinsky, C. A. Drug-like properties and the causes of poor solubility and poor permeability. J. Pharmacol. Toxicol. Methods 44, 235–249 (2001). 8. Teague, S. et al. The design of leadlike combinatorial libraries. Angew. Chem. Int. Edn Eng. 38, 3743–3748 (1999). 9. Zimmermann, J. et al. (Phenylamino)pyrimidine (PAP) derivatives: a new class of potent and highly selective PDGF-receptor autophosphorylation inhibitors. Bioorg. Med. Chem. Lett. 6, 1221–1226 (1996). 10. Zimmermann, J. et al. Potent and selective inhibitors of the ABL-kinase: phenylaminopyrimidine (PAP) derivatives. Bioorg. Med. Chem. Lett. 7, 187–192 (1997). 11. Zimmerman, J., Furet, P. & Buchdunger, E. STI571. A new treatment modality for CML. ACS Symp. Ser. 796, 245–259 (2001). 12. Schindler, T. et al. Structural mechanism for STI571 inhibition of Abelson tyrosine kinase. Science 289, 1938–1942 (2000). The first description of the structural interactions between Glivec and ABL using crystallographic studies. Provided an important insight into potential mechanisms of resistance. 13. Okuda, K. et al. ARG tyrosine kinase activity is inhibited by STI571. Blood 97, 2440–2448 (2001). 14. Druker, B. J. et al. Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr–Abl positive cells. Nature Med. 2, 561–566 (1996). The first study to document the strong efficacy of Glivec in in vitro and in vivo models of BCR–ABLpositive leukaemias. 15. Carroll, M. et al. CGP 57148, a tyrosine kinase inhibitor, inhibits the growth of cells expressing BCR–ABL, TEL-ABL, and TEL-PDGFR fusion proteins. Blood 90, 4947–4952 (1997). 16. Beran, M. et al. Selective inhibition of cell proliferation and BCR–ABL phosphorylation in acute lymphoblastic leukemia cells expressing Mr 190,000 BCR–ABL protein by a tyrosine kinase inhibitor (CGP 57148). Clin. Cancer Res. 4, 1661–1672 (1998). 17. Gambacorti-Passerini, C. et al. Inhibition of the ABL kinase activity blocks the proliferation of BCR/ABL+ leukemic cells and induces apoptosis. Blood Cells Mol. Dis. 23, 380–394 (1997). 18. Deininger, M. et al. The tyrosine kinase inhibitor CGP57148B selectively inhibits the growth of BCR–ABL-positive cells. Blood 90, 3691–3698 (1997). 19. Dan, S., Naito, M. & Tsuruo, T. Selective induction of apoptosis in Philadelphia chromosome-positive chronic myelogenous leukemia cells by an inhibitor of BCR–ABL tyrosine kinase, CGP 57148B. Cell Death Differ. 5, 710–715 (1998). 20. Le Coutre, P. et al. In vivo eradication of human BCR/ABLpositive leukemia cells with an ABL kinase inhibitor. J. Natl Cancer Inst. 91, 163–168 (1999). 21. Wolff, N. C. & Ilaria, R. L. Establishment of a murine model for therapy-treated chronic myelogenous leukemia using the tyrosine kinase inhibitor STI571. Blood 98, 2808–2816 (2001). 22. Sillaber, C. et al. STAT5 activation contributes to growth and viability in Bcr/Abl-transformed cells. Blood 95, 2118–2125 (2000). 23. Horita, M. et al. Blockade of the Bcr–Abl kinase activity induces apoptosis of chronic myeloid leukemia cells by suppressing signal transducer and activator of transcription 5-dependent expression of BCL-XL . J. Exp. Med. 191, 977–984 (2000). 24. Oda, T. et al. Crkl is the major tyrosine-phosphorylated protein in neutrophils from patients with chronic myelogenous leukemia. J. Biol. Chem. 269, 22925–22928 (1994). 25. Jonuleit, T. et al. Bcr–Abl kinase downregulates cyclindependent kinase inhibitor p27 in human and murine cell lines. Blood 96, 1933–1939 (2000). 26. Gesbert, F. et al. BCR/ABL regulates expression of the cyclin dependent kinase inhibitor p27Kip1 through the PI3K/AKT pathway. J. Biol. Chem. 50, 39223–39230 (2000). 27. Gaston, I. et al. Abl kinase but not PI3-kinase links to the cytoskeletal defects in Bcr–Abl transformed cells. Exp. Hematol. 28, 77–86 (2000). 28. Druker, B. J. et al. Activity of a specific inhibitor of the BCR–ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N. Engl. J. Med. 344, 1038–1042 (2001). The first clinical results with Glivec in CML, documenting a high level of efficacy, low level of toxicity and describing the dose–response relationship. These results confirm the crucial role of BCR–ABL in the pathophysiology of CML. 29. Druker, B. J. et al. Efficacy and safety of a specific inhibitor of the BCR–ABL tyrosine kinase in chronic myeloid leukemia. N. Engl. J. Med. 344, 1031–1037 (2001). 30. Peng, B. et al. Clinical investigation of the PK/PD relationship for Glivec (STI571): a novel inhibitor of signal transduction. Proc. Am. Soc. Clin. Oncol. 20, 280 (2001). 31. Kantarjian, H. et al. Hematologic and cytogenetic responses to imatinib mesylate in chronic myelogenous leukemia. N. Engl. J. Med. 346, 645–652 (2002). 32. Talpaz, M. et al. Glivec™ (imatinib mesylate) induces durable hematologic and cytogenetic responses in patients with accelerated phase chronic myeloid leukemia: results of a Phase 2 study. Blood 99, 1928–1937 (2002). 33. Sawyers, C. et al. Imatinib induces hematologic and cytogenetic responses in patients with chronic myeloid leukemia in myeloid blast crisis: results of a Phase II study. Blood 99, 3530–3539 (2002). 34. Weisberg, E. & Griffin, J. Mechanism of resistance to the Abl tyrosine kinase inhibitor STI571in BCR/ABL transformed hematopoietic cell lines. Blood 95, 3498–3505 (2000). 35. Mahon, F. et al. Selection and characterization of BCR–ABL positive cell lines with differential sensitivity to the tyrosine kinase inhibitor STI571: diverse mechanisms of resistance. Blood 96, 1070–1079 (2000). 36. Le Coutre, P. et al. Induction of resistance to the Abelson inhibitor STI571 in human leukemic cells through gene amplification. Blood 95,1758–1766 (2000). 37. Gorre, M. E. et al. Clinical resistance to STI-571 cancer therapy caused by BCR–ABL gene mutation or amplification. Science 293, 876–880 (2001). This study describes potential mechanisms of resistance in CML in clinical samples from treated patients. 38. Gorre, M. E. et al. Roots of clinical resistance to STI-571 cancer therapy. Science 293, 2163a (2001). 39. Barthe, C. et al. Roots of clinical resistance to STI-571 cancer therapy. Science 293, 2163a (2001). 40. Hochhaus, A. et al. Roots of clinical resistance to STI-571 cancer therapy. Science 293, 2163a (2001). 41. Von Bubnoff, N. et al. BCR–ABL gene mutations in relation to clinical resistance of Philadelphia-chromosome-positive leukaemia to STI571: a prospective study. Lancet 359, 487–491 (2002). 42. Manley, P. W. et al. Molecular interactions between Gleevec™ and isoforms of the c-Abl kinase. Proc. Am. Assoc. Cancer Res. 4196 (2002). 43. Gambacorti-Passerini, C. et al. α1 Acidic glycoprotein (AGP) binds, inhibits and causes in vivo resistance of human BCR–ABL+ leukemic cells to STI571. J. Natl Cancer Inst. 92, 1641–1650 (2000). 44. Jorgensen, H. G. et al. α1 -Acid glycoprotein expressed in the plasma of chronic myeloid leukemia patients does not mediate significant in vitro resistance to STI571. Blood 99, 713–715 (2002). 45. Hofmann, W. K. et al. Relation between resistance of Philadelphia-chromosome-positive acute lymphoblastic leukaemia to the tyrosine kinase inhibitor STI571 and geneexpression profiles: a gene-expression study. Lancet 359, 481–486 (2002). 46. Druker, B. STI571 (Gleevec/Glivec, imatinib) versus interferon (IFN)+ cytarabine as initial therapy for patients with CML: results of a randomized study. Proc. Am. Soc. Clin. Oncol.1 (2002). 47. Buchdunger, E. et al. The Abl protein-tyrosine kinase inhibitor, STI571, inhibits in vitro signal transduction mediated by c-Kit and PDGF receptors. J. Pharmacol. Exp. Ther. 295, 139–145 (2000). 48. Heinrich, M. C. et al. Inhibition of c-kit receptor tyrosine kinase activity by STI571, a selective tyrosine kinase inhibitor. Blood 96, 925–932 (2000). 49. Hirota, S. et al. Gain of function mutations of c-Kit in human gastrointestinal stromal tumors. Science 279, 577–580 (1998). This study establishes the importance of c-KIT signalling in the pathogenesis of GIST tumours. 50. Lasota, J. et al. Mutations in exons 9 and 13 of Kit gene are rare events in gastrointestinal stromal tumors. A study of two hundred cases. Am. J. Pathol. 157, 1091–1095 (2000). 51. Lux, M. L. et al. Kit extracellular and kinase domain mutations in gastrointestinal stromal tumors. Am. J. Pathol. 156, 791–795 (2000). 52. Rubin, B. P. et al. Kit activation is a ubiquitous feature of gastrointestinal stromal tumors. Cancer Res. 61, 8118–8121 (2001). 53. Heinrich, M. C. et al. STI571 inhibits the kinase activity of wild type and juxtamembrane c-Kit mutants but not the exon 17 D816V mutations associated with mastocytosis. Blood 96, 4459 (2000). 54. Tuveson, D. A. et al. STI571 inactivation of the gastrointestinal stromal tumor c-KIT oncoprotein: biological and clinical implications. Oncogene 20, 5054–5058 (2001). 55. Joensuu, H. et al. Effect of the tyrosine kinase inhibitor STI571 in a patient with metastatic gastrointestinal stromal tumor. N. Engl. J. Med. 344, 1052–1056 (2001). A proof-of-concept report of the efficacy of Glivec in KIT-expressing GIST, and rationale for further development in this indication, as well as in other solid tumours that express KIT. 56. Van Oosterom, A. T. et al. Safety and efficacy of imatinib (STI571) in metastatic gastrointestinal stromal tumors, a phase I study. Lancet 358, 1421–1423 (2001)

© 2002 Nature Publishing Group 502 | JULY 2002 | VOLUME 1 www.nature.com/reviews/drugdisc REVIEWS 57. Blanke, C. D. et al. Evaluation of the safety and efficacy of an oral molecularly-targeted therapy, STI571, in patients with unresectable or metastatic gastrointestinal stromal tumors (GISTs) expressing c-KIT (CD117). Proc. Am. Soc. Clin. Oncol. 20,1 (2001). 58. Heinrich, M. C. et al. Inhibition of Kit kinase activity: a novel molecular approach to the treatment of Kit-positive malignancies. J. Clin. Oncol. 20, 1692–1703 (2002). An excellent review of KIT as a target for anticancer therapy. 59. Krystal, G. W., Hines, S. & Organ, C. Autocrine growth of small cell lung cancer mediated by co-expression of c-kit and stem cell factor. Cancer Res. 56, 370–376 (1996). 60. Krystal, G. W., Carlson, P. & Litz, J. Induction of apoptosis and inhibition of small cell lung cancer growth by the quinoxaline tyrophostins. Cancer Res. 57, 2203–2208 (1997). 61. Krystal, G. W. et al. The selective tyrosine kinase inhibitor STI571 inhibits small cell lung cancer growth. Clin. Cancer Res. 6, 3319–3326 (2000). 62. Wang, W. L. et al. Growth inhibition and modulation of kinase pathways of small cell lung cancer lines by the novel tyrosine kinase inhibitor STI571. Oncogene 19, 3521–3528 (2000). 63. Buchdunger, E. et al. Effects of a selective inhibitor of the Abl tyrosine-kinase in vitro and in vivo by a 2-phenylaminopyrimidine derivative. Cancer Res. 56, 100–104 (1996). 64. Östman, A. & Heldin, C. H. Involvement of platelet-derived growth factor in disease: development of specific antagonists. Adv. Cancer Res. 80, 1–38 (2001). 65. Nistér, M. et al. Differential expression of platelet-derived growth factor receptors in human malignant glioma cell lines. J. Biol. Chem. 266, 16755–16763 (1991). 66. Kilic, T. et al. Intracranial inhibition of platelet-derived growth factor-mediated glioblastoma cell growth by an orally active kinase inhibitor of the 2-phenylaminopyridine class. Cancer Res. 60, 5143–5150 (2000). 67. Shimizu, A. et al. The dermatofibrosarcoma protuberansassociated collagen type 1α/platelet-derived growth factor (PDGF) B-chain fusion gene generates a transforming protein that is processed to functional PDGF-BB. Cancer Res. 59, 3719–3723 (1999). 68. Greco, A. et al. Growth inhibitory effect of STI571 on cells transformed by the COL1A/PDGFB rearrangement. Int. J. Cancer 92, 354–360 (2001). 69. Sjöblom, T. et al. Growth inhibition of dermatofibrosarcoma protuberans tumors by the platelet-derived growth factor receptor antagonist STI571 through induction of apoptosis. Cancer Res. 61, 5778–5783 (2001). 70. Awan, R. A. et al. Patients with metastatic sarcoma arising from dermatofibrosarcoma protuberans (DFSP) may respond to imatinib (STI571, Gleevec). Proc. Am. Soc. Clin. Oncol. 1637 (2002). 71. Jousset, C. et al. A domain of TEL conserved in a subset of ETS proteins defines a specific oligomerization interface essential to the mitogenic properties of the TEL–PDGFRβ oncoprotein. EMBO J. 16, 69–82 (1997). 72. Tomasson, M. H. et al. TEL/PDGFβR induces hematologic malignancies in mice that respond to a specific tyrosine kinase inhibitor. Blood 93, 1707–1714 (1999). 73. Apperley, J. F. et al. Chronic myeloproliferative diseases with t(5:12) and a PDGFRB fusion gene: complete cytogenetic remission with STI571. Blood 98, 726A (2001). A proof-of-concept study of the efficacy of Glivec in a PDGFR-driven malignancy. 74. Gullberg, D. et al. β1 Integrin-mediated collagen gel contraction is stimulated by PDGF. Exp. Cell Res. 186, 264–272 (1990). 75. Rodt, S. A. et al. A novel physiologic role for platelet-derived growth factor-BB in rat dermis. J. Physiol. (Lond.) 495, 193–200 (1996). 76. Heuchel, R. et al. Platelet-derived growth factor receptor regulates interstitial fluid homeostasis through phosphatidylinositol-3 kinase signaling. Proc. Natl Acad. Sci. USA. 20, 11410–11415 (1999). 77. Jain, R. K. Delivery of molecular medicine to solid tumors. Science 271, 1079–1080 (1996). 78. Pietras, K. et al. Inhibition of platelet-derived growth factor receptors reduces interstitial hypertension and increases transcapillary transport in tumors. Cancer Res. 61, 2929–2934 (2001). 79. Hellström, M. et al. Role of PDGF-B and PDGFR-β in recruitment of vascular smooth muscle cells and pericytes during embryonic blood vessel formation in the mouse. Exp. Cell Res. 186, 264–272 (1999). 80. Reinmuth, N. et al. Induction of VEGF in perivascular cells defines a potential paracrine mechanism for endothelial cell survival. FASEB J. 15, 1239–1241 (2001). 81. Wang, D. et al. Induction of vascular endothelial growth factor expression in endothelial cells by platelet-derived growth factor through the activation of phosphatidylinositol 3-kinase. Cancer Res. 59, 1464–1472 (1999). 82. Buchdunger, E., O’Reilly, T. & Wood, J. Pharmacology of imatinib (STI571). Eur. J. Cancer (in the press). 83. Uehara, H. et al. Blockade of PDGF-R signaling by STI571 inhibits angiogenesis and growth of human prostate cancer cells in the bone of nude mice. Proc. Am. Assoc. Cancer Res. 2192 (2001). Acknowledgements We would like to thank the members of the Glivec International Project Team for their crucial contribution to the success of this programme and their kind review of this manuscript: P. Boultbee, V. Buss, S. Dimitrijevic, A. Dortok, D. Filipovic, I. Gathmann, H. Godman, J. Jaffe, L. Letvak, P. Marbach, R. Miranda, J. Ogorka, C. Ogawa, B. Peng, S. Silberman, F. Sutter, S. Szabo, S. Wells and J. M. Ford. We also thank B. Druker for his crucial input and a fruitful collaboration throughout this programme, and N. Lydon for his contribution in the early phase of the programme. We thank also C. Schmid for her assistance in editing the manuscript. Online links DATABASES The following terms in this article are linked online to: Cancer.gov: http://www.cancer.gov/cancer_information/ acute lymphoblastic leukaemia | acute myelogenous leukaemia | chronic myelogenous leukaemia | ovarian cancer | prostate cancer | small-cell lung cancer | testicular cancer LocusLink: http://www.ncbi.nlm.nih.gov/LocusLink/ ABL | Abl | AGP | AKT | albumin | ARG kinase | ATP5A1 | ATP5C1| BAK1 | BCL-XL | BCR | Bcr | bFGF | Bruton tyrosine kinase | COL1A1 | CRKL | EGF receptor | c-ERBB2 | ERK1 | ERK2 | FGF | c-FMS | v-fms | IGF-1 receptor | INK4B | insulin receptor | JAK2 | c-KIT | p27 | PDGF-α | PDGF-β | PDGFR-α | PDGFR-β | P-GP | PKC | SCF | v-SRC | STAT5 | TEL | VEGF Medscape DrugInfo: http://promini.medscape.com/drugdb/search.asp cytarabine | Glivec | interferon | taxol | verapamil FURTHER INFORMATION FDA: http://www.fda.gov/default.htm Access to this interactive links box is free online