Clinical and Experimental Studies in Chronic Myeloid Leukemia

Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 256 Clinical and Experimental Studies in Chronic Myeloid Leukemia Studies of Treatment Outcome, In Vitro Cellular Drug Resistance and Gene Expression ULLA OLSSON-STRÖMBERG ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2007 ISSN 1651-6206 ISBN 978-91-554-6878-1 urn:nbn:se:uu:diva-7841

Per, Clara och Majajag älskar er av hela mitt hjärta!

List of Papers This thesis is based on the following papers, referred to in the text by their Roman numerals: I Olsson-Strömberg U, Simonsson B, Ahlgren T, Bjorkholm M, Carlsson K, Gahrton G, Hast R, Lofvenberg E, Linder O, Ljungman P, Malm C, Paul C, Rodjer S, Turesson I, Uden AM, Wahlin A, Killander A, Wadman B, Westin J, Vikrot O, Zettervall O, Öberg G; Swedish CML Study Group. Comparison of busulphan, hydroxyurea and allogeneic bone marrow transplantation (BMT) in chronic myeloid leukaemia: BMT prolongs survival. Hematol J. 2004; 5 (6): 462-6. II Olsson-Strömberg U, Höglund M, Björkholm M, Braide I, Carlson K, Gahrton G, Grimfors G, Hast R, Lerner R, Linder O, Ljungman P, Löfvenberg E, Malm C, Nilsson P-G, Paul C, Rödjer S, Stenke L, Tidefeldt U, Treason I, Uren A- M, Whalin A, Vilene L, Invest I, Eternal O, Öberg G and Simonsson B for the Swedish CML group. Successful mobilization of Ph-negative blood stem cells with intensive chemotherapy + G-CSF in patients with chronic myelogenous leukemia in first chronic phase. Leuk Lymphoma. 2006 Sep; 47(9):1768-73. III Olsson-Strömberg U, Åleskog A, Björnberg A, Höglund M, Simonsson B, Bengtsson M, Barbany G, Larsson R and Lindhagen E. Imatinib activity in vitro in tumor cells from patients with chronic myeloid leukaemia in chronic phase and blast crisis. Anticancer Drugs. 2006 Jul; 17(6):631-9. IV Hagberg A, Olsson-Strömberg U, Wickenberg-Bolin U, Göransson H, Isaksson A, Bengtsson M, Höglund M, Simonsson B, Barbany G. Gene expression analysis identifies genetic signature potentially associated with response to alfa- IFN in chronic phase CML patients. Leukemia Research. 2007 Jan 3.

Contents Introduction 11 Chronic myeloid leukemia 11 History 11 Clinical presentation 11 Epidemiology and etiology 12 Diagnostic criteria 12 Prognostic scores 14 Treatment 16 Aim 16 Response 16 Chemotherapy 18 Interferon-alpha 18 Autologous stem cell transplantation 19 Allogeneic stem cell transplantation 19 Imatinib 21 Resistance 25 Dasatinib 25 New therapeutic options 26 Treatment of chronic phase 27 Treatment of accelerated phase and blast crisis 29 In vitro cytotoxicity assays 31 Methodological aspects 31 Use of primary cells in drug development 32 Prediction of individual patient response 32 Assays in chronic myeloid leukemia 33 Microarray expression analysis 34 Background 34 Data analysis and applications 35 Analysis in blood malignancies 36 Aims for the doctoral project 37 Material and Methods 38 Paper I 38 Study participants and data collection 38 Statistical analysis 38

Paper II 39 Study participants and data collection 39 Treatment evaluation 39 Statistical analysis 39 Paper III 40 Cell samples and cell preparation 40 Reagents and drugs 40 Fluorometric Microculture Cytototoxic Assay 40 Statistical analysis 41 Paper IV 41 Patient samples 41 Total RNA preparation 42 Array fabrication and analysis 42 Supervised Learning 42 RT-PCR validation 42 Results and Discussion 43 Comparison of busulphan, hydroxyurea and allogeneic stem cell transplantation (Paper I) 43 Successful mobilization of Ph-negative blood stem cells with intensive chemotherapy + G-CSF (Paper II) 45 In vitro activity of imatinib (Paper III) 48 Gene expression analysis in interferon- responders (Paper IV) 51 Conclusions 54 Paper I 54 Paper II 54 Paper III 54 Paper IV 54 Future directions 55 Acknowledgments 56 References 58

Abbreviations ACA ALL AML AP BC BCR BCR-ABL BSC CCgR CFU-GM CgR CHR CI CLL CML CMolR CP DA DiSC DMSO FDA EFS FISH FMCA G-CSF GVHD HR HU IFN LFS LOOCV MCgR MEA MMolR additional chromosome abnormality acute lymphoblastic leukemia acute myeloid leukemia accelerated phase blast crisis breakpoint cluster region breakpoint cluster region-abelson blood stem cell complete cytogenetic response colony-forming-units-granulocytemacrophage cytogenetic response complete hematologic response combination index chronic lymphocytic leukemia chronic myeloid leukemia complete molecular response chronic phase daunorubicin and Ara-C differential staining cytotoxicity dimethyl sulphoxide fluorescein diacetate event free survival fluorescence-in-situ-hybridization fluorometric microculture cytotoxic assay granulocyte-colony stimulating factor graft versus host disease hematological response hydroxyurea interferon-alpha leukemia free survival leave-one-out cross validation major cytogenetic response mitoxantrone, etoposide and Ara-C major molecular response

MNC MRD MTT OS PBS PDGFR PFS Ph Ph-negative Ph-positive PMN RD RT-PCR SCT TCK TRM URD WBC mononuclear cell minimal residual disease methyl-thiazol-tetrazolium overall survival phosphate buffered saline platelet-derived growth factor receptor progression free survival Philadelphia Philadelphia chromosome negative Philadelphia chromosome positive polymorphonuclear related donor real-time polymerase chain reaction stem cell transplantation total cell kill assay transplant related mortality unrelated donor white blood cell

Introduction Chronic myeloid leukemia History The word leukemia, derived from the Greek words leukos-white and haimablood, means white blood. Chronic myeloid leukemia (CML) was first recognized as early as 1845 [1, 2] when several cases of splenomegaly, anemia and massive granulocytosis were described. Neumann deduced that the disease originated in the bone marrow and proposed the term myeloid leukemia [3]. However, it was not until 1960 that the discovery of the Philadelphia (Ph) chromosome by Nowell and Hungerford [4] led to a better understanding of the pathogenesis of the disease. In 1973, Janet Rowley [5] showed that the Ph chromosome results from a reciprocal translocation between the long arms of chromosomes 9 and 22. The therapies in the 1960s were early splenectomy and radiotherapy over the spleen. The first reports of the use of busulphan and hydroxyurea (HU) were published in the 1960s by Haut and Galton [6, 7]. These drugs are still in use today. In 1983, Talpaz published the first report on the efficacy of interferon-alpha (IFN ) which is also still in use today [8]. Clinical presentation CML is a disease that originates in an abnormal pluripotent bone marrow stem cell [9]. The proliferating malignant clone expands and forces the normal hematopoietic cells aside. In blood and bone marrow there is a dominance of granulocytopoesis. The natural course on palliative treatment, with a median duration of approximately 3.5 years, goes from chronic phase (CP), followed by accelerated phase (AP), and finally, blast crisis (BC) where the leukemic process has acquired the characteristics of acute leukemia [10]. The great majority of patients are diagnosed in CP. A considerable percentage (20-40%) of these are asymptomatic, and their disease is diagnosed 11

after routine white blood cell counts (WBC) are obtained [11]. Some patients may have symptoms such as fatigue, weight loss, sweating and skeletal pain. At diagnosis, 50% of the patients have WBC counts >100 x 10 9 /l with a dominance of granulocytopoiesis, and 50% have thrombocytosis. Atypical presentation in BC at diagnosis occurs in roughly 2% of the cases. Epidemiology and etiology CML is a rare disease. The incidence in western countries is around 10 per million persons/year. The disease represents about 15-20% of all cases of leukemia [12, 13]. The number of new adult cases in Sweden varies between 70 and 90 per year. A male predominance is observed in many studies [14, 15]. Median age at diagnosis is 60 years, and 20% of the patients are less than 40 years old (Swedish CML group, unpublished). CML is rare in children. For the majority of patients, the etiology of CML is unknown. A recent report indicates a higher incidence of CML among clean-up workers after the Chernobyl catastrophe [16]. Individuals exposed to benzene or alkylating agents also have a higher incidence of CML, probably because of mutagenic effect [17, 18]. There does not appear to be an inherited disposition. Diagnostic criteria The CML diagnosis is based on morphological examination of blood and bone marrow, and detection of the Ph chromosome by cytogenetic or molecular analysis. The bone marrow in CML CP is hypercellular with a predominance of granulocytopoiesis. The morphologic features depend on the phase of the disease. The definitions of AP and BC are presented in Table 1 [19]. These have been used to distinguish CP from AP and BC in the most recent treatment reports and differ from those of the WHO, whose definition of AP also includes platelet count >1000 x 10 9 /l, increasing spleen size, increasing WBC count, lack of response to therapy and cytogenetic evidence of clonal evolution. The WHO definition of BC states 20% blast cells in blood or /and bone marrow compared to 30% in more recent publications [19]. 12

Table 1. List of differences between CP, AP and BC that have been proposed by an international expert panel [19]. CP AP BC - Blast cells in blood or bone marrow <15%. - Blast cells in blood or bone marrow 15-29%. - Blast cells plus promyelocytes in blood or bone marrow >30%, with blast cells <30%. - Blast cells in blood or bone marrow 30%. - Extramedullary blast cell proliferation. - Persistent thrombocytopenia (<100 x 10 9 /l) unrelated to therapy. - Large foci or clusters of blast cells in the bone marrow. - Basophils in blood >20%. The Ph chromosome results from a reciprocal translocation between the long arms of chromosome 9 and 22 (Figure 1). This translocation results in the head-to-tail fusion of the breakpoint cluster region (BCR) gene on chromosome 22 at band q11 with the Abelson (ABL) proto-oncogene on chromosome 9 at band q 34 [20]. In most cases of CML, the break occurs within the major breakpoint region of the BCR gene, that spans exons 12-16, and results in a fusion of either BCR exon 13 or 14 with ABL exon 2 (known as the b2a2 or b3a2 junctions, respectively). The fusion gene is translated into a fusion protein known as p210, because it has a molecular weight of 210 kda. The BCR-ABL fusion can be detected by fluorescence-in-situ-hybridization (FISH) and the fusion mrna by reverse-transcriptase polymerase chain reaction (RT-PCR) [21, 22]. The bcr-abl protein has an anti-apoptotic activity, and affects several signal transduction pathways that influence proliferation and adhesion of the leukemic cells. The protein is believed to play a central role in the development of CML [23, 24]. The Ph chromosome is found in 95% of patients with CML; the remaining cases have no cytogenetically visible Ph chromosome, but are positive for the BCR-ABL fusion, which is then masked either as a cryptic translocation or within a complex karyotype [13]. The presence of extensive deletions in derivate chromosome 9 (der (9)) have been identified in 10-15% of CML patients in CP [25]. These deletions have variable breakpoints, and they often include sequences from both chromosome 9 and 22. 13

Figure 1. The Ph chromosome results when a piece of chromosome #9 switches places with a piece of chromosome #22. The translocation forms an extra-long chromosome 9 (called der 9) and an extra-short chromosome #22, which is the Ph chromosome that contains the abnormal, fused BCR-ABL gene. Prognostic scores Two sets of prognostic factors can be considered: one set identified prior to treatment (baseline factors) and one during treatment (response-related factors). Baseline factors; Sokal score predicts survival for patients treated with busulphan or HU [26]. This score identifies three different risk groups: low, intermediate, and high. It is calculated from age, spleen size, blood platelet and blast cell counts. Twenty-three percent of the patients have high, 38% intermediate and 39% low Sokal risk score. The Hasford score predicts survival for patients treated with IFN and is calculated on 1303 patients from 14 different IFN clinical trials [27]. The Hasford score is based on age, spleen size and blood counts of platelets, eosinophils, basophils and blast cells. Three risk groups are identified. The low (40%), the intermediate (45%) and the high (15%) risk groups with median survivals of 98, 65 and 42 months, respectively. The Sokal and Hasford scores correlate with response, (but not survival) to imatinib (imatinib mesylate, Glivec ) treatment [21, 28, 29], which is considered as first line treatment in CML (page 21). The Gratwohl score (Table 2) expresses the risk for transplant related mortality (TRM) and the chance of leukemia free survival (LFS) after allogeneic stem cell transplantation (SCT) [30]. This score was calculated on 3142 patients transplanted in 1989-1997. The patients are scored from 0 to 7. 14

The 5-year survival ranges from 72% in risk group 0-1, to 22% in risk group 5-7. The Gratwohl score is simple, based on five main factors, and provides adequate risk assessment for treatment decisions. Table 2. Gratwohl transplantation risk score. The table lists the prognostic fctors and the corresponding risk score. Prognostic factors Age < 20 years 20-40 years > 40 years Risk score 0 1 2 Interval from diagnosis to trans- plantation 1 year > 1 year Disease phase CP AP BC Donor-recipient sex match Female donor and male recipient Any other match Donor type HLA-identical sibling Any other 0 1 0 1 2 1 0 0 1 Clonal cytogenetic evolution is defined as acquisition of cytogenetic abnormalities in addition to the Ph chromosome during the course of the disease. Only abnormalities that are demonstrable in at least two metaphases should be referred to as clonal. The tumor cells that carry these abnormalities could be Ph-positive or Ph-negative. Clonal cytogenetic evolution is often seen in AP and BC, and it is not clear whether the presence of this is an 15

adverse prognostic feature in the absence of other features of advanced disease. CML patients who bear der (9) deletions and have been treated with IFN have an adverse prognosis in terms of response to therapy and survival [25, 31]. However, there are few data regarding the impact of imatinib therapy on the outcome of CML patients with der (9) deletions [32, 33]. Response-related factors: Lately, factors related to therapy response seem to be of increased importance (page 22). Treatment Aim There are two treatment strategies in CML nowadays, remission inducing and palliative, respectively. Imatinib, other tyrosine kinase inhibitors, IFN and allogeneic SCT are used for remission induction, busulphan and HU for palliation. Response The definition of response levels (i.e. remaining tumor burden) is summarized in Table 3. Response to therapy is defined as hematological, cytogenetic or molecular. The majority of recent CML treatment reports define hematological (HR) and cytogenetic responses (CgR) in a uniform way, with only minor differences [29, 34, 35]. CML is a.o. caused by bcr-abl, an active tyrosine kinase. Imatinib is a small molecule that specifically inhibits the bcr-abl tyrosine kinase, and is recommended for up-front treatment of CP CML patients. With the introduction of imatinib, the cumulated rate of complete cytogenetic response (CCgR) in treated patients is high; therefore molecular methods for measuring minimal residual disease (MRD) (Figure 2) are required. RT-PCR has emerged as the method of choice for monitoring MRD in patients in CCgR. The status when no residual BCR-ABL transcripts can be detected with the most sensitive available methods is referred to as complete molecular response (CMolR). The use of the log reduction terminology has caused some confusion, as it implies the value is a relative. For this reason, it has been proposed to introduce a standardized numerical International Scale expressing the amount of BCR-ABL mrna as a percentage of a control gene and anchored to two absolute values, based on validated reference materials of known value [22]. The degree of molecular response is an important prognostic factor. The goal for imatinib treatment is at least a 3-log reduction 16

in BCR-ABL level. In this patient group the relapse rate at 54-months was only 3%, and no progression to AP or BC [29]. A major molecular response (MMolR) after 12 months of imatinib treatment is associated with even better event-free survival (EFS) and progression free survival (PFS) [21]. Patients with less than complete hematological response (CHR) at 6 months, less than partial cytogenetic response at 12 months, and less than CCgR at 18 months treatment with imatinib have a less good prognosis [21, 29]. A rise of BCR-ABL transcript level has been consistently associated with mutations [36]. In CP patients, there is a good correlation between BCR-ABL mrna levels in blood and bone marrow. Therefore, RT-PCR analyses can be performed on blood samples, which is much more convenient for the patient [37]. Table 3. Definitions of treatment response. Level of response Definition Complete hematologic response (CHR) WBC <10x10 9 /l, platelet <450x10 9 /l and no palpable spleen Minimal cytogenetic response (Minimal CgR) 66-95% Ph-positive metaphases * Minor cytogenetic response (MinorCgR) 36-65% Ph-positive metaphases * Partial cytogenetic response (PCgR) 1-35% Ph-positive metaphases * Complete cytogenetic response (CCgR) 0% Ph-positive metaphases * Major cytogenetic response (MCgR) 0-35% Ph-positive metaphases * Major molecular response (MMolR) Complete molecular response (CMolR) BCR-ABL ratio <0.1 International Scale Negativity by RT-PCR *Based on the analysis of at least 20 metaphases. 17

10 12 Diagnosis, Pretreament or Hematologic Relapse 100 Number of leukemic cells 10 11 10 10 10 9 10 8 10 7 10 6 Complete Hematologic Response Complete Cytogenetic Response Major Molecular Response Undetectable transcript (Complete Molecular Response) 10 1 0.1 0.01 0.001 0.0001 BCR-ABL ratio (according to the International Scale) Figure 2. Approximate relationship between number of leukemic cells and level of BCR-ABL transcripts. Chemotherapy Palliative chemotherapy is mainly based on HU and busulphan. These drugs are cytotoxic antiproliferative agents that are administrated orally, and they are well tolerated. HU was introduced for use in CML 1966 [38]. Patients treated with HU survive around 10 months longer than those on busulphan [39]. Currently, HU is used to quickly reduce the myeloid hyperplasia, or as long time palliative treatment. HU has very few side-effects, all of which are reversible on stopping treatment. Nausea and anorexia are usually transient. Atrophy and scaling of the skin and partial alopecia may occur. Busulphan is an alkylating agent. It was one of the first drugs reported to be effective for hematological control in CML [40]. Side-effects of busulphan include severe prolonged myelosuppression. An uncommon side-effect is a severe idiosyncratic pulmonary reaction, a pronounced interstitial fibrosis, known as busulphan lung, which may be lethal [41]. Busulphan is not much in use as CML therapy nowadays. Interferon-alpha IFN was one of the first drugs that induced cytogenetic response in CML. The most important goal with IFN therapy is to reduce the tumor burden and thereby prolong time to BC [15, 42]. Not all patients respond to IFN 18

and the reason for this is largely unknown. The mechanism underlying the effect of IFN in CML is still poorly understood. In addition to the role in innate immunity, IFN modulates immune response and exerts antiproliferative effects in some cell types [43, 44] and it has been suggested that differences in immunological responses elicited by IFN underlie the individual differences in response [45, 46]. IFN -based regimes were the treatment of choice for more than 10 years: several randomized prospective studies [15, 47-50] and a meta-analysis indicate that IFN prolongs survival compared to palliative chemotherapy. The latest updates of the larger IFN studies [51] report a 10-year overall survival (OS) ranging from 27 to 53% for all treated patients. IFN induces CHR in about 80% of patients, CgR in about 50%, and major cytogenetic response (MCgR) in about 30%. A substantial survival prolongation is obtained only in cytogenetic responders. About 10% achieve a CCgR after 12 months treatment and here the 10-year survival is 70-80% for low and intermediate risk group patients. The optimal IFN dose is unknown. A multicentre trial comparing low dose (3 milj IE/m 2 x 1 sc, 5 days weekly) vs. high dose (5 milj IE/m 2 x 1 sc, seven days weekly) IFN indicate that the low dose was as effective as, and better tolerated than the higher dose [52]. No significant differences in OS, PFS, CHR or MCgR were observed. As a palliative treatment, IFN is used alone, or in combination with HU or Ara-C. Autologous stem cell transplantation The role of autologous SCT for treatment of CML has been subject to a number of studies and reviews for more than 20 years [53, 54]. Several observations indicate that the procedure could be useful for achieving CgR and for prolonging survival [55]. However, a meta-analysis of six trials, in which patients were randomly allocated to autologous SCT or an IFN -based regimen, did not prove advantageous for survival for autologous SCT [56]. Despite favorable PFS in imatinib-treated patients, there is still uncertainty about the durability of response. Maybe, in the future, cryopreserved autologous blood stem cells (BSC) collected at CCgR may provide an attractive autograft option either as salvage or consolidation. In this era, several different strategies are demonstrated to mobilize Ph-negative stem cells in 40-80% of the patients [57-60]. Allogeneic stem cell transplantation The first report, 30 years ago, of a successful transplant of bone marrow from twin donors in patients with CML marked the beginning of a new era in the treatment of this disease [54, 61]. The concept was rapidly followed with transplants from siblings and was soon confirmed in a series of more than 100 patients by the International Bone Marrow Transplant Registry 19

[62]. The number of transplants increased rapidly up to 1999 and then began to decline. In the beginning of this transplant era, there was a clear consensus for early transplantation [63]. Early transplant is no longer the first choice management for CML [64]. This is also reflected by an increase in transplants performed at later stages of the disease [65]. The indication for using allogeneic SCT in CML has changed since the introduction of imatinib. Clearly, the general recommendation that patients with CML and an HLA-identical donor should be transplanted within the first 12 months after diagnosis can no longer be upheld. There is an ongoing discussion whether any subgroup of patients with newly diagnosed CP should be treated with allogeneic SCT as primary therapy or not. It has been proposed that patients classified as high risk Sokal and good risk Gratwohl should be transplanted upfront. In practice, almost all patients receive imatinib as first line therapy. Allogeneic SCT is the second line treatment for patients with a suitable donor. The longest follow-up of patients who received transplants from a matched-related donor is that reported by the European Group for Blood and Marrow Transplantation (EBMT) between 1980-1990. OS at 20 years was 34% for all patients and 41% for patients in first CP [66]. In the cohort transplanted between 2000 and 2003, probability of survival at 2 years was 61 % and for the subgroup transplanted in first CP with an HLA-identical sibling 74% [66]. TRM is considerable and ranges beween 15 and 60% [67]. However, relapses were seen up to 21 years after treatment. The usefulness of the previously described main pre-transplanted risk factors for outcome [30] (Gratwohl score) were confirmed in this analysis. Survival was better for patients transplanted in CP than those transplanted in later stages or BC, due to increased transplant-related mortality and relapse rate among patients with more advanced disease. Long-term survival after allogeneic SCT in BC is <10%. Progress in molecular DNA typing of HLA alleles, management of opportunistic infections and supportive care, as well as conditioning regimes and immunosuppressive therapy, have contributed to improved outcome after allogeneic SCT [68]. For CML patients receiving conventional transplantation, the use of blood stem cells has not proved better than the use of marrow stem cells [69]. Reduced intensity conditioning is currently being evaluated for CML [70-73] and may permit allogeneic SCT also in fragile patients. The long-term impact on OS, EFS and quality of life cannot yet be assessed. Relapse after allogeneic SCT is still a problem. Donor lymphocyte infusion has in this situation been the standard therapy. It has been shown to restore full donor chimerism and produce long-term CMolR [74]. However possible side effects, such as graft versus host disease (GVHD) and myelosuppression lead to a considerable amount of morbidity and mortality. Imatinib as treatment for relapse, has shown remarkable short term efficacy 20

combined with low toxicity [75, 76]. A recently published study comparing imatinib and donor lymphocyte infusion treatment at relapse, [77] showed that imatinib treatment resulted in a higher incidence of relapse. Imatinib The BCR-ABL oncogene is thought to play a key role in the patogenesis of CML [20, 23, 24]. It encodes a chimeric bcr-abl protein having consistutively activated abl tyrosine kinase activity, and this is thought to be the underlying cause of CML. The first preclinical studies in 1996 showed apoptotic and antiproliferative effects on BCR-ABL positive cells by blocking bcr-abl-initiated signalling pathways [78, 79]. The imatinib (imatinib mesylate, Glivec, Novartis) molecule binds to the ATP-binding pocket of the abl domain and stabilizes the inactive, non-atp-binding conformation of the bcr-abl tyrosine kinase [80, 81] (Figure 3). This blocks tyrosine kinase activity and inhibits both bcr-abl mediated autophosphorylation and substrate phosphorylation, resulting in abrogated downstream cell signalling and reduced proliferation of the BCR-ABL-positive cells [78, 79, 82]. Although imatinib is highly specific for abl, the c-kit, platelet-derived growth factor receptor (pdgfr) and arg tyrosine kinase activities are also suppressed [83], which has been suggested to be part of the inhibitory function of imatinib on CML cells [84]. Figure 3. A schematic view of the mode of action for imatinib. a) BCR-ABL signalling without imatinib and b) with imatinib. Imatinib blocks the ATP-binding site and thereby inhibits tyrosine phosphorylation. 21

Early studies into bone marrow changes associated with imatinib therapy in CML patients suggested normalization of bone marrow, without any morphologic evidence of CML [85, 86]. In recent studies, long-term (mean 21 months) therapy with imatinib produced marked changes in bone marrow histopathology in most CML patients who respond to therapy [87]. Decreases in bone marrow cellularity, neutrophil granulopoiesis and the number of atypical micromegakaryocytes are accompanied by an increase in erythroid precursors and lymphoid nodules [87]. Imatinib is easy to administer orally, and has relatively few side effects. The drug is rapidly absorbed, reaching a maximum plasma concentration 2-4 hours after oral administration. It is metabolized by the liver, mainly via the cytochrome P450, and primarily eliminated via feces. Imatinib induces a much higher rate of CCgR in CP than any other therapy (except allogeneic SCT) [29]. The superiority of imatinib 400 mg daily over IFN and low dose Ara-C was established in a prospective randomized study of CP patients, the IRIS study [29] (International Randomized Study of Interferon and STI571). Patients were randomized to receive imatinib 400 mg/day or IFN ; plus low dose Ara-C. After 60 months treatment, PFS was 93%, OS was 90% and survival freedom from progression to AP/BC, as well as from hematologic or cytogenetic relapse, was 84% [29]. This ongoing study has been continuously updated and has shown thus far that response to imatinib is a most important response related prognostic factor (Table 4). The probability of PFS is strongly correlated with the level of response. Cytogenetic response seems to be the most important response-related prognostic factor. If no CgR is achieved after 3 months, there is still a 50% chance of achieving a CCgR later. If after 12 months of treatment, the CgR is partial, the probability of achieving a CCgR at 2 years is still 50%, but if the response is less than partial, the chance is less than 20%. The issue of the optimal imatinib dose is not yet settled, and the standard dose is 400 mg to CP patients. Circumstances permitting an increase in dosage from 400 to 600 mg/day, or from 600 to 800 mg/day, include disease progression and failure to achieve satisfactory HR and CgR. Results suggest that higher doses of imatinib, 800 mg/day, may be associated with faster onset and higher rate of cytogenetic response [89, 90]. The long-term outcome is not yet known. 22

Table 4. Relationship between the degree of early CgR, the CCgR rate at 2 years,and EFS at 42 months in IRIS study [88] Time of treatment and cytogenetic response 3 months Partial Minor Minimal/none Probability of CCgR at 2 years, % 90 60 50 EFS at 42 months, % Not available Not available Not available 6 months Complete/partial Minor or minimal/none Not applicable/80 50/15 95 75 12 months Complete/partial Minor,minimal, or none Not applicable/50 < 20 An ongoing question is whether CML patients should be treated with imatinib for the rest of their life. An important unanswered question for patients who achieve a CMolR is whether imatinib therapy may be discontinued safely. Only limited data are available regarding this, and the results are divergent [91-93]. In a recently published study 12 patients with undetectable residual disease for more than 2 years were included [91]. Six patients displayed a molecular relapse within 5 months, but when imatinib was reintroduced this led to a novel molecular response in most patients. Six patients still had an undetectable level of BCR-ABL transcript after a median followup of 18 months. In another report with three patients, all relapsed shortly after discontinuation of therapy [92]. A small fraction of CML patients show primary resistance to imatinib, and responding patients may later develop secondary resistance and relapse. 23

Combining imatinib with other antileukemic agents may be one way to further increase the remission rate and to prevent outgrowth of resistant clones. In in vitro studies, different drugs are shown to overcome imatinib resistance or to synergize with the effect of imatinib. These include leptomycin B, proteasome inhibitors, mtor-inhibitors, arsenic trioxide, myophenolic acid, farnesyl-transferase inhibitors, bryostatin, decitabine, histone deacetylase inhibitors, homoharringtonine and phosphoinositol-dependent kinase -1 inhibitors [94-100]. However, results are still preliminary and limited [101]. Intermittent exposure of CML cells to granulocyte-colony stimulating factor (G-CSF) enhances the effect of imatinib on CML cells [102]. Results of single agent imatinib in the management of AP and BC are less impressive than results achieved in CP [103, 104]. Some patients in AP appear to obtain a major benefit from treatment with imatinib. The responses for the majority of patients treated in BC are not maintained over a long time. Thus, for patients with advanced phases, imatinib should mainly be used only as a bridge to an allogeneic SCT. In patients with AP and BC, an initial dosage of 600 mg/day is recommended. Treatment with imatinib prior to allogeneic SCT does not influence transplant related morbidity and mortality [105, 106]. Side-effects Imatinib is generally well tolerated. Hematological adverse events are neutropenia, thrombocytopenia and anemia. This toxicity is managed by reducing the imatinib dosage or discontinuing treatment until the problem is resolved. Sometimes it is necessary to administer G-CSF. The the most common non-hematological toxicities seen with imatinib treatment are edema (60%), nausea (50%), muscle cramps (49%), diarrhea (45%), skin rash (40%) and fatigue (39%). The most common manifestation of edema is the periorbital edema that is typically worse in the morning, and no specific therapy is required for most cases. Some patients have found that limiting salt intake helps to control edema, and in severe cases, diuretics may be indicated. Musculoskeletal complains are another common side effect of imatinib and are manifested as muscle cramps and bone pain. Calcium and magnesium supplementation often provides symptomatic relief, even in patients with normal serum levels of ionized calcium and magnesium [107]. Most cases with skin rash, is mild, self-limiting, and easily manageable with antihistamines or topical steroids, whereas a short course of oral steroids can be used to treat more severe cases [29]. A recently published study [108] indicates advanced heart failure could be a side effect in imatinib-treated patients. A follow-up examination of Novartis database found however that the incidence of advanced heart failure was only 0.2% per year [109]. Another small study reported the possible development of hypophosphataemia with associated changes in bone and mineral 24

metabolism in some patients receiving imatinib [110]. The long-term side effects are not yet known. Resistance Primary hematologic resistance, defined as a failure to obtain a CHR despite therapeutic doses of imatinib, occurs in approximately 5% of the patients. Secondary or acquired resistance refers to loss of previously established response. Although little is known about the molecular mechanisms responsible for the relatively rare cases of primary hematologic resistance, the mechanisms of secondary resistance are better understood. Secondary resistance may be multifactorial, including BCR-ABL mutations of the kinase domain interfering with imatinib binding, BCR-ABL amplification or overexpression, clonal evolution or decreased imatinib bioavailability. The frequency of BCR-ABL mutations in resistant patients was reported to range from 42 [111] to 90% [112] depending on the methodology of detection, the definition of resistance, and the phase of the disease. Currently, more than 40 different mutations are associated with clinical resistance to imatinib [113-115]. The T315I mutation, and some mutations affecting the so-called P-loop, are associated with poor prognosis [116]. Other mutations can be overcome by a dose increase, and some mutations are functionally irrelevant. Mutations are more frequent in AP/BC, and are in CP often identified in patients with more than 2-fold increase of the BCR-ABL transcript levels, rather than in those with stable or decreasing levels. Thus, screening for mutations in the kinase domain of bcr-abl can provide useful information for guiding therapeutic decisions [113-115], and need to be interpreted within the clinical context. It has been postulated that cell survival mechanisms operating independently of BCR-ABL may be responsible for many cases of primary imatinib resistance, although the understanding of potential mechanisms of this BCR- ABL-independent resistance remains limited. Dasatinib The options besides allogeneic SCT to induce remissions are limited in patients who have failed imatinib treatment. Dasatinib (Sprycel, Bristol- Myers Squibb) is a novel oral kinase inhibitor that binds to both the active and inactive conformations of the ABL kinase domain. It targets both the bcrabl and the Src kinases. In preclinical studies, this compound is 300-fold more potent than imatinib, and harbors potent inhibitory activity against nearly all imatinib resistant mutants tested [117, 118], except for the T315I mutation. In 2003, the first cases to be treated with dasatinib were CML CP patients with hematologic resistance or intolerance to imatinib. A recently published study [119] showed CHR in 92% and MCgR in 45% of cases: 25

these responses were maintained in 95% of patients for a median follow-up time of more than 12 months. Myelosupression was a common, but not doselimiting adverse event. Non-malignant pleural effusions developed in 13% of the patients, and were usually successfully managed with diuretics, pleurodesis and glucocorticoides. The registered dosage is 140 mg/day administered orally in two divided doses [120]. New therapeutic options Nilotinib Nilotinib (AMN-107, Novartis) is another new oral kinase inhibitor with a 25-fold increased in vitro potency, relative to imatinib. As with imatinib, nilotinib binds to the inactive conformation of ABL. This drug harbors activity against most imatinib-resistant cell lines with mutant ABL kinases, but it has no effect on the T315I mutation [118, 121]. Unlike dasatinib, nilotinib does not inhibit Src family kinases. A recently published study [122] with nilotinib-treated imatinib-resistant CML CP patients showed CHR in 11/12 and MCgR in 6/12 patients. Common adverse events were myelosuppression, transient indirect hyperbilirubinemia and skin rashes. Bosutinib Bosutinib (SKI-606,Wyeth) is, like dasatinib, a dual Src-Abl inhibitor active against BCR-ABL-positive CML cell lines [123]. Resistance to imatinib is not entirely explained by bcr-abl mutations, and in vitro experiments have shown bosutinib activity in models where resistance was not caused by mutations [124]. In contrast to imatinib, no substantial inhibition of c-kit and pdgfr was noted with bosutinib, thus placing this molecule in a class of its own. The difference in selectivity could, compared with imatinib, result in fewer side effects since much toxicity associated with imatinib can be tracked to the inhibition of pdgfr and/or c-kit. Phase I-II clinical trials are ongoing. MK-0457 MK-0457 (Merck) is a novel kinase inhibitor with preclinical antileukemia effect. The T315I bcr-abl mutation mediates clinical resistance to imatinib, nilotinib and dasatinib. MK-0457 has in vitro activity against cells expressing mutated BCR-ABL, including the T315I mutation [125]. The first patients with CML and Ph-positive ALL and T315I mutation have been treated, and they have achieved responses [126]. These cases are the first reported to respond to this kinase inhibitor. 26

LBH589 Treatment with the histone deacetylase inhibitor LBH589 (Novartis) depletes BCR-ABL levels in vitro. Co-treatment in vitro with LBH589 and nilotinib is active against imatinib-resistant CML cells, including those with expression of BCR-ABL mutation T315I [127]. A phase I study is ongoing. Vaccination Despite the fact that imatinib is the most effective debulking therapy for CP patients, the eradication of residual disease without allogeneic SCT still seems a difficult goal for the tyrosine kinase inhibitor approach alone [128]. An alternative attempt to target the BCR-ABL gene and its derived p210 fusion protein is an active specific immunotherapy (eg, a vaccine). Sixteen CP CML patients with a stable residual disease were injected s.c 6 times with a peptide vaccine derived from the sequence p210-b3a2 (CMLVAX100). Vaccinations with CMLVAX100 were followed by an antitumor effect in most patients [129]. All patients cytogenetic responses improved after six vaccinations. Treatment of chronic phase Standard (non-investigational) treatment of CP includes HU, INF ± low dose Ara-C, imatinib and allogeneic SCT. In coutries where imatinib is available, and standard allogeneic SCT is feasible, we are now in the rather privileged situation of having two potent strategies that are both established. Imatinib is currently preferred as the initial treatment. In patients with Sokal high risk and low Gratwohl score, the choice between imatinib and allogeneic SCT should be discussed. The comparative 5-year survival for different treatments is presented in Table 5. Table 5. Approximate 5-year survival in CML patients treated with busulphan, HU, IFN, allogeneic SCT and imatinib. The figures are based on references cited above. Treatment 5-year survival Busulphan-HU 30 % IFN 27-53% Allogeneic SCT 50-60 % Imatinib 90 % 27

The motives for treatments other than imatinib are intolerance or toxicity, failure, suboptimal response and warnings (definitions are presented in Table 6). In the case of imatinib intolerance or toxicity, the treatment choices are allogeneic SCT, dasatinib, INF ± low dose Ara-C, or investigational trials with new agents. In the case of failure, the first choice should be allogeneic SCT or doseescalation of imatinib to 600 or 800 mg daily, provided that the patient tolerated 400 mg and that resistance to imatinib was not associated with the presence of a BCR-ABL mutation with a high level of insensitivity to imatinib. Dasatinib treatment is also an option, or investigational agents such as nilotinib [119, 122]. In the case of suboptimal response, the first choice could be doseescalation of imatinib to 600 or 800 mg. In this case, allogeneic SCT could be offered to patients with a low or intermediate Gratwohl risk score. In patients presenting warning features, standard treatment is still imatinib 400 mg, but any warning should alert the clinician that the patient may become eligible for imatinib dose escalation, for allogeneic SCT, or in selected cases for investigational agents [19]. There are of course several other possible scenarios. If the patient requiring imatinib reduction or frequent treatment discontinuations, due to myelosuppression or side effects, the recommendation is to continue the 400-mg dose insofar as possible. Appropriate supportive care should be provided, including G-CSF and erythropoietin. Monitoring of blood imatinib concentration is not required, but is desirable in case of failure, especially in patients who need to take drugs interfering with cytochrome P 450, and in those experiencing a severe drug-related adverse event [130]. 28

Table 6. Definition of failure, suboptimal response and warnings for upfront CML patients who are treated with imatinib 400 mg daily [19]. Time Failure Suboptimal response Diagnosis NA NA Warnings Sokal high risk Del(9q) ACA in Ph+ cells 3 months No HR (stable disease or disease progression) Less than CHR 6 months Less thanchr No CgR Less than PCgR 12 months Less than PCgR Less than CCgR Less than MMolR 18 months Less than CCgR Less than MMolR Any time Loss of CHR Loss of CCgR Mutation ACA in Ph+cells Loss of MMolR Mutation Any rise in transcript level Other chromosome abnormalities in Ph-negative cells Treatment of accelerated phase and blast crisis As described before, the natural course on palliative treatment goes from CP followed by transformation into AP, and finally BC. BC may be myeloid or lymphoid, the former being twice as common. The pathophysiology of CML in transformation is incompletely understood. Many patients show additional cytogenetic abnormalities (ACA) and it is reasonable to hypothesize that clonal evolution plays a role. The most common ACA includes duplication of the Ph chromosome, trisomy 8 and iso-chromosome 17; however, these abnormalities are also observed in patients with earlier stages of CML [131, 132]. Since many patients now are treated with imatinib or other tyrosine 29

kinase inhibitors, this change in therapeutic approach almost certainly affects the kinetics and molecular phenotype of CML in transformation. Historically, therapy for patients with BC has been disappointing, with response rates far lower than achieved with standard induction regimens for de novo acute myeloid leukemia (AML) or acute lymphoblastic leukemia (ALL). The recently available bcr-abl tyrosine kinase inhibitors nilotinib and dasatinib, broaden the options for patients developing BC on imatinib. Ideally, the choice of therapy should be based on screening for imatinibresistant BCR-ABL mutations, as both dasatinib and nilotinib are active against most of these, except for T315I [119, 122]. With dasatinib, HR is seen in 31 of 44 patients with AP, BC or Ph-positive ALL (imatinib resistant) [119], the rate of MCgR is 25%. At six months responses were 82% for AP patients, but nearly all patients with lymphoid BC and Ph-positive ALL have a relapse. Responses occurred among all BCR-ABL genotypes, with the exception of the T315I mutation. The major adverse effects of dasatinib are myelosuppression and non-malignant pleural effusions [119]. In the nilotinib study [122], 39% achieved HR, and 27% CgR among the BC patients. In AP, the corresponding figures were 71% and 47%, respectively. Common adverse events included myelosuppression and transient indirect hyperbilirubinemia [122]. It is anticipated that these drugs will be useful in patients whose BC is driven by imatinib-resistant BCR-ABL mutants or increased levels of BCR-ABL, which may comprise 30-40% of cases [133]. For more traditional chemotherapeutic agents, high-dose Ara-C-based regimens are used for the treatment of myeloid BC, which in the preimatinib era had a worse prognosis than the lymphoid form [131]. CHR rates have ranged from 20% to 30%. Thus far, no high-intensity chemotherapy regimen has demonstrated clear superiority, and durability has been poor, with median survivals of only a few months. For CML in lymphoid BC, ALL-like induction regimens have been most commonly used, including the hyper-cvad regimen [134]. As the response rates and durability have been disappointing, various new agents are being investigated for their capacity to induce remissions in CML BC. They include inhibitors of the mammalian target of rapamycin (MTOR) pathway; anti-vascular endothelial growth factor (VEGF)-based agents; histone deacetylase inhibitors; and other novel small molecule inhibitors [135]. As it is the only established curative modality, allogeneic SCT remains the ultimate salvage therapy for patients with imatinib-resistant CML. Patients in the appropriate age range should be HLA-typed at diagnosis so that the options for related donor (RD) or unrelated donor (URD) allogeneic SCT are known early in the disease course. Ideally, CML BC can be prevented with BCR-ABL inhibition. For those patients who progress despite this, the establishment of an early warning system for CML progression could allow initiation of more aggressive or alternative therapies at a stage where inter- 30

ventions have the optimal chance of being clinically meaningful. The fear is that patients will progress to BC too rapidly to allow sufficient time for allogeneic SCT or other investigative strategies to be implemented. Long-term disease-free survival /cure rate is about 10% for patients transplanted while still in BC [136]. Outcome is somewhat better for patients transplanted after achieving a second CP with salvage. In vitro cytotoxicity assays Cell culture assays for measurement of cytotoxic drug resistance in leukemia were used as early as in the beginning of the last century. In 1917 Pappenheimer added trypan blue to fresh thymic lymphocytes that had been exposed to toxic agents [137]. Trypan blue stained the dead cells blue, while the living cells remained clear. In 1954, Black and Spear compared the clinical outcome with the in vitro response of patient tumor cells [138]. Today, in vitro cytotoxicity assays are important tools for pre-clinical development of new anti-cancer agents. Basically, the leukemia cells are exposed in vitro to different cytotoxic drugs and the effect of the drugs on survival or proliferation of the cells is investigated. Methodological aspects There are two main groups of in vitro assays, total cell kill assays (TCK) and cell proliferation assays (clonogenic assays and short-term assays). The clonogenic assays are based on single cell culture in soft agar and subsequent counting of formed colonies, 2-3 weeks after exposure to cytotoxic drugs. The assay relies on the proliferative capacity of a small fraction of cells, presumed to be stem cells. The test is laborious and time-consuming. Total cell kill assays are based on the concept of total tumor cell kill in the whole, mainly non-dividing, tumor cell population. These assays are short-term assays (2-4 days). The different TCK assays are in many aspects comparable, but the main differences are the various techniques used to estimate the proportion of viable cells after incubation with cytotoxic drugs, and the different drug exposures used. Examples of these forms of assay are methyl-thiazol-tetrazolium (MTT) assay, differential staining cytotoxicity (DiSC) assay and Fluorometric Microculture Cytototoxic Assay (FMCA). The MTT assay and FMCA have many similarities. The FMCA was developed in Uppsala in the late 1980s, and is based on the uptake of fluorescein diacetate (FDA) by viable cells and its conversion to fluorescent fluorescein after exposure to cytotoxic drugs in primary cultures [139]. The end point for 31

recognizing cell death is loss of cell membrane integrity. The generated fluorescence is linearly related to the number of living cells. It is believed to measure cell death in the total population of tumor cells. Use of primary cells in drug development Tumor cell lines are the most commonly used model systems in cancer drug development, but the use of primary tumor cells from patients has gained increasing focus. The primary tumor cells from patients might predict drug activity patterns better than tumor cell lines, whereas the latter provide a practically unlimited source of samples facilitating multiple tests [140]. In vitro cytotoxicity assays can be used in several different ways. In preclinical development of new cytotoxic agents the assay can evaluate drug sensitivity patterns for groups of patients with different diagnoses [141-143] or subgroups, for example regarding genetic aberrations [144-146]. Combinations of cytotoxic agents can be evaluated regarding antagonistic or synergistic effects and can be the basis of future phase I and phase II trials. On primary tumor cells from patients, FMCA has been used to differentiate the in vitro drug effects between different diagnostic subgroups in chronic lymphocytic leukemia (CLL), AML and ALL [142, 144, 147]. Prediction of individual patient response Human cancers are heterogeneous diseases with a diversity of pathological processes, cytotoxic drug sensitivity and resistance mechanisms, and there is a considerable variation in tumor cell sensitivity to cytotoxic drugs between patients with similar histological tumor types. Many groups have investigated the TCK assay as a method for improving disease treatment in individual patients, and the results have correlated well with the subsequent patient response and survival for a variety of diseases and drug regimes [148-152]. A meta-analysis, based on published papers from 1983-1997 and reporting data on 1596 correlations between drug sensitivity in vitro and clinical response, concluded that TCK performed well, with an overall sensitivity and specificity of about 0.9 and 0.7, respectively (www.weisenthal.org). For the individual patient, the assay can also be useful in giving prognostic information [153]. The American Society of Clinical Oncology (ASCO) has established a Working Group to evaluate the potential role of chemosensitivity assays in clinical practice. This evaluation suggests that no single assay is ready for routine integration into the clinical setting [151], which reflects problems in technical success of the assays, and a lack of adequate prospective evaluation in clinical trials. This conclusion has however gained criticism. 32

Assays in chronic myeloid leukemia In most in vitro studies on primary tumor cells from CML patients, clonogenic assays are used, but this method is time-consuming [154, 155]. The drug effects are measured only in a small fraction of proliferating cells. The effects on resting cells, which might have proliferative potential in vivo, are not measured. Patients treated with imatinib to CCgR have detectable malignant primitive progenitors [128]. The primitive quiescent stem CML cells (CD34 +,CD38 - ) have previously been shown to be insensitive to imatinib in vitro, and represent less than one percent of total CD34 + cells present at diagnosis, and may explain MRD in patients. Therefore, many laboratories use methods for enriching primitive CML stem cells (CD34 +,CD38 - ) [156]. A critical issue when dealing with primitive CML populations is the coexistence of Ph-negative stem cells. The FMCA, used in this paper, is a TCK assay and is believed to measure cell death in the whole tumor cell population. This method is easier to perform and less time-consuming than the clonogenic assay, but requires a fairly pure tumor cell population, since it cannot distinguish between normal and malignant cells. A tumor cell fraction exceeding 70% is a quality criterion normally used for the FMCA, and this is possible in newly diagnosed CML patients. The need for a high proportion of Ph-positive cells in the cell preparation currently makes testing of samples from drug-treated patients difficult, as the separation of malignant cells from normal granulocytes is complex. This far most in vitro studies are performed on CML cell lines, which originate from patients in BC [157]. However, established cell lines may have lost some of their original tumor characteristics. Imatinib resistance is a clinical problem in the treatment of the patient, and mutations within the ABL kinase domain are a major cause. Therefore cell lines transfected with different mutations are used, for example Ba/F3, to develop drugs effective against specific mutations [117]. The use of in vitro assays based on primary tumor cells from patients might be an attractive alternative. Some studies also use a small number of patient samples [154, 155]. An important difference in using cell lines versus primary cells is that tumor cell lines are generally highly prolific, which is not the case for primary tumor cells. The primary patient cells generally share more properties with the tumor cells to be treated in the clinical situation. Primary tumor cells from patients more adequately reflect the diseasespecific activity of new compounds, and may therefore be more relevant than cell lines when the effect of new drugs on different tumor types is studied [141]. 33

Microarray expression analysis The past decade has witnessed a revolution in the field of expression profiling in cancer. Microarray technology has changed both the view of cancer and how research in molecular oncology can be conducted [158]. The sequencing of the human genome [159, 160] and technological refinements allow for interrogation of the whole transcriptome (all expressed genes) in a single experiment. The challenge is to use the generated information for developing new diagnostic and prognostic indicators, and to identify new targets for therapeutic intervention. Background Microarray technology allows parallel analysis of the expression of thousands of genes at a fixed point [161], and relies on the inherent ability of single-stranded DNA to base pair with a complementary sequence [162]. Microarrays can be fabricated either by mechanical deposition of DNA or by de novo synthesis of oligonucleotides on a solid support [163, 164]. In spotted arrays, the probes are either pre-synthesized oligonucleotides or PCR products generated from cdna clones. Spotted arrays are typically hybridized with cdna from a sample and a reference (e.g. from patient and healthy control) labelled with two different fluorophores. Briefly, two different fluorescent dyes are used to label the test and reference sample, which are then allowed to hybridize to known single stranded cdna or oligonucleotide probes immobilized on a glass slide (Figure 4). As hybridization occurs competitively, the amount of bound target is proportional to the level of expression [161, 162]. Following hybridization and washing, the slides are scanned to measure the emission of the two dyes and the outcome is presented as a colour image of the spots, with red and green representing up- and down-regulation of the corresponding gene in the interrogated sample relative to the reference sample. The expression of each gene is then quantified as the ratio between the intensity values of the test and reference sample. Oligonucleotide expression arrays from commercial sources (e.g. Affymetrix, Agilent and GE Healthcare), are typically onecolor-systems. In these systems, only one sample is interrogated at a time (i.e. without reference). Thus, these microarrays provide an estimate of the absolute values of gene expression, rather than ratios, as is the case with spotted arrays. 34

Image processing and subsequent data analysis are crucial for the extraction of useful information from microarray experiments. The raw data generated from scanning have to be filtered and normalized to remove noise and systematic variation. The goal of pre-processing is to exclude unreliable data and enable comparisons between individual experiments (recently reviewed in [165]). Figure 4. Brief overview of microarray technology. In this two-color-analysis, RNA samples obtained from patients and reference subjects are individually labelled and hybridized to a single DNA microarray consisting of individual gene-specific probes. Finally providing a color image with red and green representing up- and down regulation of the corresponding gene in the patient relative to the reference sample. Yellow spots designate genes with an equal expression level in the test and reference sample. Data analysis and applications Supervised and unsupervised analyses are two commonly used but conceptually different approaches for acquiring information from microarray experiments [166]. In supervised analyses, information about tumor class is used to identify gene combinations specific for particular conditions e.g for cancer subtype differentiation. The term class refers to a characteristic shared by a group of samples, but not the other samples, for example responders to treatment and non-responders. Initially, a training dataset is used to identify a combination of genes (a classifier) that can separate the different classes. The classifier is subsequently evaluated on samples not used in the training procedure, i.e. the test set. This approach was pioneered by Golub et al [167] when they performed supervised classification on leukemias. Unsupervised methods such as hierarchical clustering are exploratory methods useful for identification of previously unrecognized subgroups in tumor material, based solely on the global gene expression profiles. 35

Analysis in blood malignancies A number of studies have shown that gene profiling analysis can correctly classify leukemia [167, 168] as well as predict response to therapy and clinical outcome in pediatric ALL [169], in adult T-cell ALL [170] and follicular lymphoma [171]. Gene profiling studies have also identified a previously unknown subgroup of large B-cell lymphoma [172]. Gene expression studies of primary CML cells have concentrated on the comparison of normal cells with those obtained in CP [173-177], and on the identification of a gene signature associated with disease progression, by comparing cells obtained in CP with AP/BC cells [178-180]. Gene expression profiling in CML has also been used to identify genes that can be used to predict the response and sensitivity to imatinib treatment [181-184]. These studies have reported gene expression signatures that can be used to distinguish between responders and non-responders to imatinib, responders defined as those patients that achieve a CCgR. However no consensus signature could be identified throughout the studies and Crossman et al [185] failed to identify a molecular signature associated with imatinib response. One major drawback is that several of these studies used a limited number of samples and the samples were heterogeneous with respect to the source of cells. A recent study subtracted the normal CD34 + cells expression signature from the CD34 + CML expression signature, and identified a set of 103 differentially expressed genes [179]. At least four studies have used (CD34 + or AC133 + ) enriched cells from CML patients, with two of them focusing on identification of genes involved in disease progression, and one on prediction of an aggressive or indolent disease outcome from diagnostic CP CML samples [177, 178, 180, 186]. Recently, Radich et al [179] reported an extensive gene expression study in which 3.500 genes were identified as being differentially expressed across the different disease phases of CML. A very similar expression profile was observed between AP and BC, suggesting that, at least at the transcriptional level, CML may be considered as a two-step, rather than a three-step process. Of the genes involved in disease progression, SOCS2 has shown increased expression, whereas members of the CEBP family that regulate myeloid differentiation show reduced expression in advanced phases relative to CP CML [179, 180]. 36

Aims for the doctoral project To evaluate a randomized study comparing the effect of busulphan versus HU influence on time to BC and on survival. To evaluate the mobilization of Ph-negative stem cells with intensive chemotherapy + G-CSF. To investigate the in vitro cytotoxic activity of imatinib alone, or in combination with other agents that could be useful in clinical trials. To identify a molecular signature predictive of response to IFN. 37

Material and Methods All studies were approved by the Ethics Committee in all regions. Paper I Study participants and data collection Newly diagnosed CML patients in CP at regional and university hospitals in Sweden, from 1 January 1984 to 30 July 1989, were randomized to either busulphan or HU. Although cytogenetic analysis was not a requirement for inclusion, analysis for Ph chromosome was performed for a majority of the patients and therefore the outcome was separately analyzed for Ph-positive patients. Overall survival, time to BC, treatment side effects and outcome of allogeneic SCT were analyzed. The final follow-up of the study was year 2000. Statistical analysis Assuming 50% of the patients in one group experienced treatment failure, approximately 100 patients per treatment group were required to establish a difference of 20%, which was equivalent to 30% of the patients experiencing treatment failure in the other group, with a power of 80% and 5% significance level (two-sided). In each treatment group, 100 patients were planned, but the introduction of IFN therapy decreased doctors compliance and forced a limit to the patient number. When comparing survival, patients were censored at the time of SCT. Actuarial curves were estimated according to the Kaplan and Meier method [187]. The risk-group classification was performed according to Sokal [26]. 38

Paper II Study participants and data collection Between 1994 and 1999, 37 patients were included in the study. The patients were aged 16-55 years, had not achieving CCgR after 6 months treatment with IFN and HU, and were not eligible for allogeneic SCT. The patients were recruited from 12 regional and university hospitals in Sweden. All patients received daunorubicin and Ara-C (DA) for mobilizing BSC. Patients mobilizing poorly could receive a four-day cycle of chemotherapy with mitoxantrone, etoposide and Ara-C (MEA) or a second DA. At day +8, G-CSF s.c (lenograstim, Granocyte, Aventis) was started at a dosage of 526 µg once daily and continued until leukaphereses were completed. Daily WBC count, with differential and number of blood progenitor cells (CD34 + cells in blood: B-CD34 + ) [188], was checked from the time when WBC count was <0.1x10 9 /l until the leukaphereses were completed. The leukapheresis products were analyzed daily for the number of mononuclear cells (MNC), B- CD34 +, colony-forming units-granulocyte macrophage (CFU-GM) and for cytogenetics. Bone marrow cytogenetic analysis was performed on one of the leukapheresis days. Twenty-one patients later underwent autologous SCT, aiming to induce long-standing CCgR. The final follow-up of the study was performed in 2001. Treatment evaluation BSC harvest was considered successful when MNC 3.5 x 10 8 /kg, CFU-GM 1.0 x 10 4 /kg, CD34 + cells 2.0 x 10 6 /kg and the share of Ph-positive cells was 0% (complete success) or 1-34% (partial success). Malignant cell contamination of the BSC harvests was assessed by standard cytogenetic analysis, or when necessary, by FISH [189] or PCR [37]. The duration of hematopoietic recovery was calculated from the first to the last day of polymorphonuclear (PMN) count <0.5 x 10 9 /l and of platelet count <20 x 10 9 /l (without platelet transfusions). Statistical analysis The probability of OS was estimated according to the Kaplan and Meier method [187]. Surviving patients were censored on the last day of follow-up. 39

Paper III Cell samples and cell preparation This study contained three different parts: For the methodological study, nine (6 CP, 3 BC) samples were used and the drug sensitivity in fresh tumor cells and cryopreserved cells from blood and bone marrow were compared. In the imatinib single drug and imatinib combination study, the K562 cell line and cryopreserved tumor cells from newly diagnosed, untreated CML CP (n=37), CML BC (n=8), CLL (n=3) and AML (n=3) patients were used. None of the patients had received imatinib. Tumor cells from bone marrow or blood were isolated by 1.077 g/ml Ficoll-Isopaque density gradient centrifugation [148] within 24 h after collection. Viability was determined by the trypan blue exclusion test and the proportion of tumor cells calculated by inspection of May-Grunwall-Giemsa stained cytocentrifugate preparations on days 0 and 3. Culture medium RPMI 1640 supplemented with 10% heatinactivated fetal calf serum, 2 mm glutamine, 100 U/ml penicillin, and 100 g/ml streptomycin was used. The test was approved if viability was >70%. All CML samples were 100% Ph-positive. Reagents and drugs Cytotoxic agents were selected based on clinical usefulness in CML, diversity in mechanism of action and from literature data [94, 101, 155, 157, 190, 191]. The concentrations of the cytotoxic agents were chosen based on our own data from previous studies on CML cells (cytarabine 10 M, daunorubicin 0.2 M, arsenic trioxide 1 M, IFN 50000 IE/ml and homoharringtonine 0.1 M) (unpublished data). Imatinib was supplied from Novartis and dissolved in dimetyl sulphoxide (DMSO) to a concentration of 10 mm. To examine the concentration-response relation, the drug was tested at concentrations ranging from 100 M to 0.01 M. For the CML CP samples, the concentration 1.0 M of imatinib resulted in approximately 50% cell survival. This concentration was therefore chosen for the combination study. Experimental microtiter plates were prepared with 20 l/well of solution with the cytotoxic agents at 10 x the desired final concentration. All agents and concentrations were tested in triplicate. Fluorometric Microculture Cytototoxic Assay FMCA is a TCK assay based on the measurement of fluorescence generated from hydrolysis of FDA to fluorescent fluorescein in cells with intact cell 40

membranes, as previously described [148, 192]. After continuous 72-hour exposure to the cytotoxic agents in the 96-well microtitre plates at 37 C, the plates were centrifuged, and washed once with phosphate buffered saline (PBS). Hepes buffered saline containing FDA was added column-wise. The plates were incubated for 40 minutes before reading the fluorescence in a scanning fluorometer. The fluorometer was blanked against wells containing PBS including the dye, but without cells. The cell survival was expressed as a percentage and calculated as the ratio between the fluorescence in drugtreated wells and the fluorescence in control wells with blank values subtracted. Low numerical values thereby indicated a high cytotoxic effect. Statistical analysis Mean values for different groups were compared using the Mann-Whitney test for unpaired comparisons. Wilcoxon test was used to compare blood versus bone marrow and cryopreserved versus fresh samples. Two sided p- values with a significance limit of 0.05 were used throughout. When the effects of drug combinations were analyzed, the cytotoxic effect of the combination was compared to the effect of the most active constituent alone ( best drug ). If the ratio between the survival index after combination treatment and the survival index after treatment with the most effective single agent was 0.8-1.2, the effect of the combination was considered as caused by the best drug only. Ratios below 0.8 were defined as a positive interaction and as antagonism when above 1.2. The best drug model for interaction between two drugs was compared with the medianeffect method proposed by Chou and Talalay [193]. Drug interaction was evaluated using the combination index (CI) equation by the software CalcuSyn (Biosoft, Ferguson, MO) [194] and it was assumed that the substances acted mutually non-exclusively (conservative isobologram) [193]. Paper IV Patient samples Blood samples, or alternatively leukapheresis cells, were collected at diagnosis from 22 patients with CML CP. The majority of the patients received IFN in combination with HU or Ara-C, while one patient received IFN alone. Responders were defined as patients achieving a CCgR or MCgR after 12 months or earlier, after initiation of IFN. Patients that had no cytogenetic response at six months were considered as non-responders. Mononuclear cells from two healthy blood donors were used as reference samples in all experiments. 41

Total RNA preparation Total RNA was extracted by TRIzol preparation (Gibco-BRL) for all 22 patients and reference samples. The RNA quality of all samples was analyzed with the RNA 6000 Nano Lab Chip kit in the bioanalyzer 400 (Agilent Technologies, Palo Alto, CA, USA). Array fabrication and analysis The cdna microarray used was manufactured at the Microarray Core Facility at Rudbeck Laboratory, Uppsala University, Sweden, and contained 15.552 spots representing 7.458 CDNA clones and controls [195]. Three micrograms of patient RNA were labelled using the TSA Labelling and Hybridization Kit, mixed with the reference sample and hybridized to the array. Each patient sample was analyzed twice. The microarrays were scanned in a Gene Pix 4000B scanner. Raw data was normalized using the SMA package (Statistics for Microarray Analysis, http://www.stat.berkeley.edu/users/ terry/zarray/software/smacode.html). The gene lists for both groups were matched resulting in 2.916 genes, present on both lists, for futher analysis. Hence, 2916 genes, out of 7.458 genes, were used for supervised learning and hierarchical clustring analysis. Supervised Learning The discriminative power of selected groups of genes was evaluated through their performance in supervised classification. In a leave-one-out cross validation procedure (LOOCV), one sample at a time was withheld and not used for either gene selection, or training of the classifier [196]. In this series, the best discrimination between responders and non-responders was achieved by the top 20 genes. These genes (with the largest median differences between the groups) where then used in a selection procedure. In order to verify that the error rates obtained with the data set were unlikely to have occurred by chance, a permutation test was performed. RT-PCR validation The RT-PCR reactions were performed in duplicate. Assay on demand mixes for RNASE2, PRG2, NRGN, LTF, JARID1A and DEFA4 were purchased from (Applied Biosystem, Foster city, CA, USA). The expression in each gene was normalized to GAPDH expression in each sample. 42

Results and Discussion Comparison of busulphan, hydroxyurea and allogeneic stem cell transplantation (Paper I) The 179 patients registered in the study represented 35% of the estimated CML patients diagnosed over this five and a half year period. Ninety patients were randomized to HU, and 89 to busulphan treatment. Twenty-six patients underwent allogeneic SCT, 21 with RD, and 5 with URD. Analysis for Ph chromosome by conventional cytogenetics was obtained for 164 patients. The median age was 56 for the busulphan and 61 years for the HU group. There was no significant difference in survival between busulphan and HU treated patients (Figure 5). Median survival was 3.2 years and 3.5 years, respectively. The treatment intensity did not differ between the two groups, nor did the time from diagnosis to therapy start (mean 24 for busulphan and 42 days for HU). One hundred and sixty-seven patients died, 85 due to BC (41 busulphan and 44 HU), and 81 had other complications (infections, bleedings, heart diseases and other malignant diseases). In one patient, the cause of death was unknown. Some of the latter 81 probably also had a BC, but no post-mortem bone marrow samples were taken to prove this. Two transplanted patients died a long time after diagnosis (10 and 14 years), one in BC and one in transplantation-related complications. There was a significant survival benefit for patients who underwent allogeneic SCT (n=26) (Figure 6). In addition, when compared with nontransplanted patients in the same age group (up to 46 years) the outcome for the transplanted patients was significantly better. Thus, the age factor did not explain why allogeneic SCT was superior to conventional therapy. The median survival was 4.7 and 3.3 years, respectively. In all, 10 transplanted patients (38%) and 2 non-transplanted patients (1.3%) were still alive (both in the busulphan group). At the latest follow up, in year 2000, all the 10 surviving transplant patients were RT-PCR-negative for BCR-ABL in blood, whereas, both non-transplanted survivors were 100% Ph-positive in bone marrow. When this study began, it was not known whether HU prolonged survival compared with busulphan, although other groups were studying these two 43

drugs. The MCR CML III b [48] study did not find any difference in treatment outcome; however, the German CML I study [197] reported an advantage for HU treatment on OS. The MRC statistical group [39] published a meta-analysis of three randomized trials comparing HU and busulphan (ours, German CML I and MRC IIIb), showing that HU treatment is beneficial to survival. It might be speculated whether our study could not identify any differences. The reasons could be either that treatment intensity was the same for the two groups, or that the patient number was too small. In the German study, the proportions of patients with normal or subnormal WBC count were considerably higher in the HU group, which could explain the positive impact on survival. An important finding in our study was the positive outcome of allogeneic SCT, the 10-year survival being 46% compared to 2% for the nontransplanted group. This study demonstrated the importance of a long follow-up time for revealing the positive effects of allogeneic SCT. It took between four and five years for the different regimen survival curves to separate, and it should also be noted that these patients were transplanted during the mid-1980s, when TRM was relatively high. Both tissue typing and supportive care have improved considerably since then. Thus in the Nordic countries, allogeneic SCT results improved during the 1990s reaching 65% OS [55]. OVERALL SURVIVAL FROM START OF THERAPY Survival 1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,0 Bu n=88 Median survival 3.2 years Hu n=89 Median survival 3.5 years p=0.4647 0 1 2 3 4 5 6 7 8 9 Years 10 11 12 13 14 15 16 Figure 5. Overall survival from start of therapy. Kaplan-Meier survival curves for CML patients treated with busulphan (n=88) versus HU (n=89). 44

OVERALL SURVIVAL FROM START OF THERAPY Survival 1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,0 Transplanted n=26 Median survival 4.7 years Not transplanted n=151 Median survival 3.2 years p=0.0001 0 1 2 3 4 5 6 7 8 9 Years 10 11 12 13 14 15 16 Figure 6. Overall survival from start of therapy. Kaplan-Meier survival curves comparing allogeneic SCT patients with non-transplanted patients (treated with busulphan or HU). Successful mobilization of Ph-negative blood stem cells with intensive chemotherapy + G-CSF (Paper II) In a national studying the period 1994-1999, 37 patients were reported: 27 were mobilized with one, and 10 patients received two chemotherapy regimens. Thus, there were 47 mobilization procedures. DA and MEA were used 41 and 6 times, respectively. G-CSF was administrated for a median of 20 (range 5-60) days/cycle. No treatment-related deaths occurred. In total, 114 leukaphereses were performed in 34 patients: three patients never underwent this procedure due to poor B-CD34 + mobilization (n=2) or fever (n=1). Leukaphereses were performed at a median of 27 days (range 15-59) from the first day of chemotherapy, with a median of 2 (range 1-7) aphereses per patient. Data on cells collected are presented in Table 7. Target cell count (B- CD34 + cells >2.0 x 10 6 /kg) was reached in 30/34 patients (88%) in one or more leukaphereisis procedures: no mortality was observed during the mobilization procedure. Ninety BSC collections underwent cytogenetic analysis. In patients with sufficient collected BSC, complete success (0% Ph-positive cells) was achieved in 16 (43%), and partial success (1-34% Ph-positive cells) in 7 (19%). Thus, 23 of 37 patients (62%) were eligible for autografting. 45

Table 7. Results of harvest in 34 patients. Median Range Day of first leukapheresis 27 15-59 Number of leukaphereses 2 1-7 MNC x 10 8 /kg 3.9 0.9-17 CFU-GM x 10 4 /kg 17 0-85 B-CD34 + x 10 6 /kg 3.7 0.2-52 The patients in the failure group did not have enough cells and/or had Phpositive cells >34% in the harvest. There was a trend towards a lower percentage of Ph-positive cells in the first apheresis harvest compared with the last one (ns). Cytogenetic analysis of bone marrow and the same day s apheresis product showed no statistical difference in Ph-positivity. Twenty-three patients were eligible, but only 21 underwent autologous SCT: 15 with 100% Ph-negative and 6 patients with Ph-positive (1-34%) leukapheresis products. The remaining two patients underwent an URD allogeneic SCT, despite one having a successful, and the other a partially successful, mobilization. There was a rapid hematopoetic recovery after autologous SCT. The number of days with PMN count <0.5 x 10 9 /l was 10 (range7-49) and with platelets <20 x 10 9 /l also 10 (range 2-173). The estimated 5-year OS was 68%, with a median follow-up time of 5.2 years. CCgR after autologous SCT was achieved in 18 of 21 patients (85%) and lasted 3-78+ months (median 7.5 months). There was no transplant related mortality. Four patients subsequently underwent URD allogeneic SCT due to cytogenetic relapse. Twelve out of 21 patients were still alive 23 to 95 months post transplant (median 68 months). Eleven had a cytogenetic relapse, however: seven of these were due to treatment in CCgR at the latest follow-up. Four of these patients had received imatinib (one RT-PCR negative, two RT-PCR positive and one with unknown RT-PCR status), one IFN (RT-PCR positive) and two had undergone URD allogeneic SCT (both RT-PCR-negative). One patient did not receive any CML treatment post autographing (RT-PCRpositive) and remarkably, was still Ph-negative 10 years after transplantation with Ph-negative BSCs. The remaining four surviving patients were Phpositive and received imatinib (n=2), HU (n=1) and unknown treatment (n=1). Nine patients died (five in blast transformation, one from breast cancer, one in CNS relapse, one in GVHD and one for unknown reason). The 46

survival for all 37 patients (intention-to treat) compared with the outcome for the 16 patients not autotransplanted was not significantly different. In summary, this study showed that target cell count was reached for the majority of patients (88%). Twenty-three out of 37 (62%) were completely successfully, or partially successfully mobilized, as defined by the protocol (B-CD34 + cells >2 x 10 6 /kg and Ph-positive cells <34% in the apheresis products). Out of these, 16 patients (43%) were Ph-negative in the apheresis products. These results compare favorably with other studies using cytostatic regimens for mobilization. Thus with ICE (idarubicine, Ara-C, etoposide) successful mobilization was achieved in 31% and partially successful mobilization in 17% of the patients [198]. Corresponding figures were 43%with idarubicine and 16% with Ara-C [199]. Pratt et al [200] used high-dose HU for mobilization, with a successful outcome in 4/14 (28%) and partially successful outcome in 4/14 patients (28%). Although other groups have reported treatment-related deaths, mainly due to prolonged cytopenia and infections [53, 201, 202], no mortality occurred in this study. Collecting BSC very early during hematopoietic recovery post chemotherapy may decrease the risk of tumor cell contamination. Previous publications have reported more Ph-negative products in the first collections than in later ones. In this study, a similar trend was revealed, but it was not statistically significant, probably due to too a low patient number. Several groups have evaluated autologous SCT as an approach to treat patients with CML lacking a suitable donor, or too fragile for allogeneic SCT [53, 55]. In a review of the experience in eight centres, McGlave et al [203, 204] showed the 3-year survival after autologous SCT for selected CP patients to be approximately 60%. Bathia et al [205] reported cytogenetic responses after autologous transplantation in a range of 10-57%. Since imatinib was introduced for CML treatment, the procedure of autologous SCT is rarely performed. However, given that imatinib often induces CCgR, it may in the future prove valuable to harvest BSC at the time of CCgR or MMolR with potential use at the time of progression in patients not eligible for allogeneic SCT. Many study groups have demonstrated the possibility of successfully collecting BSC in 40-80% of the patients achieving CCgR with imatinib [57-60]. 47

In vitro activity of imatinib (Paper III) In an initial investigation, the feasibility of using the FMCA method in drug sensitivity testing on tumor cells from CML patients was evaluated (Methodological study). Many other studies have been performed using the FMCA in other hematological malignancies as well as in solid tumors [148-150, 153, 157, 192], but not in primary CML tumor cells. No difference in drug sensitivity between tumor cells obtained from blood and bone marrow was found, indicating that blood can be used to make sampling easier for the patient. On the other hand, cryopreserved cells tended to be more drug sensitive than fresh cells, suggesting that it may not be optimal to use fresh and cryopreserved together in the same study. Considering this, only cryopreserved cells were used in the rest of this study. Drug sensitivity analysis was performed with the FMCA method in samples from untreated CML CP (n=26), with CML BC (n=6), and the K 562 cell line (Imatinib single drug study). The CML samples from BC were significantly (p< 0.05, Mann-Whitney test) more sensitive to imatinib than the CP cells at the concentrations 1, 10 and 100 M. The K 562 cell line (CML BC) had a similar pattern as the CML BC patient samples (Figure 7). Most published in vitro studies of imatinib use CML cell lines that originate from patients in BC [117, 157]. The use of such cell lines to draw conclusions concerning CML CP may be sub-optimal, risking an overestimation of drug efficacy and of drug combination effects. Imatinib is very efficient in patients with CML CP, [206, 207] whereas in AP and BC, the clinical effects are moderate and transient [103, 208]. In the present study, the BC samples were significantly more sensitive to imatinib than the samples taken in CP at diagnosis. The reason for this is unclear, although a possible explanation could be that BC cells are immature cells that generally have fewer mechanisms of protection against foreign stimuli. In the clinical settings, factors in addition to cellular drug sensitivity also come into play. The K 562 cell line had drug sensitivity comparable to the BC cells. Then, the combination of imatinib with conventional (IFN, daunorubicin, cytarabine, vincristine) and experimental (arsenic trioxide, homoharrintonine) anti-cml agents was analyzed to determine if the activity of the single agent in vitro could be enhanced or reduced (Imatinib combination study). The highest mean difference in cell survival (%) of the combination imatinib and other cytotoxic agents to the most active single agent was compared. IFN, daunorubicin and arsenic trioxide were the drugs with the highest frequency of positive interactions with imatinib in CML CP samples (Figure 8 a). Unfortunately, the high single agent activity of homoharringtonine, as well as of cytarabine, made any potential benefit of combinations with imatinib difficult to evaluate. Imatinib in combination with cytarabine, vincristine, arsenic trioxide and homoharringtonine showed a positive interaction in all BC samples (Figure 8 b). Remarkably, IFN in combination 48

with imatinib had an antagonistic effect in two of the BC samples and this may explain the clinical situation, where IFN is used only in patients in CP, and not in BC. Tumor cells from patients obtained in BC were more sensitive to combinations than samples from CP (<0.05, Mann Whitney test: Figure 8 b). Survival Index (%) 100 80 60 40 20 CP BC K562 0 0.01 0.1 1 10 100 Imatinib concentration ( M) Figure 7. Mean concentration-effect curve for imatinib in samples from CML patients in CP (n=26), BC (n=6) and from the K 562 cell line. The error bars show ± 1 SEM. As primary and acquired resistance to imatinib therapy is problematic in CML therapy, trials combining imatinib with other anti-cml agents are of major interest. New oral tyrosine kinase inhibitors, such as nilotinib and dasatinib are promising drugs in case of imatinib intolerance/resistance [119, 122]. Previous combination studies performed in BCR-ABL positive cell lines, indicate synergistic effects of imatinib with cytarabine, vincristine and IFN, and both additive and synergistic effects of imatinib combined with daunorubicin and homoharringtonine [94, 154, 155, 157, 190, 191]. Also cell lines transfected with different mutations are used, for example Ba/F3, to develop drugs effective against specific mutations [117]. There is evidence that CML stem cells are not eliminated by imatinib in vivo; therefore many laboratories investigate in vitro effects on enriched CML stem cells (CD34 +, CD38 - ) [156]. Our BC results were thus similar to previous results published on CML cell lines, as well as results on Ph-positive ALL cells [147]. In the current study, there were generally smaller combination effects in cells from CP, 49

indicating that cell line studies may overestimate the effects in CML CP. The chosen concentration of 1.0 M of imatinib in the combination studies is relevant in treatment, as administration of 350 mg imatinib daily in the study of Peng et al [209] resulted in mean plasma concentration of about 1 M. Survival Index (%) 100 75 50 25 a) Imatinib X Imatinib+X 0 Cytarabine Arsenic Trioxide Vincristine Daunorubicin Interferon Homoharringtonine Survival Index (%) 100 75 50 25 b) Imatinib X Imatinib +X 0 Cytarabine ArsenicTrioxide Vincristine Daunorubicin Interferon Homoharringtonine Figure 8. The mean survival index (%) in samples from CML patients in CP (n=26) for imatinib and six drugs as single drugs, and in combination (a), and in samples from CML patients in BC (n=6) for imatinib and six drugs as single drugs, and in combination (b). The error bars show ± 1 SEM 50

Gene expression analysis in interferon- responders (Paper IV) Twenty-two patient samples were included in the microarray-based gene expression study; however, seven samples not passing the quality requirements were excluded. Of the remaining samples, six were from patients responding, and nine from patients not responding, to IFN. In a first step unsupervised hierarchical clustering analysis was performed. The samples clustered into two distinct categories according to the sample processing procedure used before RNA extraction: ficoll density sedimentation or leukapheresis separation (Figure 9). This result indicates that the largest differences in gene expression between our samples arise from the different types of cells isolated by the two different methods. Unsupervised clustering was unable to separate responders from non-responders to IFN. In a subsequent analysis we ranked the genes according to the differrences in gene expression between the responders and non-responders. The top genes that showed the largest differential expression were used in a new hierarchial clustering. Again the analysis was unable to separate the two groups. Similarly, Crossman et al [185] could not identify a molecular signature associated with response to imatinib by hierarchical clustering when analyzing a mixed sample population consisting of both blood and bone marrow samples. These problems illustrate the importance of processing samples uniformly. In order to identify genes that could potentially predict response to IFN, the LOOV procedure was performed, and the genes most frequently utilized in this selection procedure were RNASE2 (ribonuclease family 2), PRG2 (proteoglycan 2), and NRGN (neurogranin), LTF (lactotransferrin), JARID 1A and DEFA4 (defensin alpha 4) (Figure 10). Classification of the withheld samples based on these genes resulted in an error rate of 13%. A permutation test showed that it was unlikely to obtain equal or more successful classification by chance (p=0.004). L L L L F F F F F F F F F F F Figure 9. Outcome of sample classification by hierarchical clustering of diagnostic CML samples. L=samples collected by leukapheresis. F=samples processed by ficoll density centrifugation. 51

The genes selected in the LOOV seem to be biologically relevant. Twelve out of 20 genes selected for the LOOCV procedure either had a cis element for the C/EBP transcription factor family in the promoter region (RNASE2, PRG2, LTF, MEFV, BPI,MT1F and CST 7), or was coupled indirectly to the transcription factors, e.g. MYC that antagonizes the transcription factor C/EBP. Previous studies have shown that C/EBP is expressed early in the myeloid differentiation [210] and C/EBP is essential for terminal differentiation of committed granulocyte progenitor cells. This finding suggests that the ability to respond to IFN might be associated with the capacity to differentiate in CML cells. C/EBP is negatively regulated by bcr-abl [211] and the introduction of C/EBP inhibits proliferation and promotes differentiation of BCR-ABL expressing cells [212]. RNASE2 and PRG2 are primarily expressed in eosinophils and have a higher expression in the non-responder group. These two genes have been identified to distinguish CD34 + cells derived from CP and BC CML samples [180]. The PRG2 is over-expressed in CD34 + cells from CML patients with a more indolent disease [186]. DEFA4 is highly expressed in neutrophils. Several of the genes reported in this study as differentially expressed between the two categories of patients, have been reported in other gene expression studies on CML cells. DEFA4, NRGN, CAMP, BPI, LTF, MYC and CD44 have been found to be deregulated in CML cells compared to normal controls [176]. Thus, a number of the genes identified in the present study appeared to be associated with CP or indolent CML, which suggests that they might have prognostic value not only with regard to IFN response but also perhaps in a general context for CML patients, irrespective of treatment. The RT-PCR data confirmed that RNASE2, PRG2, JARID1A and DEFA4 had higher expression in non-responders, whereas, NRGN and LTF had a lower expression in this group. Analyzed individually by RT-PCR RNASE2, PRG2 and NRGN showed statistical difference among responders and nonresponders (p<0.05). The results from the RT-PCR were used to perform a principal component analysis, which correctly clustered the samples into two distinct categories, with one exception. One responder patient clustered among the non-responders, and interestingly, this patient rapidly progressed to BC, despite the good initial response to IFN. We conclude that it might be possible to use gene expression analysis to predict future response to IFN in CML diagnostic samples. Larger studies with uniformly processed samples are warranted in order to confirm our findings. 52

Figure 10. Expression level of RNASE2, PRG2, NRGN, JARID1A, DEFA4 and LTF transcripts relative to GAPDH expression determined by RT-PCR in CML samples (n=13). Filled circles represent non-responders and open circles represent responder patients. * Statistically significant. 53

Conclusions Paper I No difference between busulphan and HU therapy on OS was determined. Reports that allogeneic SCT improves survival in CML were confirmed. Paper II In the majority of patients with CML CP, intensive chemotherapy and G- CSF mobilize Ph-negative BSC sufficient for autologous SCT. Paper III FMCA may be a useful method for drug sensitivity testing in samples from CML patients, both in CP and BC. Patient tumor cells may have advantages over cell line for sensitivity testing. Combinations between imatinib and IFN, daunorubicin and arsenic trioxide may by interesting for clinical trials in patients with CML CP. Paper IV With microarray-based gene expression analysis on diagnostic samples derived from CP CML patients, a set of genes that may possibly predict response to IFN treatment were identified. In patients not responding to IFN treatment RNASE2, PRG2, JARID1A and DEFA4 showed a higher expression, whereas NRGN and LTF showed a lower expression. 54

Future directions Much effort has been invested in developing valuable prognostic scores to predict response and survival in patients with CML [26, 27]. Common for both scores are the possibility of distinguishing patients with a more indolent from patients having a more aggressive disease, and that the patients are sub grouped into low-, intermediate-, and high-risk patients. In the imatinib era, factors related to therapy response seem to be more important. Early cytogenetic response to therapy being the most important. We are currently performing microarray analysis to explore differences in gene expression pattern among CP CML patients at diagnosis, in particular differences among the three different risk groups according to Sokal. The identification of genes and pathways involved in the malignant transformation could provide biomarkers that are useful for identifying risk patients already at diagnosis. Thirty-three patients (eleven from each Sokal risk group) were included in the study. We have started to analyze the data. Later, we would like to verify our results in an independent patient cohort. Resistance to imatinib represents an important scientific and clinical issue in CML. Mutations in the ABL domain is a major cause, but we do not yet know all the mechanisms of resistance. To overcome resistance to imatinib, different strategies were developed, including dose escalation of imatinib or combination with conventional cytotoxic drugs, or new drugs such as dasatinib, nilotinib, bosutinib and MK-0457. It may be possible to combine these drugs. The FMCA system could be useful to study this, but it may be better to use primitive CML stem cells (CD34 +, CD38 - ) in the future, or mutated cell lines. Despite the fact that imatinib is the most effective debulking therapy for CP patients, the eradication of residual disease without allogeneic SCT still seems a difficult goal for the tyrosine kinase inhibitor approach alone. Peptide vaccines in eliminating and/or controlling residual cells by inducing a leukemia-specific immune reaction is an attractive solution. However, the peripheral blood of cancer patients frequently contains an increased number of T-regulatory cells, which appear to inhibit immune reactivity. We would like to investigate these T-regulatory cells in CML patients both at diagnosis and after treatment with imatinib. I also hope to continue working with academical studies concerning CML therapy, in the Swedish and Nordic CML collaboration groups. 55

Acknowledgments I would like to express my sincere gratitude to all the people who made this thesis possible. My special thanks to: Bengt Simonsson, my supervisor, for excellent scientific guidance and never-ending support and encouragement. My deepest gratitude to you for always being there with wisdom and inspiration when I needed it and for sharing your expert knowledge in your favourite field, CML. Elin Lindhagen, my co-supervisor, for introducing me into the FMCA - world. Thank you for your patience and never-ending reading of my manuscript. Gisela Barbany, my co-supervisor, for interesting scientific discussions and valuable critical comments. Thank you for excellent travel company to all CML meetings. Anette Hagberg, for all your help with laboratory matters in the microarray-field Rolf Larsson, and the staff at Clinical Pharmacology, for giving me the opportunity to come to your lab and for providing excellent laboratory facilities. Anneli Björnberg, for helping me with the analysis of all CML samples at Clinical Pharmacology. All my co-authors in the Swedish CML group. Inger Persson, for well needed statistical advice over the years. Bengt Smedmyr, head of the Department of Hematology, for creating good opportunities for research. Anna, Elisabeth and Helene, my roommates. Thank you for being such terrific collegues and friends! I stick to my room, the best place in the hospital. Special thanks to Anna for all help with my FMCA project. 56

My other colleges at the Department of Hematology: Martin Höglund, Kristina Carlson, Gunnar Birgegård, Gunnar Öberg, Sara Rosengren, Anna Eriksson and Gunnar Larfors. Thank you for good companionship and for letting me off clinical work. Special thanks to Martin for all suggestions on how to improve my manuscripts. All the staff at the Department of Hematology, 50 C and blodmottagningen, for creating a pleasant atmosphere and help with all patient samples. Gunilla Eriksson, for secretarial help. Janet Ahlgren, for skilful linguistic revision. Leif Engqvist, for introducing me into the world of hematology in Värnamo, during my first job as a doctor. My other friends, from the world outside the hospital. A special thanks to Eva and L-L for being my sisters. My family, that has been great during this period of time. I have not being the nicest person! Per, my beloved, has given me all the love and support I could possibly ask for. Clara and Maja for being the sunshine and joy in my life. You are the reason for living! My parents, for their constant support and encouragement. I am so sorry that they could not share this moment with me! Financial support from the Lions Cancer Research Foundation in Uppsala and the Swedish Cancer Society is gratefully acknowledged. 57

References 1. Bennett JH.: Case of hypertrophy of the spleen and liver, in which death took place from suppuration of the blood. Edingburgh Medical and Surgical Journal 1845. 64: p. 413-424. 2. Virchow R:Weisses Blut. Frorieps Notizen 1845. 36: p. 151-156. 3. Neumann E: Uber myelogene leukemia. Berliner Klinische Wochenschreibe 1878. 15: p. 69-72. 4. Nowell PC, Hungerford DAA: minute chromosome in human granulocytic leukemia. Science 1960. 132: p. 1497. 5. Rowley JD: A new consistent chromosomal abnormality in chronic myelogenous leukemia identified by quinacrine fluorescence and Giemsa staining. Nature 1973. 243: p. 290-293. 6. Haut A.. Busulphan in the treatment of chronic myelocytic leukemia. The effect of long intermittent therapy. Blood 1961. 17: p. 1-19. 7. Galton. DA. Chemotherapy of chronic myelocytic leukemia. Semin. Hematol. 1969. 6: p. 323-43. 8. Talpaz, M., et al., Leukocyte interferon-induced myeloid cytoreduction in chronic myelogenous leukemia. Blood, 1983. 62(3): p. 689-92. 9. Cotta, C.V. and C.E. Bueso-Ramos, New insights into the pathobiology and treatment of chronic myelogenous leukemia. Ann Diagn Pathol, 2007. 11(1): p. 68-78. 10. Kantarjian, H.M., et al., Clinical course and therapy of chronic myelogenous leukemia with interferon-alpha and chemotherapy. Hematol Oncol Clin North Am, 1998. 12(1): p. 31-80. 11. Faderl, S., et al., Chronic myelogenous leukemia: biology and therapy. Ann Intern Med, 1999. 131(3): p. 207-19. 12. Gunz, F.W., The epidemiology and genetics of the chronic leukaemias. Clin Haematol, 1977. 6(1): p. 3-20. 13. Jaffe ES, Harris HL, Stein H and Vardiman JW: World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissures. IARC Press:Lyon 2001. 14. Berger, U., et al., Gender aspects in chronic myeloid leukemia: longterm results from randomized studies. Leukemia, 2005. 19(6): p. 984-9. 15. Hehlmann, R., et al., Randomized comparison of interferon-alpha with busulfan and hydroxyurea in chronic myelogenous leukemia. The German CML Study Group. Blood, 1994. 84(12): p. 4064-77. 58

16. Gluzman, D., et al., Patterns of hematological malignancies in Chernobyl clean-up workers (1996-2005). Exp Oncol, 2006. 28(1): p. 60-3. 17. Mehlman, M.A., Dangerous and cancer-causing properties of products and chemicals in the oil refining and petrochemical industries. Part XXX: Causal relationship between chronic myelogenous leukemia and benzene-containing solvents. Ann N Y Acad Sci, 2006. 1076: p. 110-9. 18. Aksoy, M., S. Erdem, and G. DinCol, Leukemia in shoe-workers exposed chronically to benzene. Blood, 1974. 44(6): p. 837-41. 19. Baccarani, M., et al., Evolving concepts in the management of chronic myeloid leukemia. Recommendations from an expert panel on behalf of the European Leukemianet. Blood, 2006. 20. Shtivelman, E., et al., Fused transcript of abl and bcr genes in chronic myelogenous leukaemia. Nature, 1985. 315(6020): p. 550-4. 21. Hughes, T.P., et al., Frequency of major molecular responses to imatinib or interferon alfa plus cytarabine in newly diagnosed chronic myeloid leukemia. N Engl J Med, 2003. 349(15): p. 1423-32. 22. Hughes, T.P., et al., Monitoring CML patients responding to treatment with tyrosine kinase inhibitors - Review and recommendations for 'harmonizing' current methodology for detecting BCR-ABL transcripts and kinase domain mutations and for expressing results. Blood, 2006. 23. Daley, G.Q., R.A. Van Etten, and D. Baltimore, Induction of chronic myelogenous leukemia in mice by the P210bcr/abl gene of the Philadelphia chromosome. Science, 1990. 247(4944): p. 824-30. 24. Lugo, T.G., et al., Tyrosine kinase activity and transformation potency of bcr-abl oncogene products. Science, 1990. 247(4946): p. 1079-82. 25. Sinclair, P.B., et al., Large deletions at the t(9;22) breakpoint are common and may identify a poor-prognosis subgroup of patients with chronic myeloid leukemia. Blood, 2000. 95(3): p. 738-43. 26. Sokal, J.E., et al., Prognostic discrimination in "good-risk" chronic granulocytic leukemia. Blood, 1984. 63(4): p. 789-99. 27. Hasford, J., et al., A new prognostic score for survival of patients with chronic myeloid leukemia treated with interferon alfa. Writing Committee for the Collaborative CML Prognostic Factors Project Group. J Natl Cancer Inst, 1998. 90(11): p. 850-8. 28. Rosti, G., et al., Risk and early cytogenetic response to imatinib and interferon in chronic myeloid leukemia. Haematologica, 2003. 88(3): p. 256-9. 29. Druker, B.J., et al., Five-year follow-up of patients receiving imatinib for chronic myeloid leukemia. N Engl J Med, 2006. 355(23): p. 2408-17. 30. Gratwohl, A., et al., Risk assessment for patients with chronic myeloid leukaemia before allogeneic blood or marrow transplantation. 59

Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Lancet, 1998. 352(9134): p. 1087-92. 31. Huntly, B.J., et al., Deletions of the derivative chromosome 9 occur at the time of the Philadelphia translocation and provide a powerful and independent prognostic indicator in chronic myeloid leukemia. Blood, 2001. 98(6): p. 1732-8. 32. Huntly, B.J., et al., Imatinib improves but may not fully reverse the poor prognosis of patients with CML with derivative chromosome 9 deletions. Blood, 2003. 102(6): p. 2205-12. 33. Quintas-Cardama, A., et al., Imatinib mesylate therapy may overcome the poor prognostic significance of deletions of derivative chromosome 9 in patients with chronic myelogenous leukemia. Blood, 2005. 105(6): p. 2281-6. 34. Rosti, G., et al., Molecular response to imatinib in late chronic-phase chronic myeloid leukemia. Blood, 2004. 103(6): p. 2284-90. 35. Kantarjian, H.M., et al., Long-term survival benefit and improved complete cytogenetic and molecular response rates with imatinib mesylate in Philadelphia chromosome-positive chronic-phase chronic myeloid leukemia after failure of interferon-alpha. Blood, 2004. 104(7): p. 1979-88. 36. Branford, S., et al., Real-time quantitative PCR analysis can be used as a primary screen to identify patients with CML treated with imatinib who have BCR-ABL kinase domain mutations. Blood, 2004. 104(9): p. 2926-32. 37. Barbany, G., et al., Manifold-assisted reverse transcription-pcr with real-time detection for measurement of the BCR-ABL fusion transcript in chronic myeloid leukemia patients. Clin Chem, 2000. 46(7): p. 913-20. 38. Kennedy, B.J. and J.W. Yarbro, Metabolic and therapeutic effects of hydroxyurea in chronic myeloid leukemia. Jama, 1966. 195(12): p. 1038-43. 39. Hydroxyurea versus busulphan for chronic myeloid leukaemia: an individual patient data meta-analysis of three randomized trials. Chronic myeloid leukemia trialists' collaborative group. Br J Haematol, 2000. 110(3): p. 573-6. 40. Galton, D.A., Myleran in chronic myeloid leukaemia; results of treatment. Lancet, 1953. 264(6753): p. 208-13. 41. Podoll, L.N. and S.S. Winkler, Busulfan lung. Report of two cases and review of the literature. Am J Roentgenol Radium Ther Nucl Med, 1974. 120(1): p. 151-6. 42. Baccarani, M., et al., A randomized study of interferon-alpha versus interferon-alpha and low-dose arabinosyl cytosine in chronic myeloid leukemia. Blood, 2002. 99(5): p. 1527-35. 43. Stark, G.R., et al., How cells respond to interferons. Annu Rev Biochem, 1998. 67: p. 227-64. 60

44. Angstreich, G.R., et al., Effects of imatinib and interferon on primitive chronic myeloid leukaemia progenitors. Br J Haematol, 2005. 130(3): p. 373-81. 45. Paquette, R.L., et al., Interferon-alpha induces dendritic cell differentiation of CML mononuclear cells in vitro and in vivo. Leukemia, 2002. 16(8): p. 1484-9. 46. de Castro, F.A., et al., Immunological effects of interferon-alpha on chronic myelogenous leukemia. Leuk Lymphoma, 2003. 44(12): p. 2061-7. 47. Interferon alfa-2a as compared with conventional chemotherapy for the treatment of chronic myeloid leukemia. The Italian Cooperative Study Group on Chronic Myeloid Leukemia. N Engl J Med, 1994. 330(12): p. 820-5. 48. Allan, N.C., S.M. Richards, and P.C. Shepherd, UK Medical Research Council randomised, multicentre trial of interferon-alpha n1 for chronic myeloid leukaemia: improved survival irrespective of cytogenetic response. The UK Medical Research Council's Working Parties for Therapeutic Trials in Adult Leukaemia. Lancet, 1995. 345(8962): p. 1392-7. 49. Ohnishi, K., et al., A randomized trial comparing interferon-alpha with busulfan for newly diagnosed chronic myelogenous leukemia in chronic phase. Blood, 1995. 86(3): p. 906-16. 50. Randomized study on hydroxyurea alone versus hydroxyurea combined with low-dose interferon-alpha 2b for chronic myeloid leukemia. The Benelux CML Study Group. Blood, 1998. 91(8): p. 2713-21. 51. Baccarani, M., et al., Interferon-alfa for chronic myeloid leukemia. Semin Hematol, 2003. 40(1): p. 22-33. 52. Kluin-Nelemans, H.C., et al., Randomized comparison of low-dose versus high-dose interferon-alfa in chronic myeloid leukemia: prospective collaboration of 3 joint trials by the MRC and HOVON groups. Blood, 2004. 103(12): p. 4408-15. 53. Carella, A.M., G. Beltrami, and M.T. Corsetti, Autografting in chronic myeloid leukemia. Semin Hematol, 2003. 40(1): p. 72-8. 54. Goldman, J.M., et al., Bone-marrow transplantation for chronic myelogenous leukaemia. Lancet, 1985. 2(8467): p. 1295. 55. Simonsson, B., et al., Intensive treatment and stem cell transplantation in chronic myelogenous leukemia: long-term follow-up. Acta Haematol, 2005. 113(3): p. 155-62. 56. Autologous stem cell transplantation in chronic myeloid leukaemia: a meta-analysis of six randomized trials. Cancer Treat Rev, 2007. 33(1): p. 39-47. 57. Drummond, M.W., et al., Mobilization of Ph chromosome-negative peripheral blood stem cells in chronic myeloid leukaemia patients with imatinib mesylate-induced complete cytogenetic remission. Br J Haematol, 2003. 123(3): p. 479-83. 61

58. Hui, C.H., et al., Successful peripheral blood stem cell mobilisation with filgrastim in patients with chronic myeloid leukaemia achieving complete cytogenetic response with imatinib, without increasing disease burden as measured by quantitative real-time PCR. Leukemia, 2003. 17(5): p. 821-8. 59. Kreuzer, K.A., et al., Filgrastim-induced stem cell mobilization in chronic myeloid leukaemia patients during imatinib therapy: safety, feasibility and evidence for an efficient in vivo purging. Br J Haematol, 2004. 124(2): p. 195-9. 60. Perseghin, P., et al., Peripheral blood progenitor cell collection in chronic myeloid leukemia patients with complete cytogenetic response after treatment with imatinib mesylate. Transfusion, 2005. 45(7): p. 1214-20. 61. Fefer, A., et al., Disappearance of Ph1-positive cells in four patients with chronic granulocytic leukemia after chemotherapy, irradiation and marrow transplantation from an identical twin. N Engl J Med, 1979. 300(7): p. 333-7. 62. Speck, B., et al., Allogeneic bone-marrow transplantation for chronic myelogenous leukaemia. Lancet, 1984. 1(8378): p. 665-8. 63. Silver, R.T., et al., An evidence-based analysis of the effect of busulfan, hydroxyurea, interferon, and allogeneic bone marrow transplantation in treating the chronic phase of chronic myeloid leukemia: developed for the American Society of Hematology. Blood, 1999. 94(5): p. 1517-36. 64. Gratwohl, A., Does early stem-cell transplantation have a role in chronic myeloid leukaemia? For the proposal. Lancet Oncol, 2005. 6(9): p. 721-2. 65. Gratwohl, A., et al., Change in stem cell source for hematopoietic stem cell transplantation (HSCT) in Europe: a report of the EBMT activity survey 2003. Bone Marrow Transplant, 2005. 36(7): p. 575-90. 66. Gratwohl, A., et al., Allogeneic hematopoietic stem cell transplantation for chronic myeloid leukemia in Europe 2006: transplant activity, long-term data and current results. An analysis by the Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation (EBMT). Haematologica, 2006. 91(4): p. 513-21. 67. Goldman JM, Rizzo JD, Jabocinski KA, et al. Long-term outcome after allogeneic stem cell transplanration for CML.(abstract). Hematol J.2004;5:98.Abstract no.226. 68. Barrett, J., Allogeneic stem cell transplantation for chronic myeloid leukemia. Semin Hematol, 2003. 40(1): p. 59-71. 69. Oehler, V.G., et al., Randomized trial of allogeneic related bone marrow transplantation versus peripheral blood stem cell transplantation for chronic myeloid leukemia. Biol Blood Marrow Transplant, 2005. 11(2): p. 85-92. 62

70. Bornhauser, M., et al., Dose-reduced conditioning for allografting in 44 patients with chronic myeloid leukaemia: a retrospective analysis. Br J Haematol, 2001. 115(1): p. 119-24. 71. Or, R., et al., Nonmyeloablative allogeneic stem cell transplantation for the treatment of chronic myeloid leukemia in first chronic phase. Blood, 2003. 101(2): p. 441-5. 72. Weisser, M., et al., Reduced-intensity conditioning using TBI (8 Gy), fludarabine, cyclophosphamide and ATG in elderly CML patients provides excellent results especially when performed in the early course of the disease. Bone Marrow Transplant, 2004. 34(12): p. 1083-8. 73. Baron, F., et al., HLA-matched unrelated donor hematopoietic cell transplantation after nonmyeloablative conditioning for patients with chronic myeloid leukemia. Biol Blood Marrow Transplant, 2005. 11(4): p. 272-9. 74. Raiola, A.M., et al., Factors predicting response and graft-versushost disease after donor lymphocyte infusions: a study on 593 infusions. Bone Marrow Transplant, 2003. 31(8): p. 687-93. 75. DeAngelo, D.J., et al., Extended follow-up of patients treated with imatinib mesylate (gleevec) for chronic myelogenous leukemia relapse after allogeneic transplantation: durable cytogenetic remission and conversion to complete donor chimerism without graft-versushost disease. Clin Cancer Res, 2004. 10(15): p. 5065-71. 76. Kantarjian, H.M., et al., Imatinib mesylate therapy for relapse after allogeneic stem cell transplantation for chronic myelogenous leukemia. Blood, 2002. 100(5): p. 1590-5. 77. Weisser, M., et al., A comparison of donor lymphocyte infusions or imatinib mesylate for patients with chronic myelogenous leukemia who have relapsed after allogeneic stem cell transplantation. Haematologica, 2006. 91(5): p. 663-6. 78. Buchdunger, E., et al., Inhibition of the Abl protein-tyrosine kinase in vitro and in vivo by a 2-phenylaminopyrimidine derivative. Cancer Res, 1996. 56(1): p. 100-4. 79. Druker, B.J., et al., Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat Med, 1996. 2(5): p. 561-6. 80. Schindler, T., et al., Structural mechanism for STI-571 inhibition of abelson tyrosine kinase. Science, 2000. 289(5486): p. 1938-42. 81. Nagar, B., et al., Crystal structures of the kinase domain of c-abl in complex with the small molecule inhibitors PD173955 and imatinib (STI-571). Cancer Res, 2002. 62(15): p. 4236-43. 82. Deininger, M.W., et al., The tyrosine kinase inhibitor CGP57148B selectively inhibits the growth of BCR-ABL-positive cells. Blood, 1997. 90(9): p. 3691-8. 83. Buchdunger, E., et al., Abl protein-tyrosine kinase inhibitor STI571 inhibits in vitro signal transduction mediated by c-kit and platelet- 63

derived growth factor receptors. J Pharmacol Exp Ther, 2000. 295(1): p. 139-45. 84. Wong, S., et al., Sole BCR-ABL inhibition is insufficient to eliminate all myeloproliferative disorder cell populations. Proc Natl Acad Sci U S A, 2004. 101(50): p. 17456-61. 85. Braziel, R.M., et al., Hematopathologic and cytogenetic findings in imatinib mesylate-treated chronic myelogenous leukemia patients: 14 months' experience. Blood, 2002. 100(2): p. 435-41. 86. Hasserjian, R.P., et al., ST1571 (imatinib mesylate) reduces bone marrow cellularity and normalizes morphologic features irrespective of cytogenetic response. Am J Clin Pathol, 2002. 117(3): p. 360-7. 87. Thiele, J., et al., Bone marrow changes in chronic myelogenous leukaemia after long-term treatment with the tyrosine kinase inhibitor STI571: an immunohistochemical study on 75 patients. Histopathology, 2005. 46(5): p. 540-50. 88. Guilhot,F.On behaf of the IRIS study group. Sustained durability of responses plus high rates of cytogenetic responses result in long term benefit for newly diagnosed chronic phase myeloid leukemia treated with Imatinib therapy:update from IRIS study(abstract). Blood 2004.104:10a. Abstract nr 21. 89. Cortes, J., et al., Result of high-dose imatinib mesylate in patients with Philadelphia chromosome-positive chronic myeloid leukemia after failure of interferon-alpha. Blood, 2003. 102(1): p. 83-6. 90. Kantarjian, H., et al., High-dose imatinib mesylate therapy in newly diagnosed Philadelphia chromosome-positive chronic phase chronic myeloid leukemia. Blood, 2004. 103(8): p. 2873-8. 91. Rousselot, P., et al., Imatinib mesylate discontinuation in patients with chronic myelogenous leukemia in complete molecular remission for more than 2 years. Blood, 2007. 109(1): p. 58-60. 92. Cortes, J., S. O'Brien, and H. Kantarjian, Discontinuation of imatinib therapy after achieving a molecular response. Blood, 2004. 104(7): p. 2204-5. 93. Merante, S., et al., Outcome of four patients with chronic myeloid leukemia after imatinib mesylate discontinuation. Haematologica, 2005. 90(7): p. 979-81. 94. La Rosee, P., et al., In vitro studies of the combination of imatinib mesylate (Gleevec) and arsenic trioxide (Trisenox) in chronic myelogenous leukemia. Exp Hematol, 2002. 30(7): p. 729-37. 95. Hoover, R.R., et al., Overcoming STI571 resistance with the farnesyl transferase inhibitor SCH66336. Blood, 2002. 100(3): p. 1068-71. 96. Gatto, S., et al., The proteasome inhibitor PS-341 inhibits growth and induces apoptosis in Bcr/Abl-positive cell lines sensitive and resistant to imatinib mesylate. Haematologica, 2003. 88(8): p. 853-63. 97. Nimmanapalli, R., et al., Histone deacetylase inhibitor LAQ824 both lowers expression and promotes proteasomal degradation of Bcr-Abl and induces apoptosis of imatinib mesylate-sensitive or -refractory 64

chronic myelogenous leukemia-blast crisis cells. Cancer Res, 2003. 63(16): p. 5126-35. 98. Jorgensen, H.G., et al., Enhanced CML stem cell elimination in vitro by bryostatin priming with imatinib mesylate. Exp Hematol, 2005. 33(10): p. 1140-6. 99. Tseng, P.H., et al., Synergistic interactions between imatinib mesylate and the novel phosphoinositide-dependent kinase-1 inhibitor OSU-03012 in overcoming imatinib mesylate resistance. Blood, 2005. 105(10): p. 4021-7. 100. Aloisi, A., et al., BCR-ABL nuclear entrapment kills human CML cells: ex vivo study on 35 patients with the combination of imatinib mesylate and leptomycin B. Blood, 2006. 107(4): p. 1591-8. 101. Marin, D., et al., Phase I/II trial of adding semisynthetic homoharringtonine in chronic myeloid leukemia patients who have achieved partial or complete cytogenetic response on imatinib. Cancer, 2005. 103(9): p. 1850-5. 102. Jorgensen, H.G., et al., Intermittent exposure of primitive quiescent chronic myeloid leukemia cells to granulocyte-colony stimulating factor in vitro promotes their elimination by imatinib mesylate. Clin Cancer Res, 2006. 12(2): p. 626-33. 103. Kantarjian, H.M., et al., Treatment of philadelphia chromosomepositive, accelerated-phase chronic myelogenous leukemia with imatinib mesylate. Clin Cancer Res, 2002. 8(7): p. 2167-76. 104. Lahaye, T., et al., Response and resistance in 300 patients with BCR-ABL-positive leukemias treated with imatinib in a single center: a 4.5-year follow-up. Cancer, 2005. 103(8): p. 1659-69. 105. Shimoni, A., et al., Imatinib mesylate (STI571) in preparation for allogeneic hematopoietic stem cell transplantation and donor lymphocyte infusions in patients with Philadelphia-positive acute leukemias. Leukemia, 2003. 17(2): p. 290-7. 106. Bornhauser, M., et al., Allogeneic haematopoietic cell transplantation for chronic myelogenous leukaemia in the era of imatinib: a retrospective multicentre study. Eur J Haematol, 2006. 76(1): p. 9-17. 107. Deininger, M.W., et al., Practical management of patients with chronic myeloid leukemia receiving imatinib. J Clin Oncol, 2003. 21(8): p. 1637-47. 108. Kerkela, R., et al., Cardiotoxicity of the cancer therapeutic agent imatinib mesylate. Nat Med, 2006. 12(8): p. 908-16. 109. Hatfield, A., S. Owen, and P.R. Pilot, In reply to 'Cardiotoxicity of the cancer therapeutic agent imatinib mesylate'. Nat Med, 2007. 13(1): p. 13; author reply 15-6. 110. Berman, E., et al., Altered bone and mineral metabolism in patients receiving imatinib mesylate. N Engl J Med, 2006. 354(19): p. 2006-13. 111. Hochhaus, A. and T. Hughes, Clinical resistance to imatinib: mechanisms and implications. Hematol Oncol Clin North Am, 2004. 18(3): p. 641-56, ix. 65

112. Shah, N.P., et al., Multiple BCR-ABL kinase domain mutations confer polyclonal resistance to the tyrosine kinase inhibitor imatinib (STI571) in chronic phase and blast crisis chronic myeloid leukemia. Cancer Cell, 2002. 2(2): p. 117-25. 113. Hochhaus, A., et al., Molecular and chromosomal mechanisms of resistance to imatinib (STI571) therapy. Leukemia, 2002. 16(11): p. 2190-6. 114. Gorre, M.E., et al., Clinical resistance to STI-571 cancer therapy caused by BCR-ABL gene mutation or amplification. Science, 2001. 293(5531): p. 876-80. 115. Branford, S., et al., High frequency of point mutations clustered within the adenosine triphosphate-binding region of BCR/ABL in patients with chronic myeloid leukemia or Ph-positive acute lymphoblastic leukemia who develop imatinib (STI571) resistance. Blood, 2002. 99(9): p. 3472-5. 116. Branford, S., et al., Detection of BCR-ABL mutations in patients with CML treated with imatinib is virtually always accompanied by clinical resistance, and mutations in the ATP phosphate-binding loop (P-loop) are associated with a poor prognosis. Blood, 2003. 102(1): p. 276-83. 117. O'Hare, T., et al., In vitro activity of Bcr-Abl inhibitors AMN107 and BMS-354825 against clinically relevant imatinib-resistant Abl kinase domain mutants. Cancer Res, 2005. 65(11): p. 4500-5. 118. Shah, N.P., et al., Overriding imatinib resistance with a novel ABL kinase inhibitor. Science, 2004. 305(5682): p. 399-401. 119. Talpaz, M., et al., Dasatinib in imatinib-resistant Philadelphia chromosome-positive leukemias. N Engl J Med, 2006. 354(24): p. 2531-41. 120. Hochhaus, A., et al., Dasatinib induces notable hematologic and cytogenetic responses in chronic-phase chronic myeloid leukemia after failure of imatinib therapy. Blood, 2007. 109(6): p. 2303-9. 121. Weisberg, E., et al., Characterization of AMN107, a selective inhibitor of native and mutant Bcr-Abl. Cancer Cell, 2005. 7(2): p. 129-41. 122. Kantarjian, H., et al., Nilotinib in imatinib-resistant CML and Philadelphia chromosome-positive ALL. N Engl J Med, 2006. 354(24): p. 2542-51. 123. Golas, J.M., et al., SKI-606, a 4-anilino-3-quinolinecarbonitrile dual inhibitor of Src and Abl kinases, is a potent antiproliferative agent against chronic myelogenous leukemia cells in culture and causes regression of K562 xenografts in nude mice. Cancer Res, 2003. 63(2): p. 375-81. 124. Puttini, M., et al., In vitro and in vivo activity of SKI-606, a novel Src-Abl inhibitor, against imatinib-resistant Bcr-Abl+ neoplastic cells. Cancer Res, 2006. 66(23): p. 11314-22. 66

125. Carter, T.A., et al., Inhibition of drug-resistant mutants of ABL, KIT, and EGF receptor kinases. Proc Natl Acad Sci U S A, 2005. 102(31): p. 11011-6. 126. Giles, F.J., et al., MK-0457, a novel kinase inhibitor, is active in patients with chronic myeloid leukemia or acute lymphocytic leukemia with the T315I BCR-ABL mutation. Blood, 2007. 109(2): p. 500-2. 127. Fiskus, W., et al., Combined effects of novel tyrosine kinase inhibitor AMN107 and histone deacetylase inhibitor LBH589 against Bcr- Abl-expressing human leukemia cells. Blood, 2006. 108(2): p. 645-52. 128. Bhatia, R., et al., Persistence of malignant hematopoietic progenitors in chronic myelogenous leukemia patients in complete cytogenetic remission following imatinib mesylate treatment. Blood, 2003. 101(12): p. 4701-7. 129. Bocchia, M., et al., Effect of a p210 multipeptide vaccine associated with imatinib or interferon in patients with chronic myeloid leukaemia and persistent residual disease: a multicentre observational trial. Lancet, 2005. 365(9460): p. 657-62. 130. Picard, S., et al., Trough imatinib plasma levels are associated with both cytogenetic and molecular responses to standard-dose imatinib in chronic myeloid leukemia. Blood, 2006. 131. Kantarjian, H.M., et al., Chronic myelogenous leukemia in blast crisis. Analysis of 242 patients. Am J Med, 1987. 83(3): p. 445-54. 132. Marktel, S., et al., Chronic myeloid leukemia in chronic phase responding to imatinib: the occurrence of additional cytogenetic abnormalities predicts disease progression. Haematologica, 2003. 88(3): p. 260-7. 133. Hochhaus, A., Cytogenetic and molecular mechanisms of resistance to imatinib. Semin Hematol, 2003. 40(2 Suppl 2): p. 69-79. 134. Kantarjian, H.M., et al., Results of treatment with hyper-cvad, a dose-intensive regimen, in adult acute lymphocytic leukemia. J Clin Oncol, 2000. 18(3): p. 547-61. 135. Giles, F.J., et al., Accelerated and blastic phases of chronic myelogenous leukemia. Hematol Oncol Clin North Am, 2004. 18(3): p. 753-74, xii. 136. Weisdorf, D.J., et al., Allogeneic bone marrow transplantation for chronic myelogenous leukemia: comparative analysis of unrelated versus matched sibling donor transplantation. Blood, 2002. 99(6): p. 1971-7. 137. Pappenheimer A. Experimental studies upon lymphocytes I. The reactions of lymphocytes under various experimental conditions. J Exp Med 1917;25:633. 138. Black, M.M. and F.D. Speer, Further observations on the effects of cancer chemotherapeutic agents on the in vitro dehydrogenase activity of cancer tissue. J Natl Cancer Inst, 1954. 14(5): p. 1147-58. 67

139. Larsson, R. and P. Nygren, Pharmacological modification of multidrug resistance (MDR) in vitro detected by a novel fluorometric microculture cytotoxicity assay. Reversal of resistance and selective cytotoxic actions of cyclosporin A and verapamil on MDR leukemia T-cells. Int J Cancer, 1990. 46(1): p. 67-72. 140. Jonsson, E., et al., Differential activity of topotecan, irinotecan and SN-38 in fresh human tumour cells but not in cell lines. Eur J Cancer, 2000. 36(16): p. 2120-7. 141. Fridborg, H., et al., Relationship between diagnosis-specific activity of cytotoxic drugs in fresh human tumour cells ex vivo and in the clinic. Eur J Cancer, 1999. 35(3): p. 424-32. 142. Aleskog, A., et al., In vitro drug resistance in B cell chronic lymphocytic leukemia: a comparison with acute myelocytic and acute lymphocytic leukemia. Anticancer Drugs, 2005. 16(3): p. 277-83. 143. Villman, K., et al., Predictive value of in vitro assessment of cytotoxic drug activity in advanced breast cancer. Anticancer Drugs, 2005. 16(6): p. 609-15. 144. Aleskog, A., et al., VH gene mutation status and cellular drug resistance in chronic lymphocytic leukaemia. Eur J Haematol, 2004. 73(6): p. 407-11. 145. Frost, B.M., et al., Translocation t(12;21) is related to in vitro cellular drug sensitivity to doxorubicin and etoposide in childhood acute lymphoblastic leukemia. Blood, 2004. 104(8): p. 2452-7. 146. Frost, B.M., et al., Translocation t(1;19) is related to low cellular drug resistance in childhood acute lymphoblastic leukaemia. Leukemia, 2005. 19(1): p. 165-9. 147. Hallbook, H., et al., In vitro activity of imatinib in cells from patients with adult acute lymphoblastic leukemia. Anticancer Drugs, 2005. 16(6): p. 631-4. 148. Larsson, R., et al., Laboratory determination of chemotherapeutic drug resistance in tumor cells from patients with leukemia, using a fluorometric microculture cytotoxicity assay (FMCA). Int J Cancer, 1992. 50(2): p. 177-85. 149. Nygren, P., et al., In vitro drug sensitivity testing of tumor cells from patients with non-hodgkin's lymphoma using the fluorometric microculture cytotoxicity assay. Ann Oncol, 1994. 5 Suppl 1: p. 127-31. 150. Nygren, P., et al., Feasibility of the fluorometric microculture cytotoxicity assay (FMCA) for cytotoxic drug sensitivity testing of tumor cells from patients with acute lymphoblastic leukemia. Leukemia, 1992. 6(11): p. 1121-8. 151. Schrag, D., et al., American Society of Clinical Oncology Technology Assessment: chemotherapy sensitivity and resistance assays. J Clin Oncol, 2004. 22(17): p. 3631-8. 152. Loizzi, V., et al., Survival outcomes in patients with recurrent ovarian cancer who were treated with chemoresistance assay-guided chemotherapy. Am J Obstet Gynecol, 2003. 189(5): p. 1301-7. 68

153. Frost, B.M., et al., Increased in vitro cellular drug resistance is related to poor outcome in high-risk childhood acute lymphoblastic leukaemia. Br J Haematol, 2003. 122(3): p. 376-85. 154. Thiesing, J.T., et al., Efficacy of STI571, an abl tyrosine kinase inhibitor, in conjunction with other antileukemic agents against bcrabl-positive cells. Blood, 2000. 96(9): p. 3195-9. 155. Scappini, B., et al., In vitro effects of STI 571-containing drug combinations on the growth of Philadelphia-positive chronic myelogenous leukemia cells. Cancer, 2002. 94(10): p. 2653-62. 156. Copland, M., et al., Dasatinib (BMS-354825) targets an earlier progenitor population than imatinib in primary CML but does not eliminate the quiescent fraction. Blood, 2006. 107(11): p. 4532-9. 157. Kano, Y., et al., In vitro cytotoxic effects of a tyrosine kinase inhibitor STI571 in combination with commonly used antileukemic agents. Blood, 2001. 97(7): p. 1999-2007. 158. Raetz, E.A. and P.J. Moos, Impact of microarray technology in clinical oncology. Cancer Invest, 2004. 22(2): p. 312-20. 159. Ladner, D.P., et al., Multiplex detection of hotspot mutations by rolling circle-enabled universal microarrays. Lab Invest, 2001. 81(8): p. 1079-86. 160. Venter, J.C., et al., The sequence of the human genome. Science, 2001. 291(5507): p. 1304-51. 161. Duggan, D.J., et al., Expression profiling using cdna microarrays. Nat Genet, 1999. 21(1 Suppl): p. 10-4. 162. Southern, E., K. Mir, and M. Shchepinov, Molecular interactions on microarrays. Nat Genet, 1999. 21(1 Suppl): p. 5-9. 163. Schena, M., et al., Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science, 1995. 270(5235): p. 467-70. 164. Lockhart, D.J., et al., Expression monitoring by hybridization to high-density oligonucleotide arrays. Nat Biotechnol, 1996. 14(13): p. 1675-80. 165. Sievertzon, M., P. Nilsson, and J. Lundeberg, Improving reliability and performance of DNA microarrays. Expert Rev Mol Diagn, 2006. 6(3): p. 481-92. 166. Quackenbush, J., Microarray analysis and tumor classification. N Engl J Med, 2006. 354(23): p. 2463-72. 167. Golub, T.R., et al., Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science, 1999. 286(5439): p. 531-7. 168. Haferlach, T., et al., Global approach to the diagnosis of leukemia using gene expression profiling. Blood, 2005. 106(4): p. 1189-98. 169. Yeoh, E.J., et al., Classification, subtype discovery, and prediction of outcome in pediatric acute lymphoblastic leukemia by gene expression profiling. Cancer Cell, 2002. 1(2): p. 133-43. 170. Chiaretti, S., et al., Gene expression profile of adult T-cell acute lymphocytic leukemia identifies distinct subsets of patients with dif- 69

ferent response to therapy and survival. Blood, 2004. 103(7): p. 2771-8. 171. Bohen, S.P., et al., Variation in gene expression patterns in follicular lymphoma and the response to rituximab. Proc Natl Acad Sci U S A, 2003. 100(4): p. 1926-30. 172. Shipp, M.A., et al., Diffuse large B-cell lymphoma outcome prediction by gene-expression profiling and supervised machine learning. Nat Med, 2002. 8(1): p. 68-74. 173. Bruchova, H., et al., Gene expression profiling in chronic myeloid leukemia patients treated with hydroxyurea. Leuk Lymphoma, 2002. 43(6): p. 1289-95. 174. Kaneta, Y., et al., Genome-wide analysis of gene-expression profiles in chronic myeloid leukemia cells using a cdna microarray. Int J Oncol, 2003. 23(3): p. 681-91. 175. Li, H., et al., CDNA microarray analysis of chronic myeloid leukemia. Int J Hematol, 2002. 75(4): p. 388-93. 176. Nowicki, M.O., et al., Chronic myelogenous leukemia molecular signature. Oncogene, 2003. 22(25): p. 3952-63. 177. Kronenwett, R., et al., Distinct molecular phenotype of malignant CD34(+) hematopoietic stem and progenitor cells in chronic myelogenous leukemia. Oncogene, 2005. 24(34): p. 5313-24. 178. Ohmine, K., et al., Characterization of stage progression in chronic myeloid leukemia by DNA microarray with purified hematopoietic stem cells. Oncogene, 2001. 20(57): p. 8249-57. 179. Radich, J.P., et al., Gene expression changes associated with progression and response in chronic myeloid leukemia. Proc Natl Acad Sci U S A, 2006. 103(8): p. 2794-9. 180. Zheng, C., et al., Gene expression profiling of CD34+ cells identifies a molecular signature of chronic myeloid leukemia blast crisis. Leukemia, 2006. 20(6): p. 1028-34. 181. Ohno, R. and Y. Nakamura, Prediction of response to imatinib by cdna microarray analysis. Semin Hematol, 2003. 40(2 Suppl 2): p. 42-9. 182. Kaneta, Y., et al., Prediction of sensitivity to STI571 among chronic myeloid leukemia patients by genome-wide cdna microarray analysis. Jpn J Cancer Res, 2002. 93(8): p. 849-56. 183. Frank, O., et al., Gene expression signature of primary imatinibresistant chronic myeloid leukemia patients. Leukemia, 2006. 20(8): p. 1400-7. 184. Villuendas, R., et al., Identification of genes involved in imatinib resistance in CML: a gene-expression profiling approach. Leukemia, 2006. 20(6): p. 1047-54. 185. Crossman, L.C., et al., In chronic myeloid leukemia white cells from cytogenetic responders and non-responders to imatinib have very similar gene expression signatures. Haematologica, 2005. 90(4): p. 459-64. 70

186. Yong, A.S., et al., Molecular profiling of CD34+ cells identifies low expression of CD7, along with high expression of proteinase 3 or elastase, as predictors of longer survival in patients with CML. Blood, 2006. 107(1): p. 205-12. 187. Kaplan EL, Meier P. Non-parametric astimation from incomplete observation.j AM Stat Assoc (B) 1958 ; 53: 457-481. 188. To, L.B., et al., A comparative study of the phenotype and proliferative capacity of peripheral blood (PB) CD34+ cells mobilized by four different protocols and those of steady-phase PB and bone marrow CD34+ cells. Blood, 1994. 84(9): p. 2930-9. 189. Dewald, G.W., et al., Highly sensitive fluorescence in situ hybridization method to detect double BCR/ABL fusion and monitor response to therapy in chronic myeloid leukemia. Blood, 1998. 91(9): p. 3357-65. 190. Tipping, A.J., et al., Drug responses of imatinib mesylate-resistant cells: synergism of imatinib with other chemotherapeutic drugs. Leukemia, 2002. 16(12): p. 2349-57. 191. Porosnicu, M., et al., Co-treatment with As2O3 enhances selective cytotoxic effects of STI-571 against Brc-Abl-positive acute leukemia cells. Leukemia, 2001. 15(5): p. 772-8. 192. Larsson, R., et al., Chemotherapeutic drug sensitivity testing of human leukemia cells in vitro using a semiautomated fluorometric assay. Leukemia, 1990. 4(8): p. 567-71. 193. Chou, T.C. and P. Talalay, Quantitative analysis of dose-effect relationships: the combined effects of multiple drugs or enzyme inhibitors. Adv Enzyme Regul, 1984. 22: p. 27-55. 194. Steel, G.G. and M.J. Peckham, Exploitable mechanisms in combined radiotherapy-chemotherapy: the concept of additivity. Int J Radiat Oncol Biol Phys, 1979. 5(1): p. 85-91. 195. Fryknas, M., et al., Molecular markers for discrimination of benign and malignant follicular thyroid tumors. Tumour Biol, 2006. 27(4): p. 211-20. 196. van 't Veer, L.J., et al., Expression profiling predicts outcome in breast cancer. Breast Cancer Res, 2003. 5(1): p. 57-8. 197. Hehlmann, R., et al., Randomized comparison of busulfan and hydroxyurea in chronic myelogenous leukemia: prolongation of survival by hydroxyurea. The German CML Study Group. Blood, 1993. 82(2): p. 398-407. 198. Carella, A.M. and F. Frassoni, ICE, mini-ice or high-dose hydroxyurea to mobilize Philadelphia (Ph1)-negative PBPC in chronic myelogenous leukaemia. Br J Haematol, 1996. 95(1): p. 213-5. 199. Chalmers, E.A., et al., Treatment of chronic myeloid leukaemia in first chronic phase with idarubicin and cytarabine: mobilization of Philadelphia-negative peripheral blood stem cells. Br J Haematol, 1997. 96(3): p. 627-34. 200. Pratt, G., et al., Autologous stem cell transplantation in chronic myeloid leukaemia using Philadelphia chromosome negative blood pro- 71

genitors mobilised with hydroxyurea and G-CSF. Bone Marrow Transplant, 1998. 21(5): p. 455-60. 201. Khouri, I.F., et al., Results with high-dose chemotherapy and unpurged autologous stem cell transplantation in 73 patients with chronic myelogenous leukemia: the MD Anderson experience. Bone Marrow Transplant, 1996. 17(5): p. 775-9. 202. Mehta, J., et al., Myelosuppressive chemotherapy to mobilize normal stem cells in chronic myeloid leukemia. Bone Marrow Transplant, 1996. 17(1): p. 25-9. 203. McGlave, P.B., et al., Autologous transplants for chronic myelogenous leukaemia: results from eight transplant groups. Lancet, 1994. 343(8911): p. 1486-8. 204. Bhatia, R., et al., Autologous transplantation therapy for chronic myelogenous leukemia. Blood, 1997. 89(8): p. 2623-34. 205. Bhatia, R. and P.B. McGlave, Autologous hematopoietic cell transplantation for chronic myelogenous leukemia. Hematol Oncol Clin North Am, 2004. 18(3): p. 715-32, xi. 206. O'Brien, S.G., et al., Imatinib compared with interferon and lowdose cytarabine for newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med, 2003. 348(11): p. 994-1004. 207. Talpaz, M., et al., Imatinib induces durable hematologic and cytogenetic responses in patients with accelerated phase chronic myeloid leukemia: results of a phase 2 study. Blood, 2002. 99(6): p. 1928-37. 208. Sawyers, C.L., et al., Imatinib induces hematologic and cytogenetic responses in patients with chronic myelogenous leukemia in myeloid blast crisis: results of a phase II study. Blood, 2002. 99(10): p. 3530-9. 209. Peng, B., et al., Pharmacokinetics and pharmacodynamics of imatinib in a phase I trial with chronic myeloid leukemia patients. J Clin Oncol, 2004. 22(5): p. 935-42. 210. Scott, L.M., et al., A novel temporal expression pattern of three C/EBP family members in differentiating myelomonocytic cells. Blood, 1992. 80(7): p. 1725-35. 211. Perrotti, D., et al., BCR-ABL suppresses C/EBPalpha expression through inhibitory action of hnrnp E2. Nat Genet, 2002. 30(1): p. 48-58. 212. Guerzoni, C., et al., Inducible activation of CEBPB, a gene negatively regulated by BCR/ABL, inhibits proliferation and promotes differentiation of BCR/ABL-expressing cells. Blood, 2006. 107(10): p. 4080-9. 72

Acta Universitatis Upsaliensis Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 256 Editor: The Dean of the Faculty of Medicine A doctoral dissertation from the Faculty of Medicine, Uppsala University, is usually a summary of a number of papers. A few copies of the complete dissertation are kept at major Swedish research libraries, while the summary alone is distributed internationally through the series Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine. (Prior to January, 2005, the series was published under the title Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine.) Distribution: publications.uu.se urn:nbn:se:uu:diva-7841 ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2007