Chronic Myeloid Leukemia

Introduction and Historical Perspective

Chronic myeloid leukemia (CML), a rare condition, has had a profound impact on the development of hematology/oncology and modern medicine as a whole ( Fig. 69.1A and B ). The story started in the 1830s with Alfred Francois Donné of Paris, a remarkable innovator who pioneered the use of microscopy in medicine and was the first to describe the different types of blood cells. When asked to examine the blood specimen of a 44-year-old woman suffering from a painless left-abdominal mass, he noticed the excess of white cells and speculated that the patient’s blood may contain pus. Donné also organized a Cours de Microscopie , which was attended by Edinburgh pathologist John Hughes Bennet who in 1845 published a “Case of Hypertrophy of the Spleen and Liver in which Death Took Place from Suppuration of the Blood.” Only a few weeks later Rudolf Virchow of Berlin reported on a very similar patient in an article titled “Weisses Blut” (white blood). Although we do not know for sure that these patients indeed suffered from CML, they highlight CML’s most salient features: leukocytosis and splenomegaly. There was disagreement about the cause of the new disease. While Bennett believed it was an infection, Virchow recognized its neoplastic nature and coined the term “leukemia,” from the Greek words “λενκον αιμα” meaning “white blood.” The ensuing academic dispute was settled cordially: Virchow acknowledged Bennett’s priority of discovery and Bennett that leukemia was a cancer of the blood. Perhaps it was the ability of these pioneers to put collaboration first and ego second that turned leukemia research into an engine of medical progress. Initial steps however were tedious. In 1872 Ernst Neumann established the bone marrow as the origin of leukemia and of blood cells in general, but then not much progress was made for almost a century. In 1951 William Dameshek, a visionary with a rare ability of seeing an organizing concept in the unstructured, posited that CML belonged to a larger group of related disorders characterized by overproduction of blood cells in the bone marrow. He called these disorders myeloproliferative disorders (MPDs), a name that stuck.

Figure 69.1, THE HISTORY OF CHRONIC MYELOID LEUKEMIA.

Just as better microscopes had enabled the first description of leukemia, improvements in chromosome banding led to a breakthrough by cytogeneticists Peter Nowell and David Hungerford from the city of Philadelphia. In 1960 they described an abnormally small G-group chromosome in metaphase spreads from several patients with CML. Ironically, in a now mostly forgotten manuscript, a team from Edinburgh led by Ishbel Tough reported similar findings just a few months later. The discovery of a consistent karyotypic abnormality associated with a cancer settled the debate as to whether DNA or proteins were responsible for transmitting the neoplastic phenotype to the next generation of cells. Anticipating that this was just the first in a long list of cancer chromosomes, the scientists called the new chromosome Philadelphia ¹ (Ph ¹ , now usually referred to as Ph) after the place of its discovery. One lesion in all CML cells suggested that the leukemia traced its origins back to a single ancestral cell. Formal proof for this idea was provided by Philip Fialkow who showed that the neutrophils from a female CML patient expressed only one isoform of the X chromosomal gene glucose-6-phosphate dehydrogenase (G6PD). This was consistent with inactivation of either a maternal or paternal X chromosome in all cells rather than the random inactivation characteristic of polyclonal tissues.

In 1973, Janet Rowley recognized that Ph was the product of a t(9;22)(q34;q11.1) reciprocal translocation rather than only a shortened chromosome 22. Over the following 25 years, the genetic anatomy and biochemical consequences of t(9;22)(q34;q11.1) were revealed with increasing resolution. Groups led by Eli Cannani, John Groffen/Nora Heisterkamp, and Owen Witte contributed to the identification of the genes juxtaposed by t(9;22). The chromosome 9 fusion partner was found to be ABL1 (formerly ABL ), the human homolog of v-abl, the oncogene of the Abelson murine leukemia virus (A-MuLV). On the derivative chromosome 22, ABL1 sequences consistently translocated to a genetic region that became known as the “breakpoint cluster region” ( BCR ), a name that was subsequently used for the new gene fused upstream of ABL1 . The next critical step was the discovery that BCR-ABL1, similar to v-Abl, was a constitutively active tyrosine kinase and that kinase activity was required for cellular transformation. Lastly, mice transplanted with bone marrow engineered to express BCR-ABL1 were shown to develop an MPD resembling CML, another milestone in the history of cancer research.

CML therapy developed slowly. Arsenicals, in use for cancer treatment since ancient times, were the only treatment available in the 19th century, usually in the form of Fowler’s solution, which contained potassium arsenite. The German physician Heinrich Lissauer is credited with the first publication on the remarkable efficacy of Fowler’s solution in a patient with leukemia. We may suspect that the reviewers of his paper gave him a pass on novelty, since the use of arsenic for cancer therapy had been described by the poet Valmiki in the Indian Ramayana in approximately 500 BC. As a curiosity, in an 1882 Lancet paper, Conan Doyle, better known for his Sherlock Holmes detective stories, published on a patient with the clinical presentation of CML who achieved a partial response to arsenic. In the 1920s, radiotherapy was introduced for symptomatic relief of tender splenomegaly, and remained the mainstay of CML therapy for the first half of the 20th century. Busulfan (1959), an alkylating agent discovered at the Chester Beatty Laboratories in London, was the first drug that reliably controlled white blood cell counts in CML and has retained a place in conditioning regimens for allogeneic hematopoietic cell transplantation (HCT). Ten years later, hydroxyurea was introduced as the first intervention that prolonged survival, although by only a small margin. A breakthrough was achieved in the late 1970s when the Seattle group reported the disappearance of Ph in CML patients who underwent HCT, the first cures of CML. Soon after that, interferon-α (IFNα) was found to induce durable complete cytogenetic responses (CCyRs) and long-term survival in 10% to 20% of patients. In 1992, Alexander Levitzki proposed the use of ABL inhibitors to treat leukemias driven by ABL oncogenes, but unfortunately, none of his so-called tyrphostins were developed for clinical use. At about the same time, scientists at Ciba-Geigy had synthesized a potent inhibitor of ABL1 termed GCP57148B that is now known as imatinib. Of note, GCP57148B’s anti-ABL activity was the serendipitous byproduct of a drug development effort to identify inhibitors of platelet-derived growth factor receptor (PDGFR) for cardiovascular indications. Clinical trials initiated by Brian Druker, against the skepticism of the manufacturer, rapidly established the compound’s activity in patients with CML and revolutionized CML therapy. Solving the crystal structure of ABL1 in complex with imatinib provided the foundation for subsequent studies into the auto-regulation of ABL1 kinase activity and its disruption in BCR-ABL1. These studies were of fundamental importance for our understanding of kinases and their regulation in physiological and aberrant signal transduction. Despite imatinib’s general efficacy, subsets of patients developed resistance, frequently due to point mutations in BCR-ABL1, stimulating the development of subsequent generations of tyrosine kinase inhibitors (TKIs). The latest development is asciminib, an allosteric ABL inhibitor that mimics the physiological autoinhibition of ABL1.

There is compelling evidence that the most primitive CML cells are not dependent on BCR-ABL1 kinase activity, explaining why residual leukemia remains detectable in many CML patients treated with TKIs. Surprisingly a plethora of studies has demonstrated that 40% to 60% of patients with deep molecular responses (DMRs) maintain remission after discontinuation of TKIs, a state termed treatment-free remission (TFR). While survival remains the universal goal for all CML patients, TFR has gained traction as a new target for selected patient populations, particularly young individuals.

The unprecedented success of TKIs to treat CML established a paradigm for molecularly targeted therapy that informs approaches in other types of cancer. However, many scientific, clinical, and societal questions remain. CML mortality is not zero, and the prognosis for patients who progress to blast phase CML (BP-CML) remains very poor. Despite improved understanding of blastic transformation at the molecular level, no fundamentally new treatments have emerged beyond chemotherapy followed by HCT. At the other extreme, increasing the pool of TRF eligible patients and improving the success rate will be crucial to minimize long-term TKI toxicity. However, as the biological underpinnings of TFR have remained largely elusive, rational approaches to improve over the current situation are lacking. Last but not least, the exorbitant costs of TKIs, particularly in the United Stated, raise ethical and societal questions that cannot be answered on medical grounds alone. Again, CML has prepared the ground for fundamental discussions far beyond what one might expect from an orphan disease.

Epidemiology

CML occurs with an annual incidence of 1.0 to 1.5 per 100,000, without significant racial or geographic differences, but a slight male predominance of approximately 1.5. The median age at CML diagnosis is above 60 years in the developed world, but lower in developing countries, reflecting differences in access to medical care or—much less likely—disease biology. CML is generally rare in children, but cases have been reported even in infants. Familial occurrence has been observed, but there is no evidence for a general familial predisposition. The only well-documented risk factor is ionizing radiation. In the survivors of the atomic bombings in Japan, CML incidence increased, peaking 8 to 10 years post exposure. Elevated rates of CML were also seen in patients treated with radiotherapy for ankylosing spondylitis and those exposed to thorotrast, an alpha-emitting contrast medium widely used in the 1930s and 1940s. In contrast, there is no association between benzene exposure and CML. As a result of the vastly improved survival afforded by TKIs, CML prevalence in the United States is predicted to rise from 70,000 in 2010 to 180,000 in 2050. At that point CML will become the most prevalent myeloid neoplasm, with considerable implications for health care costs.

Molecular Biology

BCR-ABL1 Translocation and Fusion Gene ( Chapter 57 )

Ph is the result of a reciprocal translocation between the long arms of chromosomes 9 and 22 [t(9;22)(q34;q11.2)] that fuses sequences from the ABL1 gene on 9q34 downstream of the BCR gene on 22q11 ( Fig. 69.2A ). The chromosome 22 breaks localize almost exclusively to one of three BCRs and determine which parts of BCR are retained in the BCR-ABL1 fusion mRNA and protein. In contrast, the chromosome 9 breaks can occur over a large genetic region, 5′ of ABL1 exon Ib, 3′ of ABL1 exon Ia, or most commonly between the two alternative first ABL1 exons ( Fig. 69.3B ). Despite the different breakpoints, splicing almost invariably leads to fusion mRNAs that encompass ABL1 exons 2 to 11. Breakpoints in the minor BCR (m -BCR) generate an e1a2 fusion mRNA and p190 ^BCR-ABL1 , which is found in two-thirds of Ph ⁺ acute lymphoblastic leukemia (ALL) cases, but very rarely in CML, where it is associated with monocytosis and an aggressive clinical course. The origin and biology of p190 ^BCR-ABL1 -positive CML is mysterious. Either m- BCR rearrangements are much less likely to occur in hematopoietic stem cells (HSCs) than the major BCR (M-BCR) rearrangements typical of CML, or the cell of origin is different, or an as yet unknown concomitant somatic mutation creates a state conducive to transformation by p190 ^BCR-ABL1 . In the vast majority of CML patients, the break localizes to the M- BCR , generating e13a2 or e14a2 (formerly referred to as b2a2 and b3a2) fusion mRNAs and a p210 ^BCR-ABL1 fusion protein. p230 ^BCR-ABL1 , the largest of the fusion proteins, is derived from a break in the micro BCR (μ- BCR ) that generates e19a2 mRNA. p230 ^BCR-ABL1 is associated with neutrophilic predominance and a more benign disease. There are a number of rare BCR-ABL1 variants, some of which may be associated with a more aggressive clinical course. Additionally variants can give rise to misleading reverse transcription polymerase chain reaction (RT-PCR) results and cannot be followed using standard quantitative PCR (qPCR) assays. Approximately two-thirds of CML cases also express the reciprocal ABL1-BCR mRNA, but there is no definitive evidence that this influences disease biology or prognosis. However, absence of ABL1-BCR mRNA is an indirect indication for deletions flanking the breakpoints in BCR , ABL1 , or both. For unknown reasons, these deletions are associated with significantly reduced survival in patients treated with IFN-α, but their adverse impact is reduced or abolished by TKI therapy.

Figure 69.2, GENETIC ANATOMY OF THE BCR-ABL1 TRANSLOCATION.

Figure 69.3, STRUCTURAL DOMAINS AND SIGNALING MOTIFS IN BCR, ABL1, AND BCR-ABL1 PROTEINS.

BCR and ABL1 Proteins and Their Contribution to Cellular Transformation

BCR is a ubiquitously expressed 160kD cytoplasmic protein encompassing several functional domains, including an N-terminal coiled-coil motif, followed by a serine/threonine kinase motif whose only known substrate is the 14-3-3 family protein Bap-1, a dbl -like (DH) domain and a pleckstrin-homology (PH) domain (see Fig. 69.3A ). DH had been thought to stimulate exchange of guanidine triphosphate (GTP) for guanidine diphosphate (GDP) on Rho guanidine exchange factors, but recent structural and biochemical studies failed to confirm this. The PH domain binds to various phosphatidylinositol (PI)-phosphates, impacts subcellular protein localization, and is required for interactions with selected proteins. The BCR C-terminus contains a putative calcium-dependent lipid binding (CaLB) site and a GTPase activating function (RAC-GAP). The latter constrains activity of Ras-related C3 botulinum toxin substrate (RAC), a small GTPase that regulates actin polymerization and an nicotinamide adenine dinucleotide phosphate hydrogen (NADPH) oxidase in phagocytic cells. Bcr ^−/− mice are viable and fertile, the only known defect being an increased oxidative burst in neutrophils. BCR can be phosphorylated on several tyrosine residues, most importantly tyrosine 177, which binds growth factor receptor bound protein 2 (GRB2), an adapter molecule involved in rat sarcoma virus (RAS) activation in CML cells. The question why p190 ^BCR-ABL1 and p210 ^BCR-ABL1 are associated with ALL and CML, respectively, has long been a matter of debate. Recent work has shown that the protein networks entertained by p190 ^BCR-ABL1 and p210 ^BCR-ABL1 are profoundly different, and the differences have been attributed to the retention of the PH domain in p210 ^BCR-ABL1 . It remains unclear how this impacts lineage commitment and why HSCs are susceptible to transformation by p210 ^BCR-ABL , while p190 ^BCR-ABL1 targets pre-B cells.

ABL1 is the human homologue of the v- abl oncogene of the A-MuLV. The 145-kDa ABL1 protein is ubiquitously expressed and encodes a non-receptor tyrosine kinase. Two isoforms arise from alternative splicing of the first exon. ABL1a lacks the first 19 amino acids, which are retained in ABL1b and include a glycine in position 2 that can be myristoylated. The N-terminal region contains three SRC homology domains (SH1 to SH3). SH1 carries the tyrosine kinase function, while the SH2 and SH3 domains engage in protein-protein interactions and participate in regulation of kinase activity (see Fig. 69.3B ). The center of ABL1 features proline rich sequences that interact with SH3 domains of other proteins such as Crk. The C-terminus contains nuclear localization, DNA binding, and actin-binding motifs. Several functions have been attributed to ABL1, including inhibition of cell cycle progression and proliferation, integrin signaling, and DNA repair. Mice with germline deletion of Abl1 exhibit high perinatal mortality, runting, and skeletal and immune system defects. ABL1 kinase activity is tightly controlled in physiological conditions through allosteric mechanisms that involve a myristate-binding pocket at the base of the SH1 domain. Mice null for Abl1 and the Abl -related gene (formerly Arg , now Abl2 ) are embryonically lethal due to absence of neurulation. Gain of function in the germline is also detrimental, as shown by a recent report of mutations in the myristate binding pocket that confer constitutive kinase activation and are associated with cardiac and skeletal defects. The emerging complex picture suggests that ABL1 may be best characterized as a cellular module that integrates extra- and intracellular signals to influence decisions with regard to cell cycle and apoptosis. The latter observation raised considerable concerns about the potential side effects of clinical ABL inhibitors that fortunately were not confirmed.

Constitutive Kinase Activation in the BCR-ABL1 Fusion Protein

In contrast to ABL1, BCR-ABL1 is constitutively active and excluded from the nucleus. Revealing the mechanism by which replacing the ABL1 N-terminus with BCR sequences leads to kinase activation has led to deep insights into the regulation of ABL1. The coiled-coil domain of BCR (see Fig. 69.3A ) promotes dimerization, which allows for an initial transphosphorylation event, followed by autophosphorylation of additional tyrosine residues that fully activate the kinase. Other proteins with the ability to form dimers can substitute for BCR, such as ETV6 and NUP214 that form ABL1 fusion genes associated with a myeloproliferative neoplasm (MPN) or ALL. The N-terminal “cap” region of ABL1, when myristoylated, binds to a hydrophobic pocket at the base of the kinase domain, resulting in a conformation that resembles a latch that holds the kinase in an inactive state. However, small molecules binding to the myristate pocket inhibit kinase activity, indicating that the concept of a “mechanical” latch is too simplistic. Rather than that, an allosteric mechanism controls kinase activity that does not require the ABL1 N-terminal sequences and hence is preserved and amenable to therapeutic targeting in BCR-ABL1. How precisely information is communicated from the myristate pocket to the catalytic site is unknown. Recent data suggest that reverse signaling is also possible, as mutations in the ATP site can influence binding of small molecules to the myristate pocket. Additional levels of regulation exist: certain residues within the SH2 domain participate in the regulation of kinase activity, and trans-acting binding partners such as the ABL1 interacting proteins have been implicated as physiological inhibitors.

Several domains or amino acid residues in BCR-ABL1 are critical for cellular transformation. In the ABL1 portion they include the kinase domain, SH2, and actin-binding domains (see Fig. 69.3C ). Additionally the SH3 domain contributes to aberrant regulation of homing, adhesion, and migration. Transformation-relevant domains in the BCR portion include a coiled–coil motif contained in amino acids 1 to 63, the tyrosine at position 177, and phosphoserine–threonine-rich sequences between amino acids 192 to 242 and 298 to 413. Whether or not a particular feature is essential may depend on the cellular context. For example, SH2 deletion mutants of BCR-ABL1 are defective for fibroblast transformation, but they retain the capacity to transform cell lines to factor independence and are leukemogenic in mice.

Aberrant Signaling in BCR-ABL1 Transformed Cells

BCR-ABL1 signaling is one of the most intensely studied areas in biomedical research. Three decades of research have identified numerous BCR-ABL1 substrates, binding partners, and downstream signaling molecules involved in cellular transformation, and the number is still growing ( Fig. 69.4 ). Efforts have been directed at linking these pathways to specific characteristics of CML cells, such as increased proliferation, reduced apoptosis, abnormal adhesion and migration, or genetic instability. The accumulated data are overwhelming, and a comprehensive review of the multiple pathways implicated in BCR-ABL1 transformation is beyond the scope of this chapter. As a note of caution, despite evidence from CML cell lines and primary CML cells, mouse models using homozygous deletions of signaling proteins have identified very few pathway components with a truly essential role for BCR-ABL1–induced leukemia. This speaks to the extensive redundancy of the BCR-ABL1 transformation network and/or to the limitations of the disease models.

Figure 69.4, GROSSLY SIMPLIFIED REPRESENTATION OF SIGNALING PATHWAYS ACTIVATED IN CML CELLS.

Phosphatidylinositol-3 Kinase

BCR-ABL1 activates phosphatidylinositol-3 kinase (PI3K) through at least two separate mechanisms. The first is initiated by autophosphorylation of BCR-ABL1 tyrosine 177 (Y177), which generates a docking site for the GRB2 adapter protein. GRB2 recruits GAB2, another adapter, into a complex that activates PI3K. Alternatively, PI3K can be activated by complex formation between the PI3K p85 regulatory subunit, and the BCR-ABL1 substrates casistas B-lineage lymphoma (CBL) and CRK like (CRKL), which bind to the SH2 and proline-rich domains of BCR-ABL1. The main downstream target of PI3K is the serine-threonine kinase AKT, a major conduit for oncogenic signals in many types of cancer. In BCR-ABL1 ⁺ cells, AKT enhances survival by phosphorylating forkhead O transcription factor 3a (FOXO3a) and pro-apoptotic proteins like BCL2 associated agonist of cell death (BAD). Phosphorylated FOXO3 is excluded from the nucleus, leading to downregulation of important transcriptional targets such as p27. Phosphorylated BAD is no longer able to neutralize anti-apoptotic BCL2 family members such as BCL2 and BCLX _L , promoting survival. Furthermore, AKT activates mammalian target of rapamycin (mTOR), which phosphorylates ribosomal proteins p70S6 kinase (S6K) and 4E-BP1. S6K, a serine/threonine kinase, phosphorylates multiple substrates that collectively promote cell proliferation. Phosphorylated mTOR has been shown to inactivate 4EBP-1, releasing the translation initiation factor eIF4E from inhibition to enhance protein synthesis. In addition to activating AKT, PI3K promotes proteasomal degradation of p27 through upregulation of SKP2, the F-Box recognition protein of the SCF ^SKP2 E3 ubiquitin ligase. Accordingly, absence of SKP2 from leukemia cells prolongs survival in a murine CML model.

RAS/Mitogen-Activated Protein Kinase Pathways

A long-known pathway of RAS activation is phosphorylation of tyrosine 177 with recruitment of GRB2 and Son of Sevenless (SOS) into a complex that promotes exchange of GTP for GDP on RAS. GTP- bound RAS activates the serine-threonine kinase RAF-1, which in turn activates mitogen-activated protein kinase (MAPK). Recently, a genetic screen identified inactivation of LZTR1 as a promoter of RAS activation in TKI-treated CML cells. Mechanistically LZTR1 controls KRAS expression by enhancing interaction with the CUL3 E3 ligase, promoting proteasomal degradation. Other small GTPases like RAC1/2 have also been implicated in BCR-ABL1 signaling and genetic absence of RAC1/2 has been shown to attenuate BCR-ABL1 leukemia in a murine model.

Janus Kinase (JAK)/Signal Transducer and Activator of Transcription (STAT) Pathway

The STAT5 transcription factor is constitutively tyrosine-phosphorylated in CML cells. Initial experiments in mice failed to demonstrate a critical role for STAT5. However, it was subsequently discovered that the presumed Stat5 knockout mice used in this study were not null for Stat5 , but expressed a partially functional N-terminally deleted protein. While STAT5 is not required for the initiation of CML, there is convincing evidence that it plays a rate-limiting role in leukemogenesis. Absence of only STAT5A or overexpression of a dominant-negative STAT5 mutant in BCR-ABL1 expressing bone marrow cells attenuates CML. Conditional deletion of STAT5 prevents the establishment of active CML in mice, but does not eliminate the BCR-ABL1 expressing cells. The consequence of constitutive STAT5 activation is inhibition of apoptosis by enhancing transcription of anti-apoptotic proteins like MCL-1 and BCL-X _L . BCR-ABL1 may activate STAT5 through direct phosphorylation or indirectly by promoting phosphorylation by HCK or JAK2. The role of JAK2 itself is controversial. Previous studies had suggested that JAK2 plays an important if not essential role, while a recent report using conditional knockout technology showed that murine CML does not require JAK2. In contrast, JAK2 may have an important role in promoting survival of CML stem cells (from here onward referred to as LSCs).

Cytoskeletal Proteins

Several BCR-ABL1 substrates are involved in adhesion and migration, including focal adhesion kinase (FAK), CRKL, paxillin, p130CAS, and HEF1. It is thought that this aberrant phosphorylation is responsible for the defect in integrin-mediated adhesion of CML progenitors to bone marrow stroma and extracellular matrix. Alternatively, integrin function may be compromised through RAS activation or BCR-ABL1 binding to F-actin. Lastly one report found that a kinase inactive BCR-ABL1 variant causes an adhesion defect, consistent with a process that is independent of BCR-ABL kinase activity. Given that adhesion of integrins inhibits proliferation of hematopoietic progenitors, the defect in integrin function may contribute to the premature circulation as well as the abnormal proliferation of Ph ⁺ progenitor cells.

DNA Damage Surveillance and Repair Pathways

Without effective therapy, chronic phase of CML (CP-CML) invariably progresses to BP-CML, testimony to the profound genetic instability of CML cells. CML CD34 ⁺ cells contain more reactive oxygen species (ROS) compared to normal controls, and this is particularly pronounced in CML-BP cells. ROS cause DNA damage by generating oxidized bases and double-strand breaks (DSB), which are up to eightfold more frequent in CML than in normal cells. While ROS generation in CML cell lines and progenitor cells is controlled by BCR-ABL1 through activation of PI3K, studies in mice have shown that LSCs continue to accumulate ROS-induced DNA damage despite TKI inhibition of BCR-ABL1. This suggests that genetic evolution may continue in patients on TKIs as long as they harbor residual leukemia. Fortunately, this concern was not borne out in the clinic, at least in patients with a CCyR or major molecular response (MMR), whose progression rate is low. Additional mutagenic mechanisms operational in lymphoid blast phase cells are activation of activation-induced cytidine deaminase (AID), which promotes point mutations, and RAG-mediated recombination, which promotes structural recombination events. The consequences of increased DNA damage are aggravated by the impairment of DNA damage surveillance and repair. Multiple mechanisms have been implicated, including BCR-ABL1 kinase-independent mechanisms. For instance, BCR-ABL1 impairs the intra-S-phase cell cycle checkpoint through suppression of checkpoint kinase 1 (CHK1), either by inhibiting the nuclear protein kinase ATR or by downregulating BRCA1, a substrate of ataxia telangiectasia mutated (ATM). There is also aberrant regulation of DSB repair pathways. In addition to BRCA1, the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs) is downregulated in CD34 ⁺ CML progenitors, leading to error-prone non-homologous end-joining (NHEJ), rather than more faithful D-NHEJ, thereby enhancing the effects of BCR-ABL1–induced ROS. While these deficiencies result in genomic instability and increase the risk of progression, they also offer therapeutic opportunities. Cycling cells have access to a range of DNA repair mechanisms, including homologous recombination (HR), single strand annealing (SSA), transcription-associated homologous recombination (TA-HR), and NHEJ. In contrast, quiescent cells have limited options, using mostly NHEJ ( Fig. 69.5A ). There are differences in the usage of the various DNA repair pathways between LSCs and HSCs that in the future may provide individualized therapeutic opportunities, based on the principle of synthetic lethality. For instance BRCA1/2 is frequently downregulated in CML. Additionally, some cases have downregulated one of the components of D-NHEJ, rendering them vulnerable to a combination of PARP inhibitors such as olaparib and inhibitors of RAD51, while cycling HSCs can resort to BRCA1/2-based HR, and quiescent HSCs to D-NHEJ (see Fig. 69.5B, C ). There is little evidence that TKIs themselves cause DNA damage, although imatinib was reported to induce centrosome abnormalities in vitro. Conflicting data have been reported regarding the risk of second (non-hematologic) malignancies in CML patients on TKIs but there is consensus that any excess incidence is related to CML rather than TKI therapy. Similarly there is no evidence that the clonal cytogenetic abnormalities in Ph-negative cells (CCA/Ph ⁻ ) detected in 5% to 10% of CML patients with a cytogenetic response to TKI therapy are induced by TKIs.

Figure 69.5, DNA REPAIR PATHWAYS IN CYCLING AND QUIESCENT HSCS AND LSCS.

A wealth of data on BCR-ABL1 signaling has been amassed, yet a complete picture is still elusive. As conventional approaches have typically focused on single pathways, efforts are now underway to characterize more comprehensively BCR-ABL1 signaling by quantitative genomics. These data suggest that cellular processes in CML rely on integrated networks rather than single pathways, which cooperate to fully realize the leukemogenic potential of BCR-ABL1.

Cellular Biology

Cell of Origin

Current thinking holds that BCR-ABL1 is both necessary and sufficient to produce the CP-CML. Unlike some fusion genes associated with acute myeloid leukemia (AML), such as MOZ-TIF2 , BCR-ABL1 does not confer HSC self-renewal, implying that the initial translocation event must occur in a multipotent hematopoietic HSC already endowed with this property. Evidence in support of this comes from mouse models and clinical observations. Mice transplanted with bone marrow cells infected with a BCR-ABL1 retrovirus develop an oligoclonal CML-like MPN, suggesting that a second event is not required. Similarly mice expressing a doxycycline-repressible BCR-ABL1 allele develop CML upon withdrawal of doxycycline, which regresses upon re-exposure to doxycycline. TKIs restore polyclonal hematopoiesis as assessed by X chromosome inactivation. However the clinical variability of CP-CML, particularly the highly variable time to transformation to AP/BP-CML, indicates the presence of major disease modifying factors. Between 5% and 10% of CP-CML cases have additional clonal cytogenetic abnormalities in Ph ⁺ cells (CCA/Ph ⁺ ) at diagnosis, and this is associated with a more aggressive clinical course. In almost all cases, CCA/Ph ⁺ disappear upon achievement of a cytogenetic response, suggesting they reflect evolution of the Ph ⁺ cell clone ( Fig. 69.6 ). Next generation sequencing (NGS) has revealed mutations in epigenetic modifier genes such as TET2, DNMT3A , and ASXL1 in ∼30% of CP-CML patients at diagnosis. In most but not all cases, these mutations became undetectable at the time of cytogenetic or molecular response, indicating that CML may develop either truly de novo or on the background of clonal hematopoiesis (CH) (see Fig. 69.6 ). This model is consistent with the observation that CML can occur at any age, while complex myeloid neoplasms such as chronic myelomonocytic leukemia (CMML) almost invariably develop from preceding CH and as such are restricted to older individuals. The impression is that patients with additional somatic mutations at diagnosis have a worse prognosis, but this is less well established compared to CCA/Ph ⁺ , and more data is needed. That somatic mutations in addition to Ph may happen at any point during the evolution of CML speaks to the remarkable capacity of BCR-ABL1 to drive the CML phenotype, either alone or on a genetically modified background.

Figure 69.6, ORIGIN AND DEVELOPMENT OF THE CHRONIC PHASE OF CHRONIC MYELOID LEUKEMIA.

Chronic Myeloid Leukemia Hematopoiesis

Unlike in acute leukemia, the hierarchical organization of hematopoiesis is initially maintained in CML, although there is lineage imbalance in favor of the myeloid series and a left-shift of granulopoiesis. Although maturation is delayed, cellular function is mostly normal, evidenced by the fact that CML patients in the chronic phase are not at increased risk of infection or bleeding. BCR-ABL1 is detected in cells belonging to all hematopoietic lineages, including B cells and T cells. Consistent with the clinical presentation, the thrust of BCR-ABL1 –induced cellular expansion targets the myeloid progenitor cell compartment, which is almost exclusively Ph ⁺ (see Fig. 69.6 ). In contrast, in most newly diagnosed patients the majority of the most primitive cells (defined as long-term culture initiating cells, LTC-ICs, or quiescent CD34 ⁺ cells) are partially or even predominantly Ph ⁻ . These normal HSCs are the basis for cytogenetic and molecular responses to therapy. The cause of the myeloid bias is incompletely understood. One possible explanation is that BCR-ABL1 alone is insufficient to overcome the powerful metabolic growth restraint placed on lymphoid precursor cells, which is thought to minimize the risk of malignant transformation during V(D)J rearrangement or T-cell receptor diversification. As such BCR-ABL1-positive chronic lymphocytic leukemia does not exist. However, when disruption of orderly B-cell (and rarely T-cell) differentiation abrogates the growth restraint, BCR-ABL1-positive ALL ensues. In the B-cell lineage this is accomplished by inactivation of lineage-specific transcription factors such as IKZF1 or PAX5.

Leukemia stem cells (LSCs) and what distinguishes them from HSCs have been the topic of intense research. It is important to remember that the ultimate diagnostic test for stemness is functional—the ability of a cell to generate progeny of all lineages in vivo, while phenotypic markers are surrogates by definition. As such, the terms HSC and LSC frequently denote populations of cells that are enriched for functionally defined “true” stem cells, but not pure populations. For instance, in many studies stem cells are conveniently defined as Lineage ⁻ CD34 ⁺ CD38 ⁻ , but less than 10% of these cells are functionally stem cells. As of 2020, no universal immunophenotypic marker had been identified that separates LSCs from HSCs, but strategies using multiple markers have achieved considerable enrichment for BCR-ABL ⁺ cells. The issue is complicated by the fact that some but not all markers are regulated by BCR-ABL1 kinase activity, and that TKIs predominantly eliminate proliferating LSCs with a “late” myeloid signature, while relatively sparing the most primitive LSCs. As a consequence, the immunophenotype of LSCs in untreated CML patients is different from that in patients with residual disease on TKIs ( Table 69.1 ). Some clinical features of CML can be linked to biological aberrancies of LSCs. For instance LSCs exhibit reduced integrin-mediated adhesion to bone marrow stroma and abnormal migration to CXCL12 (SDF1), which may account for leukocytosis and extramedullary hematopoiesis in spleen and liver. Transcriptomic profiling of CD34 ⁺ 38 ⁻ cells showed that quiescent LSCs are primed to proliferate compared to HSCs, explaining their hypersensitivity to cytokines such as IL-3 and G-SCF.

Table 69.1

Immunophenotypic Markers of Primitive Chronic Myeloid Leukemia Cells in Chronic Phase at Diagnosis and on Tyrosine Kinase Inhibitor Therapy

Adapted from Warfvinge R, Geironson L, Sommarin MNE, et al. Single-cell molecular analysis defines therapy response and immunophenotype of stem cell subpopulations in CML. Blood . 2017;129(17):2384–2394.

	HSC	LSC (Diagnosis)	LSC (on TKI)
Lineage	−	−	−
CD34	+++	+++	+
CD38	Low/−	Low/−	Low/−
CD11c	+	+	−
CD25	−	++	−
CD26	Low/−	+	+
CD32 (FcγRII)	+	++	++
CD45RA	Low/−	Low/−	−
CD90	+++	++	++
CD117	+++	+	−
CD276	−	−	−
TIM3	−	−	−
IL1RAP	−	++	−
ITGB7	−	−	−

HSC , Hematopoietic stem cell; TKI , tyrosine kinase inhibitor.

A question of considerable clinical importance is whether CML stem cells are dependent (“addicted”) to BCR-ABL1 or not. The fact that residual BCR-ABL1 ⁺ cells remain detectable by PCR in most patients on TKIs has been interpreted as evidence for persistence of LSCs despite continued suppression of BCR-ABL1 kinase activity. Consistent with this, primitive CML cells survive TKI treatment ex vivo. However, recent data have revealed that long-lived B cells and less frequently T cells account for most of the positive PCR results in patients in TFR, while granulocytes are consistently BCR-ABL1-negative. These data do not prove that multipotent CML stem cells are absent, but they explain the puzzling observation of positive PCR results in patients with stable TFR for years. Conceptually the central question is whether an LSC with kinase-disabled BCR-ABL1 is equal to a normal HSC or is still “rewired.” This could happen through at least three mechanisms. First there is compelling data that some BCR-ABL1 functions are kinase-independent. Whether these functions alone are sufficient to support long-term survival of LSCs is unknown. Second epigenetic reprogramming as a result of exposure to unopposed BCR-ABL1 kinase activity could persist even when BCR-ABL1 is no longer active. Third the presence of a large multi-domain protein like BCR-ABL1 may interfere with physiological signaling, for instance through sequestration of adaptor proteins. Indeed considerable evidence has accumulated to suggest that LSCs exhibit specific vulnerabilities that may be exploitable therapeutically, either alone or in combination with TKIs ( Table 69.2 ). Although several pathways such as β-catenin, PP2A, and JAK2 are supported by multiple independent studies, translation of these data into hypothesis-driven clinical trials has been excruciatingly slow. As a result, conclusive data regarding clinical efficacy are mostly missing or inconclusive, and thus far no new clinical approaches have emerged that could be considered standard of care.

Table 69.2

Pathways Implicated in the Survival of CML Stem Cells

Pathway	Publications	Clinical Trial	Status	Final Report
WNT/β-Catenin	Zhao et al., Cancer Cell 2007 ¹ ; McWeeney et al., Blood 2010 ² ; Heidel et al., Cell Stem Cell 2012 ³ ; Schürch et al., JCI 2012 ⁴ ; Lim et al., PNAS 2013 ⁵ ; Zhang et al., Blood 2013 ⁶ ; Eiring et al., Leukemia 2015 ⁷ ; Agarwal et al., Blood 2017 ⁸	No	NA	NA
HDAC	Zhang et al., Cancer Cell 2010 ^9,10	Panobinostat	Lack of efficacy	No
Hedgehog	Dierks et al., Cancer Cell 2008; Zhao et al., Nature 2009 ¹¹	LDE225+NIL BMS-833923+DAS	Terminated Terminated	No No
5-Lipoxygenase	Chen et al., Nat Genet. 2009 ¹²	Zileuton+IM	Terminated	No
BCL6	Hurtz et al., J Exp Med. 2011 ¹³	No	NA	NA
MYC	Ravie et al., Cancer Cell 2013 ¹⁴ Abraham et al., Nature 2016 ¹⁵	No	NA	NA
PP2A	Neviani et al., J Clin Invest 2013 ¹⁶ ; Lai et al., Sci Trans Med ¹⁷	No	NA	NA
SIRT1	Bhatia et al., Cancer Cell 2012 ¹⁸	No	NA	NA
PRMT5	Jin et al., JCI 2016 ¹⁹	No	NA	NA
Rad52	Cramer-Morales et al., Blood 2013 ²⁰	No	NA	NA
PIM2	Ma et al., PNAS 2019 ²¹	No	NA	NA
BCL2	Goff et al., Cancer Stem Cell 2013 ²²	Venetoclax+DAS	Recruiting	NA
PPRγ	Prost et al., Nature 2015 ²³	Pioglitazone	Potentially active	Rousselot et al., Cancer 2017 ²⁴
Autophagy	Bellodi et al., JCI 2009 ²⁵ ; Baqero et al., Blood 2019 ²⁶	Chloroquine	Modest activity	Horne et al., Leukemia 2020 ²⁷
PML	Ito et al., Nature 2008 ²⁸	Arsenic trioxide	Terminated or not reported	NA
JAK2/STAT3/STAT5	Ye et al., Blood 2006 ²⁹ ; Traer et al., Leukemia 2012 ³⁰ ; Neviani et al., JCI 2013 ¹⁶ ; Gallipoli et al., Blood 2014 ³¹ ; Eiring et al., Leukemia 2015 ³²	Ruxolitinib+TKI	Potentially active	Sweet et al., Leuk Res 2018 ³³
Mitochondrial protein translation	Kuntz et al., Nature Medi 2017 ³⁴	No	No	NA
TGFβ	Naka et al., Nature 2010 ³⁵	No	No	NA
Musashi2	Ito et al., Nature 2010 ³⁶	No	No	NA
ADAR1	Jiang et al., PNAS 2013 ³⁷	No	No	NA
EZH2	Scott et al., Cancer Discov 2016 ³⁸ ; Xie et al., Cancer Discov 2016 ³⁹
PIM2	Ma et al., PNAS 2019 ²¹

DAS , Dasatinib; IM , imatinib; NA , not applicable; NIL , nilotinib.

1 Zhao C, Blum J, Chen A, et al. Loss of beta-catenin impairs the renewal of normal and CML stem cells in vivo. Cancer Cell . 2007;12:528–541.

2 McWeeney SK, Pemberton LC, Loriaux MM, et al. A gene expression signature of CD34+ cells to predict major cytogenetic response in chronic-phase chronic myeloid leukemia patients treated with imatinib. Blood . 2010;115:315–325. doi:10.1182/blood-2009-03-210732.

3 Heidel FH, Bullinger L, Feng Z, et al. Genetic and pharmacologic inhibition of beta-catenin targets imatinib-resistant leukemia stem cells in CML. Cell Stem Cell . 2012;10:412–424. doi:10.1016/j.stem.2012.02.017 .

4 Schurch C, Riether C, Matter MS, Tzankov A, Ochsenbein AF. CD27 signaling on chronic myelogenous leukemia stem cells activates Wnt target genes and promotes disease progression. J Clin Invest . 2012;122:624-638. doi:10.1172/JCI45977.

5 Lim S. Saw TY, Chang, M, et al. Targeting of the MNK-eIF4E axis in blast crisis chronic myeloid leukemia inhibits leukemia stem cell function. Proc Natl Acad Sci U S A . 2013;110:E2298–E2307, doi:10.1073/pnas.1301838110.

6 Zhang B, Li M, McDonald T, et al. Microenvironmental protection of CML stem and progenitor cells from tyrosine kinase inhibitors through N-cadherin and Wnt-beta-catenin signaling. Blood . 2013;121: 1824–1838. doi:10.1182/blood-2012-02-412890.

7 Eiring AM, Khorashad JS, Anderson DJ, et al. β-Catenin is required for intrinsic but not extrinsic BCR-ABL1 kinase-independent resistance to tyrosine kinase inhibitors in chronic myeloid leukemia. Leukemia . 2015;29:2328–2337. doi:10.1038/leu.2015.196.

8 Agarwal P, Zhang B, Ho Y, et al. Enhanced targeting of CML stem and progenitor cells by inhibition of porcupine acyltransferase in combination with TKI. Blood . 2017;129:1008–1020. doi:10.1182/blood-2016-05-714089.

9 Zhang J, Adrián FJ, Jahnke W, et al. Targeting Bcr-Abl by combining allosteric with ATP-binding-site inhibitors. Nature . 2010;463:501–506. doi:10.1038/nature08675.

10 Dierks C, Beigi R, Guo G-R, et al. Expansion of Bcr-Abl-positive leukemic stem cells is dependent on Hedgehog pathway activation. Cancer Cell . 2008;14:238–249.

11 Zhao C, Chen A, Jamieson CH, et al. Hedgehog signalling is essential for maintenance of cancer stem cells in myeloid leukaemia. Nature . 2009;458:776–779

12 Chen Y, Hu Y, Zhang H, Peng C, Li S. Loss of the Alox5 gene impairs leukemia stem cells and prevents chronic myeloid leukemia. Nat Genet . 2009;41:783–792. doi:10.1038/ng.389.

13 Hurtz C, Hatzi K, Cerchietti L, et al. BCL6-mediated repression of p53 is critical for leukemia stem cell survival in chronic myeloid leukemia. J Exp Med . 2011;208:2163–2174. doi:10.1084/jem.20110304.

14 Reavie L, Buckley SM, Loizou, E, et al. Regulation of c-Myc ubiquitination controls chronic myelogenous leukemia initiation and progression. Cancer Cell . 2013;23:362–375. doi:10.1016/j.ccr.2013.01.025 .

15 Abraham SA, Hopcroft LEM, Carrick E, et al. Dual targeting of p53 and c-MYC selectively eliminates leukaemic stem cells. Nature . 2016;534:341–346. doi:10.1038/nature18288.

16 Neviani P, HarbJG, Oaks JJ, et al. PP2A-activating drugs selectively eradicate TKI-resistant chronic myeloid leukemic stem cells. J Clin Invest . 2013;123:4144–4157. doi:10.1172/JCI68951.

17 Lai D, Chen M, Su J, et al. PP2A inhibition sensitizes cancer stem cells to ABL tyrosine kinase inhibitors in BCR-ABL(+) human leukemia. Sci Transl Med . 2018;10:eaan8735. doi:10.1126/scitranslmed.aan8735.

18 Li L, Wang L, Li L, et al. Activation of p53 by SIRT1 inhibition enhances elimination of CML leukemia stem cells in combination with imatinib. Cancer Cell . 2012;21:266–281. doi:10.1016/j.ccr.2011.12.020 .

19 Jin Y, Zhou J, Xu F, et al. Targeting methyltransferase PRMT5 eliminates leukemia stem cells in chronic myelogenous leukemia. J Clin Invest 2016;126:3961–3980. doi:10.1172/jci85239.

20 Cramer-Morales K, Nieborowska-Skorska M, Scheibner K, et al. Personalized synthetic lethality induced by targeting RAD52 in leukemias identified by gene mutation and expression profile. Blood . 2013;122:1293–1304. doi:10.1182/blood-2013-05-501072.

21 Ma L, Pak ML,Ou J, et al. Prosurvival kinase PIM2 is a therapeutic target for eradication of chronic myeloid leukemia stem cells. Proc Natl Acad Sci U S A . 2019;116:10482–10487. doi:10.1073/pnas.1903550116.

22 Goff DJ, Recart AC, Sadarangani A, et al. A Pan-BCL2 inhibitor renders bone-marrow-resident human leukemia stem cells sensitive to tyrosine kinase inhibition. Cell Stem Cell . 2013;12:316–328. doi:10.1016/j.stem.2012.12.011 .

23 Prost S, Relouzat F, Spentchian M, et al. Erosion of the chronic myeloid leukaemia stem cell pool by PPARγ agonists. Nature . 2015;525:380–383. doi:10.1038/nature15248.

24 Rousselot P, Prost S, Guilhot J, et al. Pioglitazone together with imatinib in chronic myeloid leukemia: a proof of concept study. Cancer . 2017;123:1791–1799. doi:10.1002/cncr.30490.

25 Bellodi C, Lidonnici MR, Hamilton A, et al. Targeting autophagy potentiates tyrosine kinase inhibitor-induced cell death in Philadelphia chromosome-positive cells, including primary CML stem cells. J Clin Invest . 2009;119:1109–1123. doi:10.1172/jci35660.

26 Baquero P, Dawson A, Mukhopadhyay A, et al. Targeting quiescent leukemic stem cells using second generation autophagy inhibitors. Leukemia . 2019;33:981–994. doi:10.1038/s41375-018-0252-4.

27 Horne GA, Stobo J, Kelly C, et al. A randomised phase II trial of hydroxychloroquine and imatinib versus imatinib alone for patients with chronic myeloid leukaemia in major cytogenetic response with residual disease. Leukemia . 2020; 34:1775-1786. doi:10.1038/s41375-019-0700-9.

28 Ito K, Bernardi R, Morotti A, et al. PML targeting eradicates quiescent leukaemia-initiating cells. Nature . 2008;453:1072–1078.

29 Ye D, Wolff N, Li L, Zhang S, Ilaria RL Jr. STAT5 signaling is required for the efficient induction and maintenance of CML in mice. Blood . 2006;107:4917–4925.

30 Traer E, MacKenzie R, Snead, J, et al. Blockade of JAK2-mediated extrinsic survival signals restores sensitivity of CML cells to ABL inhibitors. Leukemia . 2012;26:1140–1143. doi:10.1038/leu.2011.325.

31 Gallipoli P, Cook A, Rhodes S, et al. JAK2/STAT5 inhibition by nilotinib with ruxolitinib contributes to the elimination of CML CD34+ cells in vitro and in vivo. Blood . 2014;124:1492–1501. doi:10.1182/blood-2013-12-545640.

32 Eiring AM, Kraft IL, Page BD, et al. STAT3 as a mediator of BCR-ABL1-independent resistance in chronic myeloid leukemia. Leuk Suppl . 2014;3:S5–S6. doi:10.1038/leusup.2014.3.

33 Sweet K, Hazlehurst L, Sahakian E, et al. A phase I clinical trial of ruxolitinib in combination with nilotinib in chronic myeloid leukemia patients with molecular evidence of disease. Leuk Res . 2018;74:89–96. doi:10.1016/j.leukres.2018.10.002.

34 Kuntz EM, Baquero P, Michie AM, et al. Targeting mitochondrial oxidative phosphorylation eradicates therapy-resistant chronic myeloid leukemia stem cells. Nat Med . 2017;23:1234–1240. doi:10.1038/nm.4399.

35 Naka K, Hoshii T, Muraguchi T, et al. TGF-beta-FOXO signalling maintains leukaemia-initiating cells in chronic myeloid leukaemia. Nature . 2010;463:676–680. doi:10.1038/nature08734.

36 Ito T, Kwon HY, Zimdahl B, et al. Regulation of myeloid leukaemia by the cell-fate determinant Musashi. Nature . 2010;466:765–768. doi:10.1038/nature09171.

37 Jiang Q, Crews LA, Barrett CL, et al. ADAR1 promotes malignant progenitor reprogramming in chronic myeloid leukemia. Proc Natl Acad Sci U S A . 2013;110:1041–1046. doi:10.1073/pnas.1213021110.

38 Scott MT, Korfi K, Saffrey P, et al. Epigenetic reprogramming sensitizes CML stem cells to combined EZH2 and tyrosine kinase inhibition. Cancer Discov . 2016;6:1248–1257. doi:10.1158/2159-8290.cd-16-0263 .

39 Xie H, Peng C, Huang J, et al. Chronic myelogenous leukemia- initiating cells require polycomb group protein EZH2. Cancer Discov . 2016;6:1237–1247. doi:10.1158/2159-8290.cd-15-1439 .

Transformation to Accelerated and Blast Phase

As the CP-CML is compatible with life, one could argue that the main objective in CML therapy is the prevention of transformation to BP-CML, an acute leukemia with myeloid (M-BP) or pre-B-cell phenotype (L-BP). A host of genetic alterations has been associated with transformation to accelerated phase (AP)/BP ( Fig. 69.7 ). CCA/Ph ⁺ is present in 70% to 80% of cases and includes a number of nonrandom abnormalities. Classically, +8, +Ph, i(17q), and +19 were considered as “major route” abnormalities, although this is not a universally accepted definition. Several other recurrent abnormalities are also associated with a poor prognosis, including −7 and 3q26 abnormalities. Many other cytogenetic changes have been described in smaller subsets of patients, and their prognostic impact is less well established. However, the acquisition of any CCA/Ph ⁺ in a patient on TKI therapy is consistent with a diagnosis of AP-CML irrespective of other disease features. NGS has revealed point mutations, indels, and fusion genes in addition to BCR-ABL1 (see Fig. 69.7 ). For example, inactivating mutations of RUNX1 occur in ∼25% of M-BP cases, respectively, while ∼55% of L-BP cases have inactivating mutations in IKZF1 .

Figure 69.7, SOMATIC MUTATIONS ASSOCIATED WITH BLAST PHASE CML.

BCR-ABL1 expression tends to increase with progression to BP. In some patients, rising BCR-ABL1 mRNA levels are associated with duplication of Ph, but in the majority this seems to result from transcriptional upregulation due to as yet unknown mechanisms. The increase in BCR-ABL1 expression promotes expression of SET, which inhibits the tumor suppressor phosphatase PP2A by forming a complex with the PP2A catalytic subunit (PP2Ac). PP2A activates SHP-1 to de-phosphorylate BCR-ABL1 and other important substrates. Reduced PP2A activity increases BCR-ABL1 activity, further enhancing the signal in a positive feedback loop. Reduced SHP-1 activity has been associated with resistance to TKIs, emphasizing the importance of this pathway. In addition to SET, cancerous inhibitor of PP2A (CIP2A) was implicated in PP2A inhibition in CML and high CIP2A activity predicts a high risk of transformation to BP.

The most salient feature of transformation from the chronic to the blast phase is the loss of terminal differentiation capacity. Rare patients acquire AML-type translocations such as CBFβ-MYH11 at the time of transformation. Mutant RUNX1 functions in a dominant-negative fashion to block neutrophil differentiation. In the majority of M-BC patients, myeloid differentiation is impaired due to suppression of CAAT enhancer binding protein α (C/EBPα), which is the result of enhanced expression of the RNA-binding protein heterogeneous nuclear ribonucleoprotein E2 (hnRNP-E2), which inhibits C/EBPα translation. Two mechanisms underlie the increased expression and activity of hnRNP-E2. First the increase in BCR-ABL1 activity promotes expression of hnRNP-E2. Second, miR-328, which binds to hnRNP-E2 to block its activity, is down-regulated. Disruption of other myeloid transcription factors such as C/EBPβ and over-expression of MECOM (EVI-1) have also been implicated in blastic transformation. With the loss of differentiation capacity, the organization of CML hematopoiesis is disrupted, as granulocyte-macrophage progenitor cells acquire self-renewal capacity, possibly by activation of nuclear β-catenin, undermining the hierarchical structure of chronic phase hematopoiesis. Various mechanisms have been implicated in β-catenin activation, including stabilization by BCR-ABL1 tyrosine phosphorylation, inactivation of glycogen synthase kinase 3β by mis-splicing, and activation of MNK family kinases. Transformation to L-BC, which has almost invariably a pre-B-cell phenotype, is caused by inactivation of transcription factors that are critical for orderly B-cell development, such as IKZF1 or PAX-5. It is thought that physiologically these transcription factors impose a metabolic restraint on developing B cells that limits the risk of malignant transformation during antigen receptor diversification. Interestingly the types of structural rearrangements in L-BC suggest that they originate from aberrant RAG-mediated recombination.

The variability of somatic mutations associated with blastic transformation is consistent with conversion of multiple aberrancies into a relatively uniform phenotype. Recent work using multiple genome-wide “omic” platforms has implicated epigenetic dysregulation as the overarching mechanism. Specifically downregulation of polycomb repressive complex 2 (PRC2) and upregulation of PRC1 activity may explain phenotypic conversion despite heterogeneous upstream mutations. In this framework, histone and DNA methylation work hand in hand to achieve profound epigenetic reprogramming, consistent with the frequent mutation of epigenetic regulators observed in multiple studies. PRC2 directs BP DNA hypermethylation, which in turn silences key genes involved in myeloid differentiation and tumor suppressor function through epigenetic switching, while PRC1 represses an overlapping and distinct set of genes, including BP tumor suppressors. This may provide a rationale for therapeutic opportunities that are broadly applicable, for instance the combination of EZH2 (PRC2) and BMI1 (PRC1) inhibitors.

Molecular Basis of Tyrosine Kinase Inhibitor Resistance

TKI resistance is classified as primary resistance (failure to achieve a desired response) and acquired resistance (loss of response). Mechanistically TKI resistance broadly bifurcates into BCR-ABL1–dependent resistance (reactivation of BCR-ABL1 kinase activity) and BCR-ABL1–independent resistance (resistance despite continued suppression of BCR-ABL1 kinase activity). Interestingly, for unknown reasons, a given CML case tends to re-use the same strategy to escape the next line of therapy, i.e. to acquire yet another BCR-ABL1 mutation if it has used this strategy before. From a clinical standpoint the distinction between BCR-ABL1–dependent and –independent resistance is critical for rationalizing salvage strategy. (See box on Resistance, Relapse, Recurrence, Intolerance .)

Resistance, Relapse, Recurrence, Intolerance

Resistance implies the failure to achieve a desired level of response (primary resistance) or the loss of response while on therapy (secondary or acquired resistance). Relapse is used synonymously with acquired resistance, while the term recurrence is reserved (and should be used) to denote increasing leukemia burden in the context of treatment-free remission (i.e., off TKI therapy). Keeping the semantics clear is important, given that the biology and prognostic impact of relapse and recurrent are vastly different. In practice it can be challenging to distinguish between true (biological) resistance and resistance due to insufficient dose intensity because of intolerance.

BCR-ABL1–Dependent Tyrosine Kinase Inhibitor Resistance

The most common mechanism of BCR-ABL1 reactivation is mutations in the kinase domain of BCR-ABL1 that impair drug binding through steric hindrance or by stabilizing a kinase conformation from which a given inhibitor is excluded. The spectrum of mutations conferring clinically relevant resistance is broadest for imatinib, reflecting its relatively low potency. Hot spots include the ATP binding loop (p-loop), activation loop, and threonine 315, which controls access to a hydrophobic pocket in the catalytic site and is frequently referred to as the gatekeeper residue ( Fig. 69.8 ). Compared to imatinib, second-generation TKIs exhibit a much more contracted spectrum of resistance mutations, and in aggregate cover all clinical mutations except T315I. This is due to the fact that all first- and second-generation TKIs make a hydrogen bond with T315 and require access to the hydrophobic pocket. In contrast, ponatinib skirts T315I through a rigid triple carbon bond. Kinase domain mutations are more common in acquired resistance than in primary resistance and in AP/BP-CML than in CP-CML. BCR-ABL1 genotyping is part of a workup for clinical resistance, as detection of a mutation allows rationalizing TKI selection. One should bear in mind that the value of genotyping is mostly to avoid selecting an inactive TKI, while the presence of a “sensitive” mutation does not guarantee a clinical response, indicating that other mechanisms, including BCR-ABL1 independent pathways, contribute to clinical resistance. Studies in cell lines have implicated overexpression of BCR-ABL1 by gene amplification or transcriptional upregulation, but the importance of these findings for clinical resistance is less clear. All TKIs are substrates for various efflux pumps, including ABCB1 (MDR1), and several polymorphisms have been correlated with the depth of response. One study showed that overexpression of MDR1 may be sufficient to protect CML stem and progenitor cells, until they acquire a BCR-ABL1 kinase domain mutation that is able to drive overt resistance.

Figure 69.8, TYROSINE KINASE INHIBITORS (TKIS) APPROVED FOR THE TREATMENT OF CHRONIC MYELOID LEUKEMIA AND CLINICAL RESISTANCE MUTATIONS.

BCR-ABL1–Independent Resistance

Extrinsic or intrinsic activation of growth and survival mechanisms can substitute for BCR-ABL1 kinase activity. Multiple different pathways have been implicated, including several that are also downstream effectors of BCR-ABL1, such as PI3K and MAPK. This is consistent with the observation that TKI resistant CML is phenotypically still CML. The mechanistic heterogeneity of BCR-ABL1–independent resistance implies that no obvious universal therapeutic vulnerabilities exist, presenting a challenge for the development of rational therapeutic strategies. There is extensive overlap in the pathways involved in transformation to AP/BP and BCR-ABL1–independent resistance, both of which employ sweeping epigenetic reprogramming, e.g., through imbalanced activity of PRC1 and PRC2. Such common denominator may in the future inform rational strategies to overcome BCR-ABL1–independent resistance. In reality, many cases of advanced TKI-resistant CML combine BCR-ABL1–dependent and –independent mechanisms, which explains why deep responses in these patients are difficult to achieve.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here