Myeloproliferative Neoplasms of Childhood

Rachel Rau

Mignon L. Loh

Myeloproliferative neoplasms (MPNs) are a heterogeneous group of rare clonal hematopoietic stem cell disorders characterized by the abnormal proliferation of one or more of the myeloid lineages.¹ The most common type, chronic myeloid leukemia (CML), accounts for less than 5% of all childhood leukemias, resulting in approximately 100 cases per year in the US pediatric population.² Other MPNs discussed in this chapter include polycythemia vera (PV) and essential thrombocythemia (ET). Additionally, the overlap myelodysplastic/myeloproliferative neoplasms juvenile myelomonocytic leukemia (JMML) (formerly known as juvenile CML), atypical CML, and chronic myelomonocytic leukemia (CMML) will be covered.

CHRONIC MYELOID LEUKEMIA

CML is a myeloproliferative neoplasm that originates in an abnormal pluripotent hematopoietic stem cell.³ It is characterized by myeloid hyperplasia of the bone marrow, extramedullary hematopoiesis, expansion of the total body granulocyte pool, and elevation of the leukocyte count (with appearance of the complete range of granulocyte precursor cells in the peripheral blood). CML is consistently associated with the BCR-ABL1 fusion gene, which in most patients is the product of the t(9;22)(q34;q11) reciprocal translocation that results in the Philadelphia (Ph¹) chromosome.⁴

Historical Background

The cytogenetic hallmark of CML, the Ph¹ chromosome, was described by Nowell and Hungerford in 1960.⁴ This was the first specific chromosomal abnormality associated with a human malignancy, and its discovery inaugurated the era of cancer cytogenetics. The Ph¹ chromosome was initially thought to be a truncated chromosome 22 resulting from deletion of genetic material; however, in 1973, Janet Rowley demonstrated a reciprocal translocation between chromosomes 9 and 22.⁵ Over the subsequent 25 years, the pathogenesis of CML was defined by identifying the resultant fusion gene (BCR-ABL1) and its protein product, a constitutively activated tyrosine kinase (TK).

Epidemiology

CML is primarily a disease of middle age; the peak incidence is in the fourth and fifth decades. It accounts for approximately 3% of cases of newly diagnosed pediatric leukemia, with an annual incidence of approximately 1 to 2 per million in individuals less than 20 years old.² (Table 13.13, SEER 1975-2011). The diagnosis of CML is particularly rare in young children with most of the pediatric patients diagnosed after age 4 years.⁶^,⁷ No significant racial or gender bias exists, and no hereditary component is demonstrable. The only environmental factor clearly implicated in the etiology of CML is ionizing radiation. An increased incidence of adult CML has been reported in radiologists, in survivors of atomic bomb explosions, and in persons exposed to therapeutic radiation for treatment of ankylosing spondylitis and other disorders. In these patients, there is a considerable lag period following exposure before clinical evidence of CML is detected. On the basis of atomic bomb data, an average of 8 years is required between the original mutational event and the development of clinical symptoms.

Molecular Biology

The cytogenetic hallmark of CML is the Ph¹ chromosome, resulting from the reciprocal translocation t(9;22)(q34;q11) (Fig. 21.1). The molecular consequence of this event is the production of a chimeric protein (BCR-ABL1) that results in constitutive kinase activity. Breakpoints on chromosome 9 involve the ABL gene and can vary widely (i.e., >100 kilobases [kb] from case to case); on the other hand, breakpoints on chromosome 22 are virtually always restricted to a small (5.8-kb) segment of DNA known as the major breakpoint cluster region (M-BCR) (Fig. 21.2). As a result of the reciprocal translocation, two hybrid genes are formed: BCR-ABL1 on 22q- and ABL1-BCR on 9q+. Although both of these genes are transcribed, BCR-ABL1 appears to have the major role in the pathogenesis of CML.

The development of the hybrid BCR-ABL1 gene may be a more common phenomenon than is commonly recognized. When studied using a very sensitive reverse transcriptase-polymerase chain reaction (RT-PCR) screening technique, 20% of “normal” adult subjects were found to harbor this translocation in a minor subpopulation of cells.⁸ The age distribution of these BCR-ABL1-positive, but otherwise normal, individuals roughly mirrors the age distribution of CML. The mechanisms preventing these individuals from developing CML are unknown.

Properties of ABL1 and Its Protein Product (ABL1)

The wild-type ABL1 gene encodes a 145-Da protein that shuttles between the cytoplasm and the nucleus and is universally active
in hematopoietic cells at all stages of differentiation. It can localize to the mitochondria during oxidative stress. It is the human homolog of the Abelson B-cell murine leukemia virus oncogene (v-abl). The ABL1 protein participates in signal transduction and regulation of gene transcription. One major function is to catalyze the attachment of phosphate groups to the tyrosine residues of various proteins, that is, to act as a TK. Members of the TK family are frequently involved in the pathways that transmit signals from the external milieu to the cytoplasm and nucleus; in this capacity, they may act as growth factors, transmembrane receptors, or submembrane catalytic subunits of surface receptors.

Figure 21.1 Anatomy of the (9;22)(q34;q11) translocation with formation of the Philadelphia (Ph¹) chromosome (22q-) containing the hybrid BCR-ABL1 gene. (See text for further explanation.) CML, chronic myelocytic leukemia.

Figure 21.2 The normal abl tyrosine kinase (TK) component of P145 is carefully regulated by the SH2 and SH3 domains; genetic alterations that prevent interaction of the SH3 domain with proline-rich motifs within the C terminus result in derepression (constitutive activation) of the TK. In Philadelphia chromosome-positive chronic myelocytic leukemia, fusion with the bcr gene leads to production of a chimeric protein (P210) with activated TK and oncogenic properties. Two domains within bcr are required for its oncogenic effect: domain I mediates oligomerization of bcr/abl and promotes phosphorylation of tyrosine residue 177 within domain II, which, in turn, binds to a signaling adaptor molecule coupling bcr/abl with the ras signaling pathway. (See text for further explanation.)

The ABL1 protein contains other functional domains in addition to its TK activity (Fig. 21.2; Table 21.1). The carboxy-terminal portion contains a domain involved in binding to F-actin as well as a separate domain that can bind to DNA; the DNA-binding activity appears to be cell cycle-regulated by the cyclin-activated cdc-2 kinase. There are also three nuclear localization signals and one nuclear export signal. At the N-terminal region are three domains known as src-homology (SH) regions because of their kinship to the viral src oncogene. The first domain (SH1) normally manifests weak TK activity and appears to be tightly regulated by the SH2 and SH3 regions. The presence of SH2 and SH3 regions is significant because these domains play critical roles in intermolecular interactions that specifically mediate protein-protein coupling. The SH2 region can bind to substrates that are tyrosine phosphorylated, whereas the SH3 domain complexes with proline-rich regions involved in coordinating cytoskeletal interactions. The proline-rich SH3 motifs can function as docking sites for the SH3 domains of “adaptor” proteins such as CRK, GRB2 (growth factor receptor-bound protein 2), and NCK. Proteins containing SH2 and SH3 domains are classified as adaptor molecules because they couple nonreceptor TKs to downstream signaling cascades regulating gene expression.

TABLE 21.1 Functional Domains of ABL, BCR, and BCR-ABL1 Proteins

Protein	Chromosome	Domain		Function
p145^ABL	9q34	N terminal
			SH1	Tyrosine kinase
			SH2	Interacts with tyrosine-phosphorylated proteins
			SH3	Suppression of tyrosine kinase activity
			Myristoylation site	Localization of p145^ABL1 to nucleus
		C terminal		DNA binding, nuclear localization, actin binding
p160^BCR	22q11	N terminal
			Coiled-coil motif	Polymerization with other proteins
			Catalytic domain	Serine-threonine kinase
		Central
			GEF	GDP-GTP exchange factor
		C terminal
			GAP	GTPase-activating protein for RAC and RHO (RAS-related proteins)
p210^BCR/ABL	t(9;22)(q34;q11)	BCR
	(Ph¹ chromosome)		Coiled-coil motif	Increases tyrosine kinase activity of ABL1
			Tyrosine residue 177	Enables binding of F-actin by ABL1
			Serine-threonine kinase	Docking site for adaptor proteins
		ABL		Activation of signal-transduction proteins
			SH1 domain	Phosphorylation of signal and adaptor proteins
			Actin-binding domain	Cytoplasmic location. Interference with adhesion
GAP, guanosine triphosphatase-activating protein; GDP, guanosine diphosphate; GEF, guanine nucleotide exchange factor; GTP, guanosine triphosphatase.

The N terminus of ABL1 consists of a “Cap” region, which is covalently linked to a C₁₄ myristoyl moiety (Fig. 21.2). This myristoyl modification plays a critical role in controlling the activity of the
TK (SH1) domain. By engaging the C-terminal lobe of SH1, it facilitates the docking of the SH2 and SH3 domains onto it, thereby blocking the access of ATP and the peptide substrate to the active site and inactivating the TK activity. In contrast, forms of ABL1 that lack myristoylation (or replace the myristoylated Cap with unmyristoylated BCR) manifest constitutive TK activity. (reviewed in Ref. 9)

Properties of BCR

The BCR region is a component of a much larger gene known as BCR, which encodes a 160-kDa protein. BCR is composed of several regions (Fig. 21.2; Table 21.1) including (a) an N-terminal sequence that contains a coiled-coil oligerimization domain and encodes a serine/threonine kinase; (b) a central portion, which encodes a Rho-guanine nucleotide exchange factor (GEF) that allows the exchange of guanosine triphosphatase (GTP) for guanosine diphosphatase (GDP) and may also activate NF-k B; and (c) a C-terminal portion that codes for a guanosine triphosphatase (GTPase)-activating protein (GAP). A particularly important site is the tyrosine moiety located at position 177 (Tyr177) that can serve as a docking site for GRB2, GRB10, 14-3-3, and the ABL1 protein.⁹

Properties of BCR/ABL1 and Its Protein Product (BCR-ABL1)

The BCR/ABL1 gene encodes the tumor-specific BCR-ABL1 molecule, a 210-kDa hybrid protein (P210) that differs from the normal ABL1 kinase in several respects: (a) it has augmented and constitutive TK activity; (b) it has the ability to autophosphorylate; (c) it is translocated to the cytoplasm, thereby exposing it to a new spectrum of substrates of remarkable diversity; and (d) it binds to F-actin. Similar modifications of other ABL proteins have been demonstrated to mediate viral oncogenesis (e.g., v-abl, feline sarcoma virus) and to confer growth-factor independence on various cell lines.¹⁰ In each of these instances, the critical genetic alteration involves a substitution at the N-terminal end of the ABL gene (Fig. 21.3).

As shown in Figure 21.4, the BCR-ABL1 protein is capable of activating multiple downstream signaling pathways. Phosphorylated Tyr177 plays a central role in leukemogenesis by (a) serving as a docking site for GRB2, which, in turn, activates multiple pathways, including those that lead to Ras/MAPK activation and increased transcription of MYC and BCL-X genes; (b) blocking transcription of interferon consensus sequence-binding protein (ICSBP), thereby preventing its inhibition of transcription of the antiapoptotic BCL-X and BCL-2; (c) activating the RAC subfamily of GTPases that, in turn, activate downstream signaling molecules such as CRKL, ERK, c-Jun N-terminal kinase (JNK), and p38.¹¹ The consequence of these events is promotion of cell proliferation and survival.

Alternate Splicing Patterns for BCR-ABL1

In the genesis of BCR-ABL1 (P210), the break in ABL1 typically occurs in the first intron between alternate exons 1a and 1b; the break in BCR occurs most often within a 5.8-kb area encompassing exons e12-e16 (Fig. 21.2). There are also at least two other BCR-ABL fusion proteins with molecular weights of 190 kDa and 230 kDa (P190 and P230, respectively) that have been associated with neoplastic phenotypes. Since these oncoproteins share the same ABL TK sequences, their different clinical features presumably reflect modulation by the unique protein domains contributed by the various BCR breakpoints (Table 21.2).

Figure 21.3 Schematic representation of the normal protein product (P145) of the ABL1 gene and the modifications associated with leukemogenesis. Note that all variants retain the tyrosine kinase domain (blue area) and are altered at the 58 end by insertion of the viral gag component in the case of Abelson murine leukemia virus (AbMuLV) and feline sarcoma virus (FeSV) or with BCR segments in the case of Philadelphia chromosome-positive (Ph¹-positive) ALL and chronic myelocytic leukemia (CML).

P210 has been associated with some forms of Ph¹-positive acute lymphoblastic leukemia (Ph⁺ ALL). Over 40% of adults with Ph⁺ ALL carry the P210 BCR-ABL1 fusion typical of CML.¹² In the remainder, and in nearly 80% of children with Ph⁺ ALL, a different rearrangement occurs, with ABL exon 2 spliced to an exon outside the major BCR region (known as the minor BCR region or m-BCR).¹³ The resultant 190 kDa protein product (P190) has approximately a five-fold higher level of TK activity than does P210. This characteristic appears to correlate with a higher transforming ability of P190. Careful analysis has shown that the p190 transcript, traditionally associated with Ph¹-positive acute leukemia, can be detected at very low levels (corresponding to approximately 0.02% of the total BCR-ABL1 transcripts) at diagnosis in virtually all CML patients, apparently arising as a consequence of alternative or missplicing events in the BCR gene.¹⁴ A third BCRABL fusion protein (P230) was initially reported as the molecular basis for some cases of chronic neutrophilic leukemia, but now is considered a rare neutrophilic variant of CML. This translocation involves a third breakpoint cluster region in BCR (µ-bcr).¹⁵

Although the classic t(9;22) is found in approximately 90% of patients with CML, approximately 5% to 10% do not manifest the Ph¹ chromosome. In some of these patients, the Ph¹ chromosome may be masked by translocation of additional genetic material to the 22q11 region.¹⁶ Other patients may have rearrangements or breaks in the 9q34 region without the reciprocal break at 22q11.¹⁷ Molecular biology techniques permit identification of BCR rearrangements in Ph¹-negative patients who have otherwise typical CML.¹⁷^,¹⁸ In such cases, ABL1 and BCR are presumed to have been juxtaposed on the molecular level by a mechanism other than that producing the typical t(9;22); this phenomenon may result from an interstitial insertion of ABL1 into BCR or a complex translocation with the BCR-ABL1 fusion gene located on another chromosome.

Approximately 10% of patients in chronic phase (CP) have, in addition to the Ph¹ chromosome, other visible karyotypic abnormalities such as a second Ph¹ chromosome, lack of the Y chromosome, isochromosome 17, or an extra chromosome 8 or 19; these secondary changes are associated with a worse prognosis, and appear to represent a mechanism of tumor progression.¹⁹

Philadelphia Chromosome-Negative Myeloproliferative Neoplasms Clinically Resembling CML

Patients with a CML-like clinical picture but lacking both Ph¹ and BCR-ABL1 rearrangement have also been described. Such
patients were previously all classified as Philadelphia chromosome-negative CML; however now this heterogeneous population of patients can be better categorized as distinct Ph¹-negative myeloproliferative neoplasms based on clinical characteristics and genetic features.¹

Figure 21.4 Molecular signaling in BCR-ABL1-positive myeloid precursors. The phosphorylated Tyr177 residue of BCR serves as a docking site for growth factor receptor-bound protein 2 (GRB2), which binds GRB2-associated binding protein 2 (GAB2), as well as SOS (a guanine nucleotide exchanger of RAS), resulting in RAS-MAPK activation; this, in turn, promotes transcription of BCL-2 gene. Upon phosphorylation, GAB2 recruits phosphatidylinositol 3-kinase (PI3K), which activates AKT. AKT activation increases transcription of MYC gene and stabilizes MYC protein via inhibition of its degradation by GSK-3β. BCR-ABL-1 also activates STAT5, both directly and indirectly through activation of JAK2 and the SRC kinases HCK and LYN. The end result is the activation of BCL-X gene transcription. In addition, BCR-ABL1 abrogates the transcription of interferon consensus sequence-binding protein (ICSBP), thereby releasing the ISCBP-mediated inhibition of BCL-2 and BCL-X gene transcription, resulting in increased survival of myeloid precursors. Thus, the net effect of BCR-ABL1 kinase activation is the promotion of cell proliferation and survival. Pointed arrows indicate direct interactions and/or activation. Blunt-ended arrows indicate inhibitory effects. GSK-3β, glycogen synthase kinase-3β; PIP₃, phosphatidylinositol-3,4,5 triphosphate. (From Quintos-Cardama A, Cortes J. Molecular biology of bcr-abl-positive chronic myeloid leukemia. Blood 2009;113:1619, used with permission.)

TABLE 21.2 Relationship of BCR Domains to BCR/ABL1 Fusion Genes

Fusion Gene	BCR Domains		Clinical Phenotype
p190 (or p185)	Dimerization		Ph¹-positive acute leukemia
	Binding
	SH2
	Serine/threonine kinase
p210	All of the above plus:		Ph¹-positive chronic myelocytic leukemia
		Dbl-like
		Pleckstrin homology
p230	All of the above plus:		Neutrophilic variant of CML
	Calcium-phospholipid binding 1/3 of GTPase-activating gene for p21 rac (racGAP)

Such patients are occasionally found to have rearrangements involving the receptor tyrosine kinases, platelet derived growth factor receptor (PDGFR) b or fibroblast growth factor receptor 1 (FGFR1) through translocations t(5;12)(q33;p13) and t(8;13)(p11,q12) respectively.²⁰^,²¹ Eosinophilia is prominent in these disorders, and such patients are now categorized as myeloid and lymphoid neoplasms with eosinophilia and abnormalities of PDGFR or FGFR1. Imatinib, which inhibits PDGFR, has resulted in durable responses in patients with chronic myeloproliferative neoplasms with PDGFR rearrangements.²²

Atypical CML, BCR-ABL1 negative (aCML) is a rare mixed myelodysplastic/myeloproliferative syndrome with an estimated incidence of 1 to 2 cases for every 100 patients with BCR-ABL1-positive CML. While clinically such patients resemble CML, they lack both Ph¹ and BCR-ABL1 rearrangement as well as have no rearrangement of PDGFRA, PDGRFB, or FGFR1. Other diagnostic criteria include peripheral blood leukocyte count ≥13×10⁹/L, neutrophil precursors (promyelocytes, myelocytes, metamyelocytes) ≥10% of peripheral blood leukocytes, minimal (usually <2%)
basophils, minimal absolute monocytosis (<10%), hypercellular bone marrow with granulocytic proliferation and granulocytic dysplasia and <20% blasts in the blood and marrow. Chromosomal changes have been reported in 20% to 88% of aCML cases, most commonly, trisomy 8 and del (20q). Recently, much has been learned about the molecular basis of this rare disease through next generation sequencing efforts. Non-mutually exclusive mutations in the receptor for colony-stimulating factor 3 (CSF3R), the Set binding protein, SETBP1, and JAK2^V617F have been identified in 40%, 25%, and 10% of aCML patients, respectively. The identification of such mutations may be useful as a diagnostic criterion for aCML,²³^,²⁴ and may have therapeutic implications. CSF3R mutations signal through JAK or SRC kinase pathways and thus JAK inhibitors such as ruxolitinib or SRC kinase inhibitors such as dasatinib may prove therapeutically useful.²⁵

Also in the differential diagnosis for patients lacking BCR-ABL1 rearrangements with clinical features resembling CML is CMML, and juvenile myelomonocytic leukemia (JMML). These mixed myelodysplastic/myeloproliferative disorders are additionally characterized by an absolute monocytosis, and are discussed later in this chapter.

Philadelphia Chromosome-Positive Acute Leukemia

Although characteristic of CML, the Ph¹ chromosome is not exclusive to this disease. This chromosomal abnormality is also found in approximately 3% to 5% of childhood acute leukemias, in 2% to 3% of adult acute myeloid leukemias, and in 25% to 33% of adult acute lymphoid leukemias. Ph¹-positive acute leukemias have no antecedent CML features and are clinically and hematologically indistinguishable from other acute leukemias except for a relatively poorer prognosis. A detailed discussion of Ph¹-positive ALL is found in chapter 19.

Biology of CML

The cytogenetic changes in the CML precursor cell are expressed in its progeny by a variety of cytologic alterations that, in turn, produce the neoplastic phenotype. These biologic features are discussed in the subsequent sections.

Clonality

Independent lines of evidence derived from cytogenetic data and analysis of isoenzyme patterns indicate that CML is an acquired disorder of unicellular origin, and the target of neoplastic transformation is a multilineage stem cell with the potential for generating all the myeloid lineage cells (erythrocytes, neutrophils, basophils, eosinophils, monocytes, megakaryocytes) and, in at least some instances, the lymphoid lineages.³ This multilineage potential accounts for the cytologic heterogeneity of the blastic phase of CML (discussed later in the “Natural History” section).

The Initial Transformation Event

Some suggestions that the BCR-ABL1 translocation may not be the primary event in the neoplastic sequence have been raised, but there is very little evidence of any acquired molecular aberration preceding the t(9;22) translocation. Based on the bulk of available data, generation of the BCR-ABL1 fusion gene in a pluripotent hematopoietic stem cell is sufficient to initiate the expansion of a hematopoietic clone that leads to CML. The idea that acquisition of the BCR-ABL1 fusion gene is the first step in the genesis of CML is supported by murine models in which mice transplanted with cells transfected with a BCR-ABL1 gene have developed a disease that closely recapitulates human CML.²⁶

Mechanisms Underlying Growth Advantage of CML Cells

The expansion of the CML clone from a single transformed cell to predominance in bone marrow and blood is dramatic evidence of its ability to overgrow the normal hematopoietic elements. The mechanisms by which it achieves preeminence have been of great interest to students studying the disease. Some clues have been provided by analysis of proliferative kinetics and in vitro colony production.

Cytokinetics. In CP CML, the myeloid mass is greatly increased owing to expansion of mature elements and to increased numbers of progenitor and stem cells.²⁷ Multiple factors likely contribute to this massive expansion, including prolonged cell survival and increased production. CML cells have a significantly longer half-life compared with their normal counterparts, likely secondary to resistance to apoptosis.²⁸ The BCR-ABL1 oncoprotein protects cells from apoptosis via several different mechanisms, among which are inhibition of upstream preapoptotic mitochondrial events (e.g., release of cytochrome C or altered fas/fas ligand interaction) and downstream inhibition of caspase 3.²⁹ Many of these effects are mediated by influencing the relative expression levels of apoptosis inhibitors (BCL-2, BCL-X_L) and promoters (BAX, BAD, BCL-X_S) and/or the subcellular localization of these factors. Signal transducers and activators of transcription (STAT) proteins appear to be constitutively phosphorylated by BCR-ABL1. Activation of STAT1 and STAT5 upregulates BCL-X_a and contributes to cellular proliferation and survival.³⁰^,³¹

Also contributing to the expansion of the myeloid mass is the hyperproduction of myeloid cells.²⁷^,²⁸ An overproduction of committed myeloid precursor cells can be demonstrated by assays of the granulocyte-monocyte colony-forming unit (CFU-GM) from the bone marrow and blood of CML-CP patients.²⁸^,³² Colony growth is qualitatively normal (in CP); mature GM forms are produced, and the process depends on the same CSFs that are obligatory for in vitro colony formation by normal GM-CFU. The basic defect appears to be discordant maturation with preferential expansion of the progenitor compartment.²⁷ This preferential expansion may relate in part to constitutive expression by leukemic progenitors of growth-stimulating factors, notably interleukin-3 and granulocyte colony-stimulating factor (G-CSF).³³

Abnormalities in Feedback Regulation. Under normal circumstances, myelopoiesis is regulated by at least three negative feedback molecular species: lactoferrin (LF), prostaglandin E (PGE), and acidic isoferritins (AIF). LF, the product of mature polymorphonuclear leukocytes (PMNs), downregulates granulocytopoiesis by reducing monocyte-macrophage production of CSF.³⁴ PGE and AIF, which are derived from subpopulations of monocytes and macrophages, inhibit proliferation of normal granulocyte-monocyte precursor cells expressing HLA-DR (Ia) antigens.³⁴^,³⁵^,³⁶

CML cells appear to be relatively insensitive to feedback inhibition by virtue of deficient LF production by the PMN, decreased responsiveness of the monocyte-macrophage to regulation by LF, and decreased sensitivity of progenitor cells to PGE and AIF.³⁷ The PGE and AIF resistance may reflect deficient HLA-DR antigen expression by CML cells or a decreased proportion of sensitive target cells. CML cells may augment their proliferative advantage by releasing humoral factors such as neutrophil elastase, proproteinase C (also known as leukemia-associated inhibitor, or LAI), and cathepsin G to which they are relatively resistant but which suppress normal hematopoietic progenitors. These enzymes are located in the azurophil granules of neutrophils and provide feedback regulation by enzymatically degrading hematopoietic growth factors and/or their receptors.³⁸^,³⁹ The leukocytosis of CML promotes high levels of neutrophil elastase and high levels of other serine proteases (e.g., LAI) that can digest G-CSF and downregulate Ph¹-negative granulocytopoiesis.⁴⁰^,⁴¹

Altered Adhesive Interactions. There are data to suggest that BCR-ABL-induced integrin dysfunction leads to defective binding of CML progenitors to marrow stroma.⁴² Additionally, BCR-ABL^P210 has been shown to downregulate the expression of several genes implicated in cell adhesion and migration, such as L-selectin, intercellular adhesion molecule-1 and the chemokine receptor
CCR7.⁴³ This defective adhesion to the bone marrow stroma may lead to release of immature cells into the circulation and may facilitate hematopoiesis in extramedullary sites.

TABLE 21.3 Criteria for Chronic, Accelerated, and Blast Phases of CML

Chronic Phase	Accelerated Phase	Blast Phase
Must meet all the following:	Must meet ≥1 of the following:	Must meet ≥1 of the following:
Documented t(9;22) or BCR-ABL fusion gene	Blasts in PB or BM 10%-19%	Blasts in PB or BM ≥20%
BM blasts <10%	Basophils in PB ≥20%	Extramedullary blast proliferation (apart from spleen)
Does not meet any criteria for accelerated phase or blast crisis	Persistent thrombocytopenia (<100 × 10⁹/L) unrelated to therapy Cytogenetic evidence of clonal evolution Increased spleen size and increasing WBC unresponsive to therapy	Large foci or clusters of blasts in BM biopsy

Abnormal Angiogenesis. CML patients manifest abnormal bone marrow vessel production (angiogenesis). Bone marrow vascularity and vascular endothelial growth factor (VEGF) levels are significantly increased. Overexpression of BCR-ABL1 upregulates VEGF, leading to phosphorylation of endothelial VEGF regulator, VEGF-R2/KDR, and PKB. Akt, a serine-threonine kinase regulator of endothelial cell activation.⁴⁴^,⁴⁵

Natural History

The natural history of CML is divided into chronic, accelerated, and blast phases (for definitions of each phase, see Table 21.3). These phases represent the progressive shift in the nature of the disorder from one of hyperproliferation, with production of mainly mature hemic elements, to one characterized by differentiation arrest, with hyperproduction of predominantly immature (blast) cells characterized by reversion to stem cell phenotype, block of apoptosis, and therapeutic resistance.

Chronic Phase

CP is characterized by marked expansion of the hematopoietic pools; morphologically mature blood cells are produced that show only subtle functional abnormalities. In general, the neoplastic cells are restricted to the bone marrow, liver, spleen, and peripheral blood. Therefore, symptoms are related to organ infiltration, hyperviscosity, and the metabolic consequences of hyperproliferation, all of which are relatively easy to control. Similar to adults, approximately 95% of children with CML will present in CP.⁷

Figure 21.5 Peripheral blood smears of chronic myelocytic leukemia. A: Chronic phase—marked leukocytosis showing the entire range of myeloid cells from myeloblast to mature polymorphonuclear leukocytes; a hypergranular eosinophil and basophil are present as well. B: Blast phase—blast cells are now more prominent, and there is a hiatus in myeloid maturation. (Courtesy of Drs. Hiroko Shinoda and William Rezuke.)

Symptoms. Patients usually present with nonspecific complaints, such as fever, night sweats, weakness, left upper quadrant pain or fullness, and bone pain. Neurologic dysfunction, respiratory distress, visual difficulties, or priapism may complicate cases characterized by marked hyperleukocytosis.

Physical Findings. The usual physical findings in CP are pallor, low-grade fever, ecchymoses, hepatosplenomegaly, and sternal tenderness. Signs relating to leukostasis (neurologic abnormalities, papilledema, retinal hemorrhages, and tachypnea) are seen in patients with extreme hyperleukocytosis.

Laboratory Findings. A mild normochromic, normocytic anemia, marked leukocytosis with shift to the left and thrombocytosis are common laboratory findings. The median hemoglobin at presentation in children (11.1 g/dL) is significantly less than that seen in adults.⁶^,⁷ The leukocyte count at diagnosis ranges from approximately 8,000 to 800,000/mm³; the median count in children (approximately 250,000/mm³) is higher than that seen in adults.⁶^,⁷ Extreme hyperleukocytosis (>500,000/mm³) is also more common in children. The peripheral blood smear shows myeloid cells at all stages of differentiation; myeloblasts and promyelocytes generally comprise less than 15% of the differential count (Fig. 21.5A). An absolute increase in the numbers of basophils
and eosinophils is noted. Hybrid eosinophilic-basophilic granulocytes may also be seen⁴⁶; because similar chimeric granules may be found in normal immature granulocytes, this phenomenon may reflect incomplete maturation. The mean platelet count in children is approximately 500,000/mm³, which is not significantly higher than that in adults.⁶^,⁷ Serologic findings include elevation of uric acid, lactate dehydrogenase, vitamin B₁₂, and vitamin B₁₂-binding protein (transcobalamin 1).

The bone marrow is hypercellular, mainly reflecting granulocytic (and often megakaryocytic) hyperplasia; orderly granulocyte maturation, eosinophilia, and basophilia are present. There are normal to increased numbers of megakaryocytes found in clusters. These megakaryocytes are often small and hypolobulated (micromegakaryocytes). The bone marrow and spleen occasionally contain lipid-laden histiocytes that resemble Gaucher cells or sea-blue histiocytes.

Differential Diagnosis. The differential diagnosis of CP CML includes leukemoid reaction, juvenile myelomonocytic leukemia (JMML), and other myeloproliferative disorders.⁴⁷

In leukemoid reactions, splenomegaly is usually not marked, the Ph¹ chromosome is absent, and an inflammatory focus is often demonstrable. In JMML, the Ph¹ chromosome is absent; leukocytosis and splenomegaly are less marked than in CML, and involvement of skin, lymphoid tissue, and the monocytic lineage is more pronounced. Other molecular abnormalities are found in JMML and are part of routine diagnostic assays. CML can be distinguished from other myeloproliferative neoplasms by the disproportionate involvement of the granulocyte series and the presence of the Ph¹ chromosome and/or BCR-ABL1 transcripts.

Accelerated Phase

Progression to a more aggressive phase generally proceeds as a gradual multistep evolution. About 50% of patients develop a progressive maturation defect, resulting in a hematologic picture similar to that of de novo acute leukemia; the remaining 45% have the gradual evolution of a myeloproliferative syndrome. Sudden onset of blastic phase (defined as onset within 3 months from a previously documented complete hematologic response [CHR]) is an unusual event within the first 3 years following diagnosis: the incidence is 0.4% in the first year, 1.8% in the second year, and 2.6% in the third year.⁴⁸ Most cases of sudden blastic crisis have lymphoblastic morphology.

The onset of accelerated phase (AP) is characterized by progressive systemic symptoms (fever, night sweats, weight loss), increasing leukocyte counts with a high proportion of immature cells, basophilia, and increasing resistance to chemotherapy. Along with these features is evidence of karyotypic evolution. Mutations in the tumor suppressor gene, TP53 may play a significant role in transformation: p53 mutations are detectable in late CP and may indicate increasing genomic instability and early progression to blast transformation. New karyotypic abnormalities (most commonly, duplication of the Ph¹ chromosome, isochromosome 17, or trisomy 8) appear in accelerated disease. Occasionally, the first manifestation of metamorphosis is extramedullary (i.e., meningeal leukemia or a chloroma arising in soft tissue or bone) disease; such findings usually herald the imminent blast transformation of the marrow.

Blast Phase/Blast Crisis

Blast crisis (BC) is characterized by loss of the leukemic clone’s capacity to differentiate. As a consequence, the clinical picture resembles that of an acute leukemia, with anemia, thrombocytopenia, and increased numbers of blast cells in both the peripheral blood (Fig. 21.5B) and the bone marrow. A marrow blast percentage of 20% or more is diagnostic of blast phase. The signs and symptoms are those of a de novo acute leukemia; if basophilia is extreme, the patient may also have hyperhistaminemic symptoms (pruritus, cold urticaria, gastric ulceration).

Cytogenetic Heterogeneity of Blast Phase. As a reflection of the pluripotent nature of the leukemic stem cells in CML, BC may involve any of the lymphohematopoietic lineages. In approximately 60% to 70% of cases, the blast cell morphology is myeloblastic; unlike de novo acute myeloid leukemia, however, the blast cells are usually peroxidase negative and rarely have Auer rods. Approximately one-third of patients have blast cells with lymphoid morphology. These cells generally express a phenotype corresponding to an early B cell. In rare cases, blast cells may express T-lineage markers; these patients usually have marked extramedullary involvement (especially in lymph nodes) and frequently lack a preceding CP. In some patients, the blast cells manifest features of more than one myeloid line or have mixed myeloid-lymphoid features.

BC is frequently associated with new specific molecular and cytogenetic abnormalities. At the molecular level, the most common gene mutations involve the TP53 gene (mutated in 25% to 30% of myeloid conversions) and the INK4A/ARF exon 2 (homozygously deleted in approximately 50% of lymphoid conversions)⁴⁹; other genes potentially involved in phenotypic progression include IKZF1, JUNB, FOS, PRAME, MXF1, EF1δ, and WNT/β-catenin.⁵⁰ These changes may serve to promote genomic instability and or dysregulation of cell cycle progression. The most commonly identifiable karyotypic alterations are duplication of the Ph¹ chromosome (⁺Ph¹), trisomy 8 (+8), trisomy 19 (+19), and isochromosome 17q (i17q). A rare consistent chromosomal abnormality, a reciprocal t(3;21)(q26;q22), has been described in patients with CML either before or at the onset of blast transformation.⁵¹^,⁵² In general, these secondary cytogenetic changes are nonspecific and may indicate a generalized genomic instability rather than directly relate to blastic progression; however, other changes clearly mirror the karyotypic features associated with specific subtypes of de novo acute leukemia and may be directly related to blastic transformation.⁵³ The phenotypic features of blast transformation may be determined by these specific secondary genetic events; for example, rearrangements involving chromosome regions containing Ig genes and T-cell receptor genes have been associated with B-lymphoid and T-lymphoid acute transformations, whereas involvement of region 3q21 may be accompanied by dysmegakaryopoiesis and conversion to acute megakaryoblastic leukemia.⁵¹

Differential Diagnosis. Because most instances of BC occur after a well-documented CP, the diagnosis is usually clear-cut. The rare patient who presents in BC without a recognized preceding CP may pose diagnostic difficulty, however. The combination of marked splenomegaly, basophilia, and the Ph¹ chromosome (P210 with BC CML, P190 with Ph⁺ ALL) distinguishes BC from most types of de novo acute leukemia.

Prognostic Considerations

Recent advances in the management of CML have improved median survival significantly. Currently, the most important prognostic factor for patients in CP is response to tyrosine kinase inhibitor (TKI) therapy (see discussion later in this chapter). For adults, factors at diagnosis such as age, spleen size, and laboratory parameters are used to categorize patients as high, intermediate or low risk by one of the commonly used prognostic systems.⁵⁴^,⁵⁵ Prior to the routine use of TKI, such factors predicted the risk of early transformation, but have still proven useful predictors of response to TKI therapy and outcome. The role of these factors in the prognosis of pediatric patients with CML is less clear; in one study, only peripheral blood and marrow blast counts at presentation were of prognostic significance.⁶

Once patients have entered BC, the only parameters that correlate with survival are blast cell phenotype and cytogenetic findings. In general, lymphoblastic phenotype and minimal karyotypic evolution augur a more favorable response to therapy.

Therapy

Special Management Problems

Metabolic Disorders. Metabolic consequences of rapid cytolysis (e.g., hyperuricemia, hyperkalemia, hyperphosphatemia) should be anticipated and treated appropriately with hydration, alkalinization, and allopurinol. A detailed discussion of metabolic management is found in Chapter 38.

Hyperleukocytosis. The extremely high leukocyte count associated with some cases of CML can cause leukostatic complications in several organs, especially brain, lung, retina, and penis. Because leukocytes are less deformable than erythrocytes, the viscosity of the blood increases dramatically as the fractional volume of leukocytes (leukocrit) increases. Myeloblasts, which are larger and more rigid than other leukocytes, contribute disproportionately to viscosity; thus, the patient with myeloblastic transformation is at particularly high risk. If hyperleukocytosis is symptomatic or extreme (leukocytes >200,000/mm³ or blast count >50,000/mm³), it should be treated with the simultaneous use of cytotoxic drugs (e.g., hydroxyurea, 50 to 75 mg/kg/d by intravenous infusion) and leukapheresis (or exchange transfusion); erythrocyte transfusions (which increase blood viscosity) should be avoided if possible until the leukocyte is reduced to a safe level.

Thrombocytosis. Thrombocytosis may be associated with thromboembolic or hemorrhagic complications. If thrombocytosis does not respond to the CML treatment regimen, the use of anagrelide (an agent that prevents megakaryocyte maturation); hydroxyurea, 50 to 75mg/kg/d; anti-aggregants or apheresis may be considered.⁵⁶

Priapism. Persistent painful penile erection may result from sludging and mechanical obstruction by leukemic cells, coagulation within the corpora cavernosa secondary to thrombocytosis, or impingement by the spleen on abdominal veins and nerves. Treatment includes analgesia, hydration, application of warm compresses, radiotherapy (to penis or spleen), and initiation of high-dose chemotherapy (e.g., hydroxyurea, 50 to 75 mg/kg/d).

Meningeal Leukemia. Meningeal leukemia is almost nonexistant in CP and is rare in BC. Intrathecal methotrexate is effective therapy, but most patients eventually die of the hematologic consequences of the blast transformation. The role of prophylactic central nervous system therapy is undefined.

TABLE 21.4 Criteria for Response to CML Therapy

Response	Level of Response	Criteria
Hematologic	Complete (CHR)	Complete normalization of PB counts with WBC < 10 × 10⁹/L Platelet count < 450 × 10⁹/L No immature cells (myelocytes, promyelocytes, blasts) in PB <5% basophils in PB No residual signs/symptoms of disease, resolution of splenomegaly
Cytogenetic	Complete (CCyR)	No Ph¹-positive metaphases
	Partial (PCyR)	1%-35% Ph¹-positive metaphases
	Major (MCyR)	0%-35% Ph¹-positive metaphases (complete + partial)
	Minor	>35%-95% Ph¹-positive metaphases
Molecular	Complete (CMR)^a	BCR-ABL mRNA undetectable by RT-PCR
	Major (MMR)	≥3 log reduction in IS of BCR-ABL mRNA
^aUse of term complete molecular response is controversial, EuropeanLeukemiaNet recommends replacing with molecularly undetectable leukemia with specification of the number of the control gene transcript copies.
PB, peripheral blood; WBC, white blood cell count; IS, International Scale. IS is the ratio of BCR-ABL transcripts to ABL1 transcripts (or other internationally recognized control transcripts) and is expressed/reported as BCR-ABL % on a log scale, where 10%, 1%, 0.1%, 0.01%, 0.0032%, and 0.001% correspond to a decrease of 1, 2, 3, 4, 4.5 and 5 logs respectively.^56,67

Management of Chronic Phase

The initial goal of treatment for CP has been to ameliorate leukocytosis and organomegaly. Early treatment regimens with busulfan or hydroxyurea achieved this goal but rarely achieved permanent remissions due to failure to eradicate all cells of the CML clone. Interferon was the first agent to achieve a significant rate of cytogenetic remissions. Historically, the median survival of patients diagnosed in CP was 5 to 7 years, with 50% to 60% alive at 5 years and 30% surviving to 10 years. With the advent of BCR-ABL1-targeted therapy with directed TKIs such as imatinib and dasatinib, complete cytogenetic molecular responses have now been achieved. Currently, TKIs are the frontline therapy for CML and have improved overall survival to 80% to 90% at 6 years.⁵⁷ However, stem cell transplantation (SCT) remains the only therapy with a track record in producing long-term cure. (For definitions of hematologic, cytogenetic, and molecular responses, see Table 21.4)

Interferon-α. In the 1980s, interferon-α (IFN-α) was the standard therapy for CML and is currently used in patients who are intolerant of TKIs. IFN-α appears to exert a direct antiproliferative effect against both normal and CML myeloid precursors, particularly those of the late progenitor compartment, which is preferentially expanded in CML.

IFN-α is effective in reversing splenomegaly and normalizing the white blood cell and platelet count in early CP. IFN-α-based protocols can achieve CHR (Table 21.5) in more than 80% of patients; approximately 5% to 30% of these patients will also achieve complete cytogenetic response (CCyR) (Ph¹-negative), while another 10% to 38% show a major cytogenetic response (MCyR) (<35% Ph¹-positive cells). Patients achieving a MCyR have projected 5-year survival approaching 90%.⁵⁸ The combination of IFN-α with low-dose cytarabine is superior to IFN-α alone; adult patients who achieved MCyR or CCyR within 2 years of diagnosis achieved a 7-year survival rate of 85%.⁵⁹^,⁶⁰ In a phase II trial testing IFN-α 2b and cytarabine in children and adolescents with CML-CP, 7 of the 12 assessable patients achieved a CHR, after a median time of 3 months. Half the patients achieved MCyR and 14% achieved CCyR. At 13 months, overall survival was 100% with a progression-free survival (PFS) of 71%.⁶¹

Tyrosine Kinase Inhibitors. An exciting new chapter in the treatment of CML specifically and, by extension, for cancer treatment, began with the development of imatinib mesylate (STI571, Gleevec, CGP57148B),⁶² the first agent designed as a molecularly
specific suppressor of the neoplastic clone.⁶³^,⁶⁴ Imatinib was approved by the United States Food and Drug Administration (FDA) for adults with CP CML in 2001 and for children with CP CML in 2003. Since then, three second-generation TK inhibitors (TKIs), dasatinib, nilotinib, and bosutinib have also received FDA approval for adult patients resistant to or intolerant of imatinib. With additional adult studies demonstrating the superiority of dasatinib (DASISION study) and nilotinib (ENESTnd study) over imatinib, therapeutic recommendations issued by both the National Comprehensive Cancer Network and the EuropeanLeukemiaNet include the upfront use of imatinib, dasatinib, or nilotinib for adult patients in CP.⁵⁶^,⁶⁵^,⁶⁶^,⁶⁷

TABLE 21.5 Results of Treatment of Chronic Phase CML

Drug	Dose	Dosing Schedule	Time Point	CHR (%)	CCyR (%)	MMR (%)	OS (%)	PFS (%)
Adult Studies
Interferon α + cytarabine (IRIS)⁶⁹			18 mo	69	15		95.1	91.5
Imatinib (IRIS)⁶⁹	400 mg	Daily	18 mo	97	69		97.2	96.7
Imatinib (DASISION)⁶⁵	400 mg	Daily	36 mo		83	55	93.2	90.9
Imatinib (ENESTnd)^66,257	400 mg	Daily	36 mo		77 24 mo	53	94	93.5
Imatinib⁷¹	800 mg	Daily	18 mo		75	68	>95	>94
Dasatinib (DASISION)⁶⁵	100 mg	Daily	36 mo		87	69	93.7	91.0
Nilotinib⁸²	400 mg	BID	18 mo	77	41		91	67
Nilotinib	300 mg	BID	36 mo		87	73	95.1	96.7
(ENESTnd)^66,154	400 mg	BID	36 mo		24 mo 85 24 mo	70	97	98.1
Pediatric Studies
Interferon α + cytarabine⁶¹			13 mo	58	14		100	71
Imatinib (COG-phase I)⁷²	260-570 mg/m²	Daily		100	83
Imatinib (AAML0123)⁷³	340 mg/m²	Daily	36 mo	80^a	72	MMR NR CMR 27	98	96
Imatinib⁷⁴	260 mg/m²	Daily	12 mo 36 mo	98	61 77	31 57	98 98	100 98
Dasatinib (COG-phase I)⁸⁰	50-80 mg/m²	Daily			37.5
^aCHR after 2 months of therapy
CCyR, complete cytogenetic response; CHR, complete hematologic response; MMR, major molecular response; CMR, complete molecular response (BCR-ABL1 transcript not detectable); OS, overall survival; PFS, progression-free survival.

Regardless of which TKI is selected for initial therapy, response to TKI therapy is the most important prognostic factor and is determined by the achievement of hematologic, cytogenetic, and molecular responses as defined in Table 21.4.

Imatinib Mesylate. Imatinib is a small 2-phenylaminopy-rimidine molecule that blocks the activity of BCR-ABL protein by occupying its kinase pocket, thereby blocking the binding of ATP (Fig. 21.6); this prevents the constituent activation of the TK and abrogates its downstream phosphorylation of target proteins responsible for proliferation and for inhibition of apoptosis. It is relatively selective in its activity, with the ability to block all of the abl kinases (including p210, p185, v-abl, and c-abl variants), but relatively few other TKs, with the exception of arg, stem cell factor (cKit), and PDGFR. ⁶²

When compared to historical controls treated with IFN regimens, patients treated with imatinib were more likely to achieve a cytogenetic response and have overall better survival outcomes (Table 21.5).⁶⁴^,⁶⁸ The IRIS (International Randomized Study of Interferon and STI571) study in newly diagnosed adult patients in CP randomly compared imatinib 400 mg daily dose versus interferon plus ara-C. Imatinib achieved higher rates of CHR (95% vs. 56%) and CCyR (76.2% vs. 14.5%) at 18 months.⁶⁹ In addition, the rate of progression to AP or BC decreased (3.3% vs. 8.5%). In the 6-year follow-up of patients on the imatinib arm, the cumulative best CCyR was 82% with the estimated 6-year eventfree survival, overall survival, and PFS of 83%, 88% (95% when non-CML-related deaths were censored), and 93%, respectively.⁵⁷ Higher doses of imatinib (600 to 800 mg/d) have been found to produce higher and more rapid rates of cytogenetic and molecular response.⁷⁰^,⁷¹ For example, a large randomized multicenter trial comparing imatinib 400 mg/d to tolerability-adapted imatinib at a starting dose of 800 mg/d demonstrated that, overall, more patients in the higher dose group achieved CMR (56.8% vs. 45.5%).⁷¹ However, higher dose regimens have been associated with higher rates of hematologic toxicity and fluid retention. Therefore, 400 mg/d is the currently recommended starting dose of imatinib for adults with CML-CP.⁵⁶^,⁶⁷

A phase I study of imatinib was conducted by the Children’s Oncology Group in pediatric patients with Ph¹-positive CML refractory to IFN therapy, relapsing after SCT, or with refractory/relapsed Ph¹-positive ALL.⁷² Doses of 260 to 570 mg/m² were well tolerated. A maximum tolerated dose was not determined and dose-limiting toxicity was not seen. CHR was achieved within 30 days in 100% of CP patients. Eighty-three percent obtained a CCyR within 2 to 6 months.⁷² The most common side effects were mild nausea, vomiting, and diarrhea. Moderate anemia, thrombocytopenia, and neutropenia were seen in one-third of the patients. Other toxicities included diarrhea, abdominal pain, headaches, fatigue, stomatitis, and bone pain. The adverse effect profile differed slightly from that seen in the adult population, where the most frequently reported toxicities were gastrointestinal, dermatologic (rash, edema), and musculoskeletal disturbances.

In the pediatric phase II Children’s Oncology Group trial AAML0123, the efficacy of imatinib at a dose of 340 mg/m² was
assessed in newly diagnosed children (ages 2 to 19 years) in CP.⁷³ CHR was attained by 80% of patients by 2 months of therapy. CCyR was achieved in 38% of patients by 3 months and overall in 72% of patients at a median time of 5.6 months. Ninety-one percent of the patients who achieved CCyR did so by 9 months of therapy. The rate of CMR was 27%. Progression-free and overall survival at 3 years were 72% and 92%, respectively.⁷³ A French multicenter phase IV trial treated 44 newly diagnosed children with CP CML with 260 mg/m² imatinib. At 36 months, PFS was 98%. The rates of CCyR at MMR at 12 months were 61% and 31%, respectively. Cumulative CCyR and MMR rates were 77% and 57%, respectively.⁷⁴

Figure 21.6 Mechanism of action of imatinib mesylate. In the untreated state (left), BCR-ABL protein has an open pocket accessible to ATP. This facilitates transfer of a phosphate moiety from ATP to a tyrosine residue on a target substrate molecule. The activated molecule is then released to interact with downstream effector molecules, which can promote oncogenesis. Imatinib inhibits this process by competing with ATP for the kinase pocket, thereby preventing phosphorylation of substrate and effector molecules. (Courtesy of Eleanore Rhodes.)

Dosing in children of 260 mg/m² and 340 mg/m² provide similar imantinib exposures to adults receiving 400 mg/d and 600 mg/d, respectively.⁷² The most commonly recommended starting dose for children is in the range of 340 mg/m².⁷⁵

Second-generation ABL kinase inhibitors.

Dasatinib. Dasatinib (Sprycel^TM, formerly BMS-354825) is a dual SCR-family kinase and ABL kinase inhibitor that also inhibits ephrin receptor kinases, PDGFR, Kit and other tyrosine and serine/threonine kinases. It is more potent than imatinib in inhibiting the growth of resistant cell lines except those harboring the T315I BCR-ABL1 mutation. Dasatinib binds to both the active and the inactive conformation of the ABL kinase domain.⁷⁶ In the phase II START trials of dasatinib, 70-mg twice-daily was given to patients resistant to or intolerant of imatinib.⁷⁷^,⁷⁸ Efficacy was demonstrated in patients with imatinib-resistant CML in CP, AP and BC (see Tables 21.5 and 21.6). The multicenter DASISION trial, in which adult patients with newly diagnosed CML-CP were randomized to receive either dasatinib 100 mg/d or imatinib 400 mg/d, showed that dasatinib resulted in faster and deeper cytogenetic and molecular responses compared with imatinib.⁶⁵^,⁷⁹ After a minimum follow-up of 3 years, the percentage of patients achieving confirmed CCyR (cCCyR, CCyR on two separate occasions at least one month apart) at any time was 82% with dasatinib versus 77% with imatinib, which was not statistically different.⁶⁵ However, the median time to cCCyR was 3.1 months with dasatinib versus 5.8 months with imatinib.⁶⁵^,⁷⁹ The cumulative MMR rate by 36 months was 69% for dasatinib versus 55% for imatinib (HR, 1.62; p < 0.0001). PFS and OS did not differ significantly between the two arms at 36 months.⁶⁵

Only gold members can continue reading. Log In or Register to continue