The first cases of what was probably CML were described by Bennett and Virchow in the mid 1840s.¹ In 1960, Philadelphia cytogeneticists Nowell and Hungerford² described a “minute” chromosome 22 in CML cells that became known as the Philadelphia chromosome (Ph). In 1973, Janet Rowley³ discovered that Ph is in fact the result of a reciprocal translocation between chromosomes 9 and 22 [t(9;22) (q34;q11)]. The genes juxtaposed by the translocation were subsequently identified as ABL (Abelson) on 9q34⁴ and breakpoint cluster region (BCR) on chromosome 22q11 (Fig. 32.1). The next critical discoveries were that the constitutive tyrosine kinase activity of BCR-ABL is required for cellular transformation and that the clinical disease was reproducible in a murine model.⁵,⁶ According to the World Health Organization the presence of BCR-ABL in the context of a myeloproliferative neoplasm is diagnostic of CML, although the translocation is also found in a subset of patients with acute lymphoblastic leukemia (ALL) and rare cases of acute myeloid leukemia (AML).

The breakpoints within ABL occur upstream of exon 1b, downstream of exon 1a, or, more frequently, between the two. Regardless of the exact breakpoint location, splicing of the primary transcript yields an mRNA in which BCR sequences are fused to ABL exon a2. Breakpoints within BCR localize to one of three breakpoint cluster regions (bcr). More than 90% of CML patients and one-third of Ph+ ALL patients express the 210kDa isoform of BCR-ABL, in which the break occurs in the 5.8-kb major breakpoint cluster region (M-bcr), which spans exons e12-e16 (formerly exons b1-b5). Alternative splicing gives rise to either b2a2 (e13a2) or b3a2 (e14a2) transcripts.⁵ In remaining Ph+ ALL patients and in rare CML cases, the breakpoints are further upstream in the 54.4-kb minor breakpoint cluster region (m-bcr), generating an e1a2 transcript that is translated into p190^BCR-ABL.⁷ A third breakpoint downstream of exon 19 in the micro breakpoint cluster region (μ-bcr) gives rise to an e19a2 BCR-ABL mRNA and p230^BCR-ABL and is associated with neutrophilia.

Other BCR-ABL variants, including b2a3, b3a3, e1a3, e6a2 or e2a2, have been observed in isolated cases.⁸ Deletions flanking the breakpoints are detected in a subset of patients, which confer a poor prognosis to patients on interferon therapy, but probably not on imatinib.⁹,¹⁰,¹¹ The reciprocal ABL-BCR transcript, although
detectable in approximately two-thirds of patients, does not seem to play any significant role in pathogenesis.¹²

p210^BCR-ABL contains several distinct domains (Fig. 32.2).¹³ The N-terminal coiled-coil domain of BCR allows BCR-ABL dimerization, which is critical for kinase activation. The p210^BCR-ABL protein also retains the serine/threonine kinase and Rho guanine nucleotide exchange factor homology (Rho-GEF) domains of BCR, which are deleted in p190^BCR-ABL, which may explain the differences in disease phenotype associated with the two variants. In contrast to BCR, the ABL sequence is almost completely retained, including SRC homology domains 2 and 3, the tyrosine kinase domain, a proline-rich sequence, and a
large C terminus with nuclear localization signal, DNA-binding, and actin-binding domains. The N-terminal “cap” region of ABL negatively regulates kinase activity by binding to a hydrophobic pocket at the basis of the kinase domain, which in the Ib isoform is mediated by N-terminal myristoylation. It is believed that the replacement of the cap with BCR sequences contributes to constitutive kinase activation.

Numerous substrates and binding partners of BCR-ABL have been identified (Fig. 32.2). Current efforts are directed at linking these pathways to the specific phenotypic defects that characterize CML, such as increased proliferation, decreased apoptosis, defective adhesion to bone marrow stroma, and genetic instability.¹⁴ As comprehensive review of the multiple implicated pathways is beyond the scope of this chapter, we will focus on those for which strong evidence supports a ratelimiting role in disease pathogenesis.

The most commonly used murine model of CML is retroviral expression of BCR-ABL in bone marrow followed by transplantation into lethally irradiated syngeneic recipients, which develop a CML-like myeloproliferative neoplasm.³⁹ Recently, an inducible transgenic mouse model has been developed, in which conditional BCR-ABL expression is under the control of the 3′ enhancer of the murine stem cell leukemia (SCL) gene. This model serves as a promising new tool for studying leukemogenic mechanisms in hematopoietic stem cells during disease initiation and progression.⁴⁰ Lastly, xenograft models use various strains of immunodeficient mice for engraftment of primary CML cells.⁴¹ The problem with xenograft models is that the engraftment of CP cells is low, probably because of their compromised interactions with the microenvironment as a result of species differences in cytokines and adhesion molecules. More promising results have been obtained by injecting CML cells directly into the livers of newborn mice.⁴²

The origin of CML in a pluripotent hematopoietic stem cell (HSC) was elegantly demonstrated in the late 1970s in studies showing clonality of granulocytes, erythrocytes, and platelets of female patients heterozygous for glucose-6-phosphate dehydrogenase, a polymorphic X-chromosomal gene.⁴³ BCR-ABL does not confer self-renewal, implying that Ph must be acquired by an HSC already endowed with this capacity.⁴⁴ For unknown reasons the main cellular expansion occurs in the progenitor cell compartment, while at least initially the majority of HSC are Ph-negative.⁴⁵ Serial xenograft studies have shown that CML leukemia stem cells (LSCs) reside within the quiescent CD34+ 38− fraction of bone marrow cells. Significant progress has recently been made by the identification of the IL-1 receptor-associated protein (IL-1RAP) as a surface marker specifically expressed on CD34+38− CML HSC.^45a If confirmed, this will enable studies into the biology of these cells. Several genes were shown to have a critical role for leukemia-initiating cells maintenance in CML, including promyelocytic leukemia,⁴⁶ the hematopoietic-specific Rac2 GTPase,⁴⁷ smoothened, which is an essential component of the hedgehog pathway,⁴⁸ and β-catenin. The fact that these genes are also critical for maintenance and self-renewal of normal HSCs may be an obstacle to exploiting them as therapeutic targets.

Disease progression is believed to be due to the accumulation of molecular abnormalities that lead to a loss of terminal differentiation capacity of the leukemic clone, which however continues to depend on BCR-ABL activity. BCR-ABL mRNA and protein levels are higher in BP than in CP cells, including CD34+ granulocyte macrophage progenitors (GMPs), which are expanded in BP.⁴⁹ Another mechanism that enhances BCR-ABL activity in BP is inactivation of PP2A through up-regulation of SET, which in turn inhibits the PP2A phosphatase and its substrate protein tyrosine phosphatase I (SHP1). As SHP1 normally promotes BCR-ABL degradation through dephosphorylation, its reduced activity stabilizes BCR-ABL.⁵⁰

The most striking feature of BP, the loss of differentiation capacity, suggests that the function of key myeloid transcription factors must be compromised. Occasionally, the differentiation block can be ascribed to mutations that result in the formation of dominant-negative transcription factors such as AML1-EVI or NUP98-HOXA9, which block differentiation or favor preferential growth of immature precursors.⁵¹,⁵²,⁵³ Isolated cases of myeloid transformation have been associated with the acquisition of core binding factor mutations typical of AML. A more universal mechanism appears to be the BCR-ABL-induced down-regulation of CCAAT/enhancer binding protein-α (CEBPα) through the stabilization of the translational regulator heterogeneous nuclear ribonucleoprotein E2 (hnRNP E2), which is low or undetectable in CP but readily detectable in BP CML.⁵⁴ Interestingly, dominant-negative mutations of CEBPα are fairly common in AML,⁵⁵ but rare in CML-BP.⁵⁶

Aberrant activation of Wnt/β-catenin signaling is believed to contribute to CML progression by conferring self-renewal capacity to GMPs. Activated β-catenin regulates self-renewal by undergoing translocation to the nucleus where it interacts with lymphoid enhancer factor/T-cell factor and regulates the transcription of genes such as Myc and Cyclin D1.⁵⁷ The acquisition of self-renewal by GMPs is expected to greatly increase the pool of LSCs in BP. Interestingly, expression microarray studies have implicated β-catenin activation not only in disease progression but also in resistance to tyrosine kinase inhibitors, supporting the view that drug resistance and disease progression share a common genetic basis.⁵⁸,⁵⁹ This has implications for prognostication as well as for the development of strategies to prevent progression and overcome resistance.

Attempts to identify the origin of CLL with respect to a normal B-cell counterpart has also occurred and remains a controversial area.⁶⁶,⁶⁷,⁶⁸ Unlike most other B-cell lymphomas and leukemia with the exception of mantle cell lymphoma, CLL expresses typical mature B-cell markers with coexpression of CD5. This prompted many to hypothesize that CLL may be derived from CD5+ B cells whose immunoglobulin (Ig)V_H is unmutated. However, the overall phenotype of CLL with expression of CD5, CD23, and CD19, and low levels of sIgM or IgD is not observed in any other type of normal B-cell counterpart. Additionally, investigators identified that approximately 40% of CLL cases are IgV_H unmutated, whereas the remainder are IgV_H mutated.⁶⁹,⁷⁰ These two groups also were shown to have distinct clinical features, prompting the hypothesis that CLL may represent two distinct diseases.⁶⁹,⁷⁰