Whereas hematologic malignancies are of clonal origin (i.e., they are derived from a single transformed cell), individual neoplastic cells from a patient ’s malignancy exhibit a great deal of phenotypic diversity and acquire secondary mutations that affect proliferation, drug sensitivity, and resistance. This diversity likely arises from the progeny of clonal populations and subsets of stem cells. In animal models, it has been shown that the clones themselves can give rise to progeny that can transmit the clonal malignancy after transplantation into secondary recipients, suggesting that stem cells are not required to transmit the malignant phenotype.
New evidence indicates that leukemia stem cells are more quiescent, have higher levels of protective proteins such as efflux pumps for drugs, and have higher levels of DNA repair proteins or antiapoptotic proteins than the more abundant cell making up the circulating population of cells. Tumor cell heterogeneity arises as a consequence of spontaneous mutational events, changes in gene promoter methylation, abnormal expression of transcription factors, lymphoid reactivity, and cytokine responsiveness. For example, a mutation or change in expression that renders a hematopoietic cell clone autonomous or growth factor–independent would be expected to render such cells less susceptible to adverse environmental conditions (e.g., growth factor withdrawal). Similarly, one would also predict that a genetic change facilitating cell cycle entry or disruption of cellular maturation would ultimately lead to overgrowth of affected clones. For obvious reasons, mutations that interfere with drug metabolism or the cell death pathway itself would provide a net survival advantage, particularly under the selection pressure of cytotoxic drug treatment.
Malignant myeloid and lymphoid cells have many reasons to have increased mutational rates. Genomic instability can arise from dysregulation of the cell cycle machinery because of a number of events, including perturbations of cyclins leading to MYC overexpression; AKT (protein kinase B) activation; disruption of replication sequences; loss of DNA repair enzymes such as mismatch repair (MMR) enzymes; loss of proper homologous recombination from a defect in the BRCA (BReast CAncer gene 2) –Fanconi pathways; and loss of ATM (ataxia telangiectasia mutated) /ATR (ataxia telangiectasia and Rad3 related) kinases, which can give rise to chromosomal recombination, loss, and microsatellite instability, and loss of check point regulation. These events can give rise to intraclonal emergent point mutations, translocations, and intragenic losses that might not only result in malignant transformation, but also lead to disruption of genomic stability and selection in favor of proliferative and apoptosis resistant subclones. Leukemic clonal evolution favors drug resistance.
Common mechanisms may be involved in events associated with malignant transformation and the development of mutations that result in tumor heterogeneity. For example, the cell cycle checkpoint and tumor suppressor gene, TP53, is induced during DNA damage, leading to G 1 arrest and, if the damage is too severe to repair, cell death by apoptosis occurs. The presumed goal of this process is to eliminate cells that develop deleterious mutations as a result of damage to the genome. Loss of TP53 may not only increase cellular survival by inhibiting the cell death process, but may also promote the transmission of mutations that would otherwise be deleted. In this manner, a defect of the cell death pathway can have multiple consequences, including (1) selection of cells exhibiting a growth advantage over their normal counterparts, (2) development of drug resistance, and (3) promotion of mutations that result in either (1) or (2), as well as neoplastic cell heterogeneity. Age-dependent changes in these processes may explain the more favorable behavior of leukemias and lymphomas in response to chemotherapy in young patients than older patients.
A model of the relationship between tumor growth rate, the occurrence of spontaneous mutations, and the development of drug resistance was first described by Goldie and Coldman and is referred to as the Goldie and Coldman hypothesis. In this model, the size of a tumor depends on a complex interaction between tumor growth rate and cell loss, the latter stemming from the status of the cell death process, exhaustion of available nutrients, and outstripping of the blood supply. As tumors increase in size, the cell death rate tends to increase. The heterogeneous nature of additional mutations makes it likely that multiple mechanisms of resistance will develop as well. From an operational standpoint, this model has clear implications for the rational design of therapeutic strategies and provides a basis for early and intensive combination drug therapy. The successful implementation of this strategy is exemplified by the administration of dose-intensive multidrug regimens (i.e., the BEACOPP [bleomycin, etoposide, Adriamycin, cyclophosphamide, vincristine (Oncovin) procarbazine, and prednisone] regimen in Hodgkin lymphoma, CODOX-M-IVAC [cyclophosphamide, vincristine, doxorubicin, high-dose methotrexate alternating with ifosfamide, etoposide and high-dose cytarabine] in Burkitt lymphoma [non-Hodgkin lymphoma (NHL)]) and combinations of cytotoxic agents with monoclonal antibodies, such as CHOP (cyclophosphamide, hydroxydaunorubicin, vincristine [Oncovin], and prednisone)–rituximab, which are potentially curative when given early in the course of the disease. Other examples include combined use of multitargeted agents such as lenalidomide and bortezomib for myeloma, fludarabine, cyclophosphamide and rituximab for chronic lymphocytic leukemia (CLL), or maneuvers to overcome resistance of pretreated disease outside of cell-based therapies. However, as predicted by the model, administration of these or other intensive regimens in patients with relapsed or late stage disease generally fails because of a generalized resistance of tumor cells to all classes of chemotherapeutic agents.
