A role for altered microtubule polymer levels in vincristine resistance of childhood acute lymphoblastic leukemia xenografts.

The microtubule-depolymerizing drug, vincristine, is effective in the treatment of acute lymphoblastic leukemia (ALL). Although vincristine resistance mechanisms have been extensively characterized in cell lines, their clinical relevance is poorly understood. The aim of the current study was to define clinically relevant mechanisms of vincristine resistance in a panel of childhood ALL xenografts established in immune-deficient (nonobese diabetic/severe combined immunodeficient) mice. We also studied two independent xenograft sublines that were selected by in vivo vincristine exposure. In vitro vincristine sensitivity determined by a stromal coculture, murine bone marrow stromal cell line (MS-5), assay, but not methyl-thiazolyl-tetrazolium metabolic activity assay, significantly correlated (P = 0.05) with the length of the patients' first remission. Investigations into mechanisms of resistance revealed no association with steady-state vincristine accumulation or increased activity and/or expression of ATP-binding cassette transporters, although increased intracellular levels of polymerized tubulin significantly correlated with resistance (r = 0.85; P = 0.0019). Two xenograft sublines selected by in vivo vincristine exposure exhibited a 2-fold increase in polymerized tubulin levels compared with the parental subline (P < 0.05), reflecting their in vivo vincristine resistance. In this study, a vincristine-resistant xenograft with high levels of polymerized tubulin was relatively sensitive to the microtubule-polymerizing drug paclitaxel. These results indicate that the balance between polymerized and nonpolymerized tubulin may be an important determinant of response to Vinca alkaloid-based chemotherapy regimens in childhood ALL.

Variegated clonality and rapid emergence of new molecular lesions in xenografts of acute lymphoblastic leukemia are associated with drug resistance.

The use of genome-wide copy-number analysis and massive parallel sequencing has revolutionized the understanding of the clonal architecture of pediatric acute lymphoblastic leukemia (ALL) by demonstrating that this disease is composed of highly variable clonal ancestries following the rules of Darwinian selection. The current study aimed to analyze the molecular composition of childhood ALL biopsies and patient-derived xenografts with particular emphasis on mechanisms associated with acquired chemoresistance. Genomic DNA from seven primary pediatric ALL patient samples, 29 serially passaged xenografts, and six in vivo selected chemoresistant xenografts were analyzed with 250K single-nucleotide polymorphism arrays. Copy-number analysis of non-drug-selected xenografts confirmed a highly variable molecular pattern of variegated subclones. Whereas primary patient samples from initial diagnosis displayed a mean of 5.7 copy-number alterations per sample, serially passaged xenografts contained a mean of 8.2 and chemoresistant xenografts a mean of 10.5 copy-number alterations per sample, respectively. Resistance to cytarabine was explained by a new homozygous deletion of the DCK gene, whereas methotrexate resistance was associated with monoallelic deletion of FPGS and mutation of the remaining allele. This study demonstrates that selecting for chemoresistance in xenografted human ALL cells can reveal novel mechanisms associated with drug resistance.

Pharmacokinetic modeling of an induction regimen for in vivo combined testing of novel drugs against pediatric acute lymphoblastic leukemia xenografts.

Current regimens for induction therapy of pediatric acute lymphoblastic leukemia (ALL), or for re-induction post relapse, use a combination of vincristine (VCR), a glucocorticoid, and L-asparaginase (ASP) with or without an anthracycline. With cure rates now approximately 80%, robust pre-clinical models are necessary to prioritize active new drugs for clinical trials in relapsed/refractory patients, and the ability of these models to predict synergy/antagonism with established therapy is an essential attribute. In this study, we report optimization of an induction-type regimen by combining VCR, dexamethasone (DEX) and ASP (VXL) against ALL xenograft models established from patient biopsies in immune-deficient mice. We demonstrate that the VXL combination was synergistic in vitro against leukemia cell lines as well as in vivo against ALL xenografts. In vivo, VXL treatment caused delays in progression of individual xenografts ranging from 22 to >146 days. The median progression delay of xenografts derived from long-term surviving patients was 2-fold greater than that of xenografts derived from patients who died of their disease. Pharmacokinetic analysis revealed that systemic DEX exposure in mice increased 2-fold when administered in combination with VCR and ASP, consistent with clinical findings, which may contribute to the observed synergy between the 3 drugs. Finally, as proof-of-principle we tested the in vivo efficacy of combining VXL with either the Bcl-2/Bcl-xL/Bcl-w inhibitor, ABT-737, or arsenic trioxide to provide evidence of a robust in vivo platform to prioritize new drugs for clinical trials in children with relapsed/refractory ALL.

Characterization of childhood acute lymphoblastic leukemia xenograft models for the preclinical evaluation of new therapies.

Continuous xenografts from 10 children with acute lymphoblastic leukemia (ALL) were established in nonobese diabetic/severe combined immunodeficient (NOD/SCID) mice. Relative to primary engrafted cells, negligible changes in growth rates and immunophenotype were observed at second and third passage. Analysis of clonal antigen receptor gene rearrangements in 2 xenografts from patients at diagnosis showed that the pattern of clonal variation observed following tertiary transplantation in mice exactly reflected that in bone marrow samples at the time of clinical relapse. Patients experienced diverse treatment outcomes, including 5 who died of disease (median, 13 months; range, 11-76 months, from date of diagnosis), and 5 who remain alive (median, 103 months; range, 56-131 months, following diagnosis). When stratified according to patient outcome, the in vivo sensitivity of xenografts to vincristine and dexamethasone, but not methotrexate, differed significantly (P =.028, P =.029, and P =.56, respectively). The in vitro sensitivity of xenografts to dexamethasone, but not vincristine, correlated significantly with in vivo responses and patient outcome. This study shows, for the first time, that the biologic and genetic characteristics, and patterns of chemosensitivity, of childhood ALL xenografts accurately reflect the clinical disease. As such, they provide powerful experimental models to prioritize new therapeutic strategies for future clinical trials.