Title

 

Authors

G. Stocco ^1,2, W. Yang ¹, K. Crews ¹, W. Thierfelder ¹, S. Paugh ¹, B. Diouf ¹, R. Mc Corkle ¹, M. Den Boer ³, M. Relling ¹, R. Pieters ³, W. Evans ^1
 
1St.Jude Children’s Research Hospital, Memphis, USA
²University of Trieste, Trieste, Italy
³Erasmus Medical Center, Rotterdam, The Netherlands

 

Abstract

Drug resistance profiles of blasts to anti-leukemic medications at diagnosis, as measured with the in vitro drug cytotoxicity assay MTT, have prognostic value in childhood acute lymphoblastic leukemia (ALL) and may therefore, be used to improve patients’ risk-group stratification (Holleman & Cheok, NEJM 2004). Previous studies have shown that in vitro resistance to prednisone, vincristine and asparaginase correlates with an increased incidence of early events (non-response or relapse within 2.5 years of diagnosis) (Den Boer et al., 2003). Differential expression of a relatively small number of genes, identified by high-throughput agnostic microarray analysis, is associated with drug resistance evaluated by the MTT assay and treatment outcome in childhood ALL (Holleman & Cheok et al., 2004). These studies have provided new insights in the pharmacodynamics of important anti-leukemic medications, for example identifying the contribution of members of the SWI/SNF transcription complex, such as SMARCB1, as associated with glucocorticoid resistance in ALL cells (Pottier et al., 2008). High-throughput gene expression studies have not yet been performed to assess ALL resistance to the thiopurine antimetabolite 6-mercaptopurine (6MP). Therefore an analysis was done to identify sets of differentially expressed genes in primary ALL cells that were sensitive or resistant to 6MP. At St. Jude in Memphis TN USA and at Erasmus Medical Center (EMC) in Rotterdam (Netherlands) we tested leukemia cells from children for sensitivity in vitro to 6MP; the cells were then subjected to an assessment of gene expression with the use of 22,277 probe sets to identify differentially expressed genes in drug sensitive and drug resistant ALL. Complete data was obtained for 121 patients at St. Jude (104 B lineage, 17 T lineage) and 156 patients at EMC (127 B lineage, 29 T lineage). In these patients, no significant relationship of patient’s gender, age or ethnicity could be found with 6MP sensitivity. ALL cell gene expression analysis was done by comparing gene expression in the most sensitive vs the most resistant quantiles of the population, using t-test or in the whole population using Spearman’s rank correlation. We identified sets of differentially expressed genes in ALL cells that were sensitive or resistant to 6MP in the St. Jude or EMC cohorts. Application of  gene expression signatures identified in one population and random forest models to estimate sensitivity and resistance to 6MP in the other cohort showed a significant correlation between the estimated values and the measured values (for example, expression of the top most significant 100 probesets according to the Spearman’s test in St.Jude predicted IC50 in the EMC cohort with a rho of 0.4 and a p-value < 0.0001). The gene signatures identified contained genes plausibly related to 6MP induced cytotoxicity, such as HMGB1 and ADA. Pathway analysis done by gene set enrichment analysis, using the KEGG pathways as the reference gene set collection, showed that among the genes significantly related to 6MP sensitivity in both populations there was an over-representation of genes in the cell cycle and DNA replication pathways.
These results indicate that global patterns of gene expression are related to 6MP sensitivity in ALL blasts at diagnosis. Validation of candidate genes is being performed using appropriate experimental techniques (e.g. gene knockdown), in order to clarify the molecular mechanism connecting thiopurine sensitivity to genes identified. This characterization of 6MP pharmacodynamics in ALL through pharmacogenomics could provide insights clinically translatable to optimize the therapy of pediatric ALL, in order to increase the efficacy of treatment and to reduce the adverse events of chemotherapy.
Holleman A. & Cheok M. et al. (2004). N Engl J Med. 351, 533-42.
Den Boer M. et al. (2003). J Clin Oncol. 21, 3262-8.
Pottier et al. (2008). J Natl Cancer Inst. 100,1792-803.