Changing Gene Expression Patterns during Early Treatment of B-Precursor Acute Lymphoblastic Leukemia May Be Predictive of Relapse

• Published in: Blood Vol. 110; no. 11; p. 232
• Main Authors: de Haas, Valerie, Dee, Rob, Cheroutre, Goedele, van den Berg, Henk, Caron, Huib, Van der Schoot, C. Ellen
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 16-11-2007
• ISSN: 0006-4971; 1528-0020
• DOI: 10.1182/blood.V110.11.232.232

Treatment of pediatric ALL is based on the concept of tailoring the intensity of treatment to a patients risk. Clinical studies have shown that it is possible to stratify patients according to the levels of minimal residual disease after induction therapy and early during further treatment, since it has been demonstrated that the MRD level is the best predictive level for disease outcome. More recently, it has been shown that gene expression profiles of leukemic cells at diagnosis might be correlated with outcome. In previous studies we reported that slow responding subclones represent the clones causative for a leukemic relapse in oligoclonal disease. Based on these results, we hypothesized that the gene expression profile of the slow responding subclones present after the first weeks of chemotherapy might be more predictive than the profiles of all leukemic cells at diagnosis. Twenty-four genes were selected; most signalling molecules, transcription factors and functions relevant for oncogenesis, drug resistance and metastasis. Selection of genes was based on the presently available data on prognostic cDNA microarry studies of cytogenetically defined subgroups of childhood ALL. In particular, we analyzed results of recently published studies that compared gene expression levels between diagnosis and relapse in B-precusor acute lymphoblastic leukemia. (Staal, 2003 and Beesley, 2005). Gene sequences were obtained from public databases. Genes were tested on different leukemic cell lines. For all cell lines differences in gene expression level were demonstrated. The same panel of genes was tested on diagnostic samples of 16 ALL patients, subsequently followed by investigation of paired diagnosis - day 15 - relapse samples of 3 relapsed ALL patients. Leukemic material was obtained from cryopreserved bone marrow samples. All leukemic cells were purified by MACS purification based on markers expressed on the tumour, i.e. CD34, CD19 and CD10. RNA extraction and cDNA synthesis was performed according to the TRIZOL protocol. Expression levels were determined in a SYBR Green based real-time PCR assay. We were able to show different gene expression profiles in the 16 tested diagnostic samples. For the paired samples from relapsed B-precursor ALL patients, the expression level of several genes at day 15 was different (ΔCT>1) in regard to diagnosis. Moreover, the changed expression at day 15 was comparable to the expression level of this gene at relapse. We conclude that indeed we were able to demonstrate that some of the genes have a changing pattern of expression during early therapy (day15), a pattern which is comparable to the pattern of gene expression at relapse and which is different from the pattern at diagnosis. We also demonstrated that purification of the bone marrow samples is necessary to be certain that the gene expression shown is relevant for the leukemic cells and not contaminated by other cells, i.e. T-cells.
[Display omitted]