Minimal Residual Disease Assessment in Childhood Acute Lymphoblastic Leukemia

Abstract: Traditionally, response to treatment in hematological malignancies is evaluated by light microscopy of bone marrow (BM) smears, but due to more effective therapies more sensitive methods are needed. Today, detection of minimal residual disease (MRD) using immunological and molecular techniques can be 100 times more sensitive than morphology. The main aim of this thesis was to compare and evaluate three currently available MRD methods in childhood acute lymphoblastic leukemia (ALL): (i) real-time quantitative PCR (RQ-PCR) of rearranged antigen receptor genes, (ii) multicolor flow cytometry (FCM) of leukemia-associated immunophenotypes and (iii) real-time quantitative PCR of fusion gene transcripts (RT-PCR).In paper I, we assessed the applicability of RQ-PCR in a population-based cohort of childhood ALL diagnosed in Sweden between 2002-2006. Clonal IG/TCR rearrangements were identified in the 96% of the 279 ALL cases. Using RQ-PCR, the quantitative range of 10-3 was reached in 93% of B-cell precursor (BCP) ALL and 86% of T-cell ALL (T-ALL) by at least one target gene. In paper II, we compared MRD detection using both RQ-PCR and FCM in the context of NOPHO ALL-2000 protocol. By applying the stratification threshold of ?0.1% MRD late during induction therapy (day 29), we could demonstrate that both methods can predict the risk of BM relapse but not extramedullary relapse. However, the threshold of ?0.2% MRD appears to be more optimal using RQ-PCR in BCP ALL, whilst in T-ALL, the results indicate that RQ-PCR is preferable for MRD assessment.The stability of RNA in vitro is a critical factor when using sensitive molecular techniques such as MRD detection. In paper III, we evaluated the influence on MRD detection when blood is collected in tubes with RNA stabilization reagents (PAX gene Vacutatiner®) compared to collection in EDTA-tubes (non-stabilized). We analyzed 68 matched samples from chronic myeloid leukemia patients and the results indicated that non-stabilized blood processed within 30 hours is preferable for MRD detection.In paper IV, follow-up samples from eight children with Philadelphia positive (Ph+) ALL were evaluated with the three available MRD methods. MRD measured by the fusion gene transcripts (BCR-ABL1) appeared to be the most sensitive method, however, precise quantification can be difficult and the other methods are thus complementary.In conclusion, all three applied MRD methods are useful and correlate to each other, although not necessary exchangeable in individual patients. We also conclude that MRD assessment by RQ-PCR, based on rearranged IG/TCR genes and multicolor FCM are predictive for identification of high risk childhood ALL patients.