ALBERT

Description: Approximately 25% of children with acute lymphoblastic leukemia (ALL) relapse after frontline therapy. Current stratification methods using clinical/biological criteria fail to identify a significant proportion of these children. Numerically more patients classified as standard/medium risk relapse than those with high-risk ALL. Early identification of this group with intensification of therapy may prevent this. We compare two techniques, multi-parameter flow cytometry (FCM) and real-time polymerase chain reaction (RQ-PCR) to monitor minimal residual disease (MRD) in bone marrow after initial therapy in newly diagnosed patients. We report our initial experience in patients with B-lineage ALL from a single institution (CHW) treated on the ANZ Haematology-Oncology Group Study VIII protocol for ALL. Multiparameter FCM was performed using two different four colour combinations of antibodies: CD19APC/CD45PerCP/CD10FITC/CD20PE and CD19APC/CD45PerCP/CD34FITC/CD9PE. A series of dual parameter displays were generated to define normal B cell differentiation pathways and this allowed the discrimination of malignant precursor B cells. RQ-PCR used clone specific primers and Taqman probes selected after sequencing the immunoglobulin heavy chain and/or T cell receptor gene rearrangements detected in the diagnosis DNA. The RQ-PCR assays had greater sensitivity and were highly reproducible. The timepoints used for analysis were Day 33 (D33) and D79, following induction and consolidation respectively. Of 45 patients, 44 (98%) had informative RQ-PCR markers: 12 (27%) with high MRD levels at D33 (〉10−3) while 3 (6.7%) had high levels at D79. In contrast by FCM 5 (11.1%) had positive MRD (〉0.1%) on D33 compared with 7 (15.6%) on D79. The 3 with high MRD on RQ-PCR at D79 were also identified by FCM. 2 of these 3 patients had been classified as standard/medium risk by conventional criteria. These two techniques may be complementary in identifying patients with significant MRD. Only a proportion of patients were identified by both methods. Long term follow up and ongoing recruitment will help delineate the optimal surveillance/stratification strategy by correlating these findings with outcome.