Flow Cytometric Leukemic Blasts Detection in Cerebrospinal Fuid of Children with Acute Leukemias

Abstract 1449

Central nervous system (CNS) involvement is one of the important risk factors in childhood acute leukemia (AL). Tumor cells detection in cerebrospinal fluid (CSF) is one of the main signs of CNS lesion. Traditionally blasts presence in CSF is assessed by conventional cytomorphology (CM) of cytospin slides. However, sensitivity of this method is relatively low. Flow cytometry (FC) having a higher sensitivity could provide better diagnostic applicability for CSF blasts detection. Aim. To compare results of tumor cells detection in CSF of children with AL by flow FC and CM. Methods. 183 samples from 52 boys and 31 girls aged from 5 months to 15 years with different types of acute lymphoblastic leukemia (ALL) (77 patients), acute myeloid leukemia (AML) (5 patients) and acute biphenotypic leukemia (1 patient) were investigated. 17 positive samples obtained by traumatic lumbar puncture were excluded from analysis because tumor blasts were also detected in peripheral blood. Comparison between FC and CM data was performed in 166 samples. Among these samples 61 was taken at the time of initial diagnostics, 34 – during AL follow-up, 17 – at relapse and 54 – during relapse monitoring. Monoclonal antibodies panels were constructed according to immunophenotype of tumor cells in bone marrow. Results. In 24 out of 166 samples (14.5%) tumor cells were detected by CM. In all these cases blasts were also found by FC, while FC allowed finding blasts in other 35 samples. Thus the total number of FC-positive samples was 59 out of 166 (35.5%). This frequency was significantly higher than rate of CM-positive cases (g < 0.0001). Among initial diagnostics samples there were 20 FC-positive and only 10 CM-positive patients (32.8% vs. 16.1%, p=0.0585). At relapse 9 (52.9%) patients were FC-positive, while 6 (35.3%) were CM-positive (p=0.4897). In both B-lineage and T-lineage ALL, analyzed separately, FC detected blasts in CSF frequently than CM (p=0.0098 and p=0.0002 respectively). Absolute blast count in 1 ml in CSF samples, positive by both methods was significantly higher than in samples, positive only by FC (median = 418, range 8–158171 and median = 34, range 5–2762 respectively, g = 0.0002). Thus FC allows detecting tumor cells in CSF much more frequently than conventional CM, which could be explained mainly by higher FC sensitivity. Moreover FC is applicable also for qualitative and quantitative monitoring of CNS lesion. Nevertheless prognostic impact of FC CSF investigation is questionable. Among 13 patients in whom discordant results were obtained in initial diagnostics samples and at relapse, only for one patient risk stratification could have been changed. For all other patients there were other risk factors, that decreased significance of FC leukemic blast detection in CSF. Conclusion. Flow cytometry allows more frequent detection of tumor blasts in CSF of children with AL, while prognostic significance of these findings is still unclear and needs to be confirmed in large prospective trials.