Comparative analysis between cytomorphology and flow cytometry methods in central nervous system infiltration assessment in oncohematological patients

ABSTRACT Introduction Oncohematological patients require the evaluation for possible infiltration of the central nervous system (CNS) by neoplastic cells at diagnosis and/or during the monitoring of the chemotherapeutic treatment. Morphological analysis using conventional microscopy is considered the method of choice to evaluate the cerebrospinal fluid (CSF) samples, despite technical limitations. Objective This study aimed to compare the performance of the cytomorphology and flow cytometric immunophenotyping (FC) in the detection of CNS infiltration. Method We evaluated 520 CSF samples collected from 287 oncohematological patients for whom the detection of neoplastic cells was simultaneously requested by cytomorphology and FC. Results Laboratory analyses revealed 435/520 (83.7%) conclusive results by the two methods evaluated, among which 385 (88.5%) were concordant. Discordance between the methods was observed in 50/435 (11.5%) samples, 45 (90%) being positive by FC. Furthermore, the FC defined the results in 69/72 (95.8%) inconclusive samples by cytomorphology. The positivity of FC was particularly higher among hypocellular samples. Among 431 samples with a cell count of < 5/μL, the FC identified neoplastic cells in 75 (17.4%), while the cytomorphology reported positive results in 26 (6%). Among the samples that presented adequate cell recovery for evaluation by both methods (506/520), the comparative analysis between FC and cytomorphology revealed a Kappa coefficient of 0.45 (CI: 0.37-0.52), interpreted as a moderate agreement. Conclusion The data showed that the CSF analysis by FC helps in the definition of CNS infiltration by neoplastic cells, particularly in the cases with dubious morphological analysis or in the evaluation of samples with low cellularity.

Incidence and risk factors for central nervous system relapse in children and adolescents with acute lymphoblastic leukemia

BACKGROUND: Despite all the advances in the treatment of childhood acute lymphoblastic leukemia, central nervous system relapse remains an important obstacle to curing these patients. This study analyzed the incidence of central nervous system relapse and the risk factors for its occurrence in children and adolescents with acute lymphoblastic leukemia. METHODS: This study has a retrospective cohort design. The studied population comprised 199 children and adolescents with a diagnosis of acute lymphoblastic leukemia followed up at Hospital das Clinicas, Universidade Federal de Minas Gerais (HC-UFMG) between March 2001 and August 2009 and submitted to the Grupo Brasileiro de Tratamento de Leucemia da Infância - acute lymphoblastic leukemia (GBTLI-LLA-99) treatment protocol. RESULTS: The estimated probabilities of overall survival and event free survival at 5 years were 69.5% ( 3.6%) and 58.8% ( 4.0%), respectively. The cumulative incidence of central nervous system (isolated or combined) relapse was 11.0% at 8 years. The estimated rate of isolated central nervous system relapse at 8 years was 6.8%. In patients with a blood leukocyte count at diagnosis > 50 x 10(9)/L, the estimated rate of isolated or combined central nervous system relapse was higher than in the group with a count < 50 x 10(9)/L (p-value = 0.0008). There was no difference in cumulative central nervous system relapse (isolated or combined) for the other analyzed variables: immunophenotype, traumatic lumbar puncture, interval between diagnosis and first lumbar puncture and place where the procedure was performed. CONCLUSIONS: These results suggest that a leukocyte count > 50 x 10(9)/L at diagnosis seems to be a significant prognostic factor for a higher incidence of central nervous system relapse in childhood acute lymphoblastic leukemia.