Peripheral blood flow cytometry for the diagnosis of pediatric acute leukemia: Highly reliable with rare exceptions

Background Recent data have demonstrated the high sensitivity and specificity of peripheral blood flow cytometry (PBFC) for the diagnosis of pediatric leukemia; however, diagnostically significant immunophenotypic discrepancies between PBFC and bone marrow (BM) evaluation, which result in different lineage assignment and treatment protocols, can rarely occur. Here, we sought to further characterize the performance of PBFC for pediatric leukemia and highlight the exceptions when PBFC can result in misdiagnosis. Methods Results An institutional database was searched between 2012 and 2016 for cases of acute leukemia with concurrent PBFC and BM evaluation. Immunophenotyping results from the peripheral blood and BM using four or eight color flow cytometry, as well as BM cytochemical staining and immunohistochemistry, were compared. Two hundred ninety PBFC samples with concurrent BM evaluation were identified. Based on the final immunophenotypic classification, the cases were distributed as follows: 108 B-lymphoblastic leukemia (B-ALL), 57 T-lymphoblastic leukemia (T-ALL), 116 acute myeloid leukemia (AML), and 9 mixed-phenotype acute leukemia (MPAL). Among all cases, five had a diagnostically significant discrepancy between PBFC and BM evaluation. In three cases, the immunophenotype by PBFC was consistent with early T-cell precursor ALL (ETP-ALL), whereas BM evaluation demonstrated MPAL. Two cases were suspicious for acute megakaryoblastic leukemia (AMKL) and MPAL, T/myeloid by PBFC but were diagnosed as B-ALL and T-ALL in the BM. Conclusion Immunophenotypic classification by PBFC is accurate (>98%) in almost all cases of pediatric leukemia with the rare exceptions of suspected ETP-ALL, MPAL, and AMKL. These PBFC diagnoses should be confirmed with BM immunophenotyping.