Automated prediction of acute promyelocytic leukemia from flow cytometry data using a graph neural network pipeline.

OBJECTIVES: Our study aimed to develop a machine learning (ML) model to accurately classify acute promyelocytic leukemia (APL) from other types of acute myeloid leukemia (other AML) using multicolor flow cytometry (MFC) data. Multicolor flow cytometry is used to determine immunophenotypes that serve as disease signatures for diagnosis. METHODS: We used a data set of MFC files from 27 patients with APL and 41 patients with other AML, including those with uncommon immunophenotypes. Our ML pipeline involved training a graph neural network (GNN) to output graph-level labels and identifying the most crucial MFC parameters and cells for predictions using an input perturbation method. RESULTS: The top-performing GNN achieved 100% accuracy on the training/validation and test sets on classifying APL from other AML and used MFC parameters similarly to expert pathologists. Pipeline performance is amenable to use in a clinical decision support system, and our deep learning architecture readily enables prediction explanations. CONCLUSIONS: Our ML pipeline shows robust performance on predicting APL and could be used to screen for APL using MFC data. It also allowed for intuitive interrogation of the model's predictions by clinicians.

Leukocyte immunoglobulin-like receptor B1 and B4 (LILRB1 and LILRB4): Highly sensitive and specific markers of acute myeloid leukemia with monocytic differentiation.

BACKGROUND: Acute myeloid leukemia (AML) with monocytic differentiation (M-AML) remains a diagnostic challenge largely due to lack of sensitive and specific markers for immature monocytes. The immunoglobulin-like inhibitory receptors, LILRB1 and LILRB4, are expressed on monocytes but have not yet been systematically evaluated in the clinical setting. METHODS: We evaluated the diagnostic performance of LILRB1 and LILRB4 as monocytic markers for both immature and mature monocytes in comparison to other myelomonocytic markers including CD14, CD15, CD33, CD36, and CD64 in eight cases of control bone marrow (BM, 5) and peripheral blood (PB, 3), 64 cases of (M-AML), and 57 cases of AML without monocytic differentiation (NM-AML) by flow cytometric immunophenotyping. RESULTS: In control BM, LILRB1 and LILRB4 were consistently expressed on monocytes at all stages of maturation, from CD34+ /CD14- monocytic precursors to CD14-/dim+ maturing and CD14+ mature monocytes. In M-AML, LILRB1 and LILRB4 were consistently expressed on monocytes, regardless of the degree of maturity, from CD14-/dim+ monoblasts/promonocytes to CD14+ mature monocytes but were not expressed on myeloblasts. The diagnostic performances as a monocytic marker assessed by sensitivity/specificity were 100%/100% for LILRB1/LILRB4, 100%/82% for CD11b, 80%/100% for CD14, 100%/81% for CD64, 100%/58% for CD15/CD33, and 89%/97% for CD36/CD64. CONCLUSION: The co-expression of LILRB1/LILRB4 outperformed other myelomonocytic markers as a highly sensitive and specific marker for monocytes at all stages of maturation and could reliably distinguish M-AML from NM-AML. LILRB4 additionally represents a novel therapeutic target for treating M-AML.

Lenalidomide-Associated Secondary B-Lymphoblastic Leukemia/Lymphoma-A Unique Entity.

OBJECTIVES: Autologous stem cell transplant with lenalidomide maintenance therapy has greatly improved the relapse-free and overall survival rates of patients with multiple myeloma but also has been associated with an increased risk of secondary B-lymphoblastic leukemia/lymphoma (B-ALL). METHODS: We report a comprehensive review of the clinicopathologic features of 2 patients with multiple myeloma who developed secondary B-ALL during lenalidomide maintenance. RESULTS: Our observations showed that the disease may initially present with subtle clinical, morphologic, and flow-cytometric findings. The flow cytometry findings in such cases may initially mimic an expansion of hematogones with minimal immunophenotypic variation. Both patients achieved complete remission of secondary B-ALL after standard chemotherapy; however, one patient continues to have minimal residual disease, and the other experienced relapse. Next-generation sequencing of the relapse specimen showed numerous, complex abnormalities, suggesting clonal evolution. CONCLUSIONS: Our findings suggest the need for increased awareness and further study of this unique form of secondary B-ALL.

Novel r(2)(p25q31) cytogenetic abnormality in a pediatric patient with acute leukemia of ambiguous lineage.

We describe a case of acute leukemia of ambiguous lineage with a novel cytogenetic abnormality. A 1-year-old boy presented with abnormal complete blood count findings, and was found to have blasts and mild dysgranulopoiesis. The blasts showed immunophenotypic evidence of myeloid and T-lineage differentiation. Subsequent cytogenetic analysis showed r(2)(p25q31) as the sole stem line cytogenetic defect with clonal evolution. While cytogenetic abnormalities can have a critical role in the classification and prognostication of acute lymphoblastic and acute myeloid leukemia, the significance of cytogenetic abnormalities in acute leukemia of ambiguous lineage remains unclear. This finding has not been reported previously to the best of our knowledge.

Notch1 in primary effusion lymphoma: a clinicopathological study.

Primary effusion lymphoma is a human herpes virus 8 (HHV-8)-associated large cell lymphoma of body cavities. Detailed large-scale clinicopathological studies are rarely reported, and the underlying mechanism of lymphomagenesis remains elusive. In the present report, we studied the clinicodemographic, immunophenotypic, and cytomorphological features on a cohort of 12 cases of primary effusion lymphoma. In contrast to HHV-8, which was positive in all nine cases tested (100%), HIV was found in 75% (9/12) of cases, whereas the three HIV-negative cases were either in elderly patients (one with hepatitis C virus infection and one with asbestoses exposure) or in a heart transplantation recipient. By flow cytometry, the antigens expressed in descending order were CD38, CD71, HLA-DR, CD30, and CD45RO. B-cell markers were largely negative. Cytomorphologically, all cases showed atypical to anaplastic morphology. Notch1, a member of transmembrane signal transduction family, was found in six of seven HHV-8-positive cases (86%). In agreement with in vitro studies using human primary effusion lymphoma cell lines, we have found that Notch1 was expressed in the majority of HHV-8-positive primary effusion lymphoma cases, corroborating the notion that Notch1 may have an important role in HHV-8-mediated lymphomagenesis of primary effusion lymphoma.

Significant CD5 expression on normal stage 3 hematogones and mature B Lymphocytes in bone marrow.

B-cell maturation from hematogones to mature B cells in bone marrow exhibits a consistent, complex spectrum of sequential antigen expression. CD5 expression, however, has not been characterized. We studied the dynamics of CD5 expression on developing B cells by 4-color flow cytometry in 32 patients, aged 9 months to 63 years, with hematogone hyperplasia (>3.5% of total events). The mean percentage of hematogones was 8.1%. We demonstrate consistent CD5 expression on normal, polytypic B cells in a continuum, predominantly at later stages of maturation, specifically on stage 3 hematogones and mature B cells. Awareness of this normal pattern of CD5 expression on B-cell subsets has implications in the analysis of minimal residual disease of CD5+ B-lineage non-Hodgkin lymphomas.

Multiparameter flow cytometric analysis reveals low percentage of bone marrow hematogones in myelodysplastic syndromes.

Diagnosis of myelodysplastic syndromes (MDS) could be difficult. We explored the usefulness of the enumeration of maturing B-lineage precursors (hematogones) by multiparameter flow cytometric analysis in the diagnosis of MDS in bone marrow (BM) specimens. We evaluated 111 MDS, 120 non-MDS (most with cytopenias; control group 1), and 41 noncytopenic lymphoma staging BM (control group 2) specimens. The percentage of total hematogones was significantly lower in MDS (median, 0%; mean, 0.10%) compared with non-MDS (control group 1, median, 0.38%, and mean, 0.91%; control group 2, median, 0.38%, and mean, 0.60%; P < .0001), as was the percentage of the most immature (stage I) hematogones. Thus, hematogone enumeration may serve as a biomarker to aid in the diagnosis of MDS. Interestingly, the percentage of hematogones was not significantly different between MDS subgroups or patients with MDS with and without chromosomal abnormalities, implying that a defect in maturing B-cell precursors may be an early event in the pathogenesis of MDS.