3D CNN-based Deep Learning Model-based Explanatory Prognostication in Patients with Multiple Myeloma using Whole-body MRI.

Although magnetic resonance imaging (MRI) data of patients with multiple myeloma (MM) are used to predict prognosis, few reports have applied artificial intelligence (AI) techniques for this purpose. We aimed to analyze whole-body diffusion-weighted MRI data using three-dimensional (3D) convolutional neural networks (CNNs) and Gradient-weighted Class Activation Mapping (Grad-CAM), an explainable AI, to predict prognosis and explore the factors involved in prediction. We retrospectively analyzed the MRI data of a total of 142 patients with MM obtained from two medical centers. We defined the occurrence of progressive disease after MRI evaluation within 12 months as a poor prognosis and constructed a 3D CNN-based deep learning model to predict prognosis. Images from 111 cases were used as the training and internal validation data; images from 31 cases were used as the external validation data. Internal validation of the AI model with stratified 5-fold cross-validation resulted in a significant difference in progression-free survival (PFS) between good and poor prognostic cases (2-year PFS, 91.2% versus [vs.] 61.1%, P = 0.0002). The AI model clearly stratified good and poor prognostic cases in the external validation cohort (2-year PFS, 92.9% vs. 55.6%, P = 0.004), with an area under the receiver operating characteristic curve of 0.804. According to Grad-CAM, the MRI signals of the spleen and bones of the vertebrae and pelvis contributed to prognosis prediction. This study is the first to show that image analysis of whole-body MRI using a 3D CNN without any other clinical data is effective in predicting the prognosis of patients with MM.

Loss of or decrease in CD30 expression in four patients with anaplastic large cell lymphoma after brentuximab vedotin-containing therapy.

Brentuximab vedotin (BV), CD30 specific antibody drug conjugate, has been used to treat anaplastic large cell lymphoma (ALCL) and classic Hodgkin lymphoma (CHL); it is also used in the treatment of other CD30-positive peripheral T-cell lymphomas. We aimed to investigate the incidence and clinicopathological characteristics of patients with ALCL or CHL with loss of or decrease in CD30 expression after BV-containing therapy. Twelve and nine patients with refractory/relapsed CHL and ALCL, respectively, were analyzed after receiving BV-containing therapy. In four ALCL patients (44%), CD30 expression was lost/decreased in re-biopsy materials, including one with complete loss and three with a reduction of less than 20%. All 12 CHL patients showed consistent CD30 expression levels after BV treatment. Compared with five ALCL patients with consistent CD30 expression, four ALCL patients with a loss of/decrease in CD30 expression received a higher cumulative dose of BV (P = 0.014) and revealed a lower intensity of CD30 expression in initial biopsy materials (P = 0.017). The subtypes of ALCL (ALK positive, ALK negative, and primary cutaneous) were not related to the loss of/decrease in CD30 expression. In conclusion, 44% of ALCL patients, regardless of histological subtypes, showed a loss of/decrease in CD30 expression after receiving BV-containing therapy, but this phenomenon was not observed in CHL patients. A higher cumulative dose of BV and a lower amount of CD30 antigen in tumor cells in the initial biopsy materials might be predictors of a loss of/decrease in CD30 expression in ALCL patients.

Heterogeneity or change in cell of origin in diffuse large B-cell lymphomas determined using hans algorithm.

This study aimed to analyze the heterogeneity or change in cell of origin (COO) in diffuse large B-cell lymphoma (DLBCLs) using the Hans algorithm including 156 patients with multiple DLBCL specimens. COO was detected via immunohistochemical staining for CD10, BCL6, and MUM1. The COO of the main tumor at initial diagnosis was germinal center B-cell (GCB) and non-GCB type in 50 (32%) and 106 (68%) patients, respectively. It did not change in 126 patients (81%). However, it changed in 30 patients (19%), from GCB to non-GCB in 12 patients and vice versa in 18 patients. The COO was heterogeneous or changed in 14% of simultaneous samples at other sites during the initial diagnosis, in 7% of primary refractory sites, and in 20% of samples obtained in the relapse phase other than the primary site. Changes in CD10, BCL6, and MUM1 expression were observed in 15%, 23%, and 24% samples, respectively. A low incidence of change in COO was observed in DLBCL with CD10+/BCL6+/MUM1- (4%), CD10-/BCL6-/MUM1+ (3%), and CD10-/BCL6-/MUM1- (0%) patterns, whereas DLBCL with other patterns showed COO changes at rates of 20-37%. In conclusion, COO was heterogeneous or changed in 19% of DLBCL cases. The COO should be re-examined in other biopsy samples to determine the optimal treatment.

Clinicopathological differences between T-lymphoblastic leukemia/lymphoma, early T-precursor lymphoblastic leukemia/lymphoma, and mixed-phenotype acute leukemia with T lineage: An analysis of 41 adult cases.

The histopathological diagnosis of T-lymphoblastic leukemia/lymphoma, NOS (T-ALL), is based on morphology and positivity for CD3 and TdT. Early T-precursor lymphoblastic leukemia/lymphoma (ETP-ALL) and mixed-phenotype acute leukemia (MPAL), T/M, and/or B rarely occur and are usually diagnosed using flow cytometry. Using only formalin-fixed paraffin-embedded tissue raises the risk of misdiagnosis due to underestimation. Immunostaining markers for T cell (CD1a, CD4, CD5, CD8), B cell (CD19, CD10, CD22, CD79a), and stem/myeloid-related cell (CD33, CD34, CD117, MPO, lysozyme) diagnosed 25 T-ALL cases (61%), 7 MPAL (17%), 6 ETP-ALL (15%), and 3 near ETP-ALL (7%), with subsequent analysis of their clinicopathological characteristics. Patients with MPAL had significantly poorer 2-year progression-free survival (14.3% vs. 60.4%, P = 0.012) and 5-year overall survival (28.6% vs. 65.9%, P = 0.011) than did those with T-ALL, whereas ETP-ALL and near ETP-ALL did not. Of the seven patients with MPAL, three were classified as T/B, two as T/M, and two as T/M/B. Because most MPALs (6/7) share the ETP-ALL phenotype, immunohistochemistry for CD19 and MPO should be performed to avoid misdiagnosing MPAL as ETP-ALL. All three patients with TdT-negative MPAL died of the disease. Four patients with MPO-positive MPAL relapsed during the early phase (1-9 months). Five patients received the ALL regimen, but two patients received acute myeloid leukemia and lymphoma regimens, respectively. In this study, MPAL exhibited a poorer prognosis compared to T-ALL, unlike ETP-ALL. Thus, immunohistochemical classification with multiple antibody panels is useful for accurate diagnosis and treatment.