Seheult, Jansen N

Background COVID‐19 and other respiratory viral infections can cause cardiovascular complications. SARS‐CoV‐2 virions are found in the blood, and circulating viral RNA levels are associated with death. We hypothesized that viremia can induce thrombotic endotheliopathy that contributes to death and studied this relationship in patients hospitalized for COVID‐19 and enrolled in the ACTIV‐4a (Accelerating COVID‐19 Therapeutic Interventions and Vaccines) randomized trial of antithrombotic therapy. Methods We quantified SARS‐CoV‐2 nucleocapsid RNA and protein in plasma and measured their associations with clinical outcomes and biomarkers of thromboinflammation and endotheliopathy. We used Cox regression and Fine–Gray competing risk models to analyze survival and thrombosis. We conducted causal mediation analysis to explore whether thrombotic endotheliopathy mediates the relationship between viral RNA and death. Results In 93 patients, SARS‐CoV‐2 RNA and N‐antigen were higher in nonsurvivors. Baseline viral RNA levels were associated with increased 90‐day death (hazard ratio [HR], 1.31 [95% CI, 1.17–1.46]) and thrombosis (HR, 1.35 [95% CI, 1.24–1.47]). Viral RNA more effectively predicted survivorship than N‐antigen levels. Soluble thrombomodulin, a biomarker of thrombotic endotheliopathy, positively correlated with viral RNA levels. Mediation analysis revealed that soluble thrombomodulin accounts for 12.2% (95% CI, 0.1%–32.3%; P =0.048) after adjustment for age and sex of the relationship between viral RNA and the 90‐day mortality rate. Conclusions Elevated plasma SARS‐CoV‐2 RNA levels are associated with death in ACTIV‐4a, which is causally mediated in part by soluble thrombomodulin. We propose that lung–blood viral dissemination is a potential mechanism for cardiovascular complications of respiratory viruses. Registration URL: https://www.clinicaltrials.gov ; Unique Identifier: NCT04505774.

Abstract Background Substituting group A whole blood (GAWB) might relieve some of the burden on the low titer group O whole blood (LTOWB) inventory. This simulation examined the effect of implementing GAWB at a large trauma hospital. Study Design and Methods A civilian trauma center with LTOWB and GAWB par levels set to 22 and 10 units, respectively, was modeled as the baseline. The average daily number of WB recipients was 0.5 who received an average of four WB units per patient. LTOWB was issued in packages of five and all had to be consumed before a group A patient could be switched to GAWB. Several variations on the baseline simulation were also modeled. Results In the baseline simulation, there was a median (interquartile range, IQR) of 1 (0–2) GAWB and 16 (13–19) LTOWB recipients per month. The median (IQR) GAWB and LTOWB units transfused per month was 2 (0–9) and 56 (42–72), respectively. Fourteen percent of group A patients received any quantity of GAWB. The variations that had the most impact on GAWB usage included increasing the number of WB recipients to 2.5 per day where the median (IQR) monthly number of GAWB recipients and units transfused increased to 4 (3–5) recipients and 21 (12–29) units, respectively, and when the average WB requirements increased to eight units per patient resulting in transfusing a median (IQR) of 10 (5–15) GAWB units per month to 3 (1–4) recipients. Discussion Under routine civilian hospital circumstances, implementing GAWB will have a modest impact on LTOWB inventory.

Abstract Artificial intelligence (AI) and machine learning (ML) are beginning to enhance key workflows across blood banking (BB) and transfusion medicine (TM), including donor engagement, inventory forecasting, immunohaematology interpretation, haemovigilance and clinical decision support. These tools can extract signals from large, diverse datasets to improve efficiency, consistency and timeliness of care while enabling more proactive patient blood management. At the same time, TM is uniquely sensitive to safety and trust, so the field must pair innovation with disciplined validation and governance. This review surveys emerging applications and distils a balanced ethics and oversight framework tailored to TM and BB practice. Core elements include outcome‐focused validation and continuous monitoring; targeted fairness testing where biological or scientific plausibility suggests subgroup differences; layered transparency that distinguishes transparency, interpretability and explainability; and privacy by design with proportionate consent. We emphasize human oversight at every step: clinicians and laboratorians evaluate, monitor and troubleshoot models, perform tasks AI cannot, integrate multiple inputs to make decisions and interact with AI through guardrails such as intended‐use statements, confidence thresholds and clear override pathways. Accountability is shared across users, service directors, institutions and regulators. When implemented with these safeguards, AI can reduce preventable errors, improve resource allocation and support more equitable access to safe transfusion. The goal is principled pragmatism that protects patients and donors while enabling timely adoption of well‐validated tools.

Abstract Background: Waldenström Macroglobulinemia (WM) is an indolent B-cell lymphoplasmacytic lymphoma characterized by heterogeneous clinical behavior and variable progression patterns. While MYD88L265Pand CXCR4 mutations are frequently encountered, the broader genomic landscape underpinning disease evolution from smoldering WM (SWM) to active treatment naïve WM (TNWM), post-treatment remission WM (PTWM) and relapsed/refractory WM (RRWM) remains incompletely understood. To address this gap, we studied the mutational profile across the WM disease spectrum. Methods We performed whole exome sequencing (WES) on sorted (CD19 and/or CD138) bone marrow aspirates of patients across the spectrum of WM. Paired germline sequencing was performed using peripheral blood mononuclear cells (PBMCs). Unique molecular identifier-tagged reads were deduplicated using fgbio. WES data were processed through the Sarek pipeline. Any valid mutation from either caller, present on the index or in COSMIC was considered. Copy number variations (CNVs) were inferred via CNVkit and summarized at the gene and arm level (≥50% directional change). Mean tumor and germline coverages were 174X and 74X, respectively. Fisher's exact test was used for enrichment and hypergeometric test for overlap analyses. Results A total of 78 high-quality WM samples [15 SWM, 44 TNWM, 13 RRWM, 6 PTWM in remission (all with paired TNWM samples)] were sequenced. The mutational burden varied across disease states (median SWM=1.48 m/Mb, TNWM= 1.12 m/Mb, RRWM = 1.59 m/Mb; p=0.04). Density plots revealed bimodal VAF distributions, consistent with clonal and subclonal architectures in WM. Overall, 77% harbored a MYD88L265P mutation by WES. Among MYD88^WT^ patients, AS-PCR identified MYD88L265P in 4 (25%) not detected by WES, raising the overall mutation rate to 83%, plus one with a non-L265P variant. Low VAF (<10%) MYD88L265P mutations were seen in 6% of matched PBMCs. MYD88 copy number gain was seen in 2 (17%) MYD88WT TNWM patients. CXCR4 mutations were detected in 38% by WES (18% each frameshift and nonsense). Two (4%) CXCR4WT samples were positive by PCR, giving an overall rate of 40%. CXCR4 mutations were more frequent in RRWM (54%) vs TNWM (41%) and SWM (27%), though not statistically significant. Del(6q) and Del(17p) were noted in 16% (n=12) and 7% (n=5) of patient samples. All del(17p) events were noted in TNWM samples and in one paired post-treatment sample. Overall rate of del(17p) and TP53 mutation was 14% (n=11). RRWM was enriched in ARID1A alterations, a gene implicated in chromatin remodeling and plasma cell identity formation [38% (n=5) in RRWM vs. 11% (n=6) in TNWM and SWM], compared to TNWM and SWM p=0.03]. Alterations in NF-kB signaling gene MAP3K14 were more frequent in RRWM (31%, n=4) compared to the SWM/TNWM cohort [(7%), n=4; p=0.036). We noted an enrichment in the TP53 mutation rate in RRMW [n=3, 23%, OR 7.7] compared to TNWM and SWM cohorts [n=2, 4%), p=0.04]. Mutations in FUBP1 (p=0.03), an upstream regulator of MYC, and CARD11 (p=.043), a B-cell signaling scaffold protein were also more frequent in RRWM. Notably, none of the SWM patients harbored an ARID1A alteration (p=0.057), suggesting acquisition during progression. CNV analysis revealed that MAP3K14 amplifications were frequent in RRWM (n=3, 23%) compared to SWM/TNWM [n=2 (4%), p=0.043]. RRWM samples showed increased CD79B mutations (15%) and significant overlap with the C5/ABC diffuse large B-cell lymphoma (DLBCL) genomic cluster (Chapuy et al., Nat. Med. 2018; p=0.007). This overlap remained significant after excluding MYD88L265P (p=0.019), driven by CARD11, CD79B, and IGLL5, suggesting that a subset of RRWM may acquire an ABC DLBCL-like mutational profile. These findings warrant further investigation using transcriptional profiling. Conclusions This study provides a genomic map of WM evolution from precursor to relapse. Increasing somatic mutations and CNVs support stepwise progression. Enrichment of genes involved in chromatin remodeling and plasma cell identity formation (ARID1A), NF-kB or BCR signaling (CARD11, MAP3K14), and DNA damage repair (TP53) were noted at relapse, all of which have been shown to drive aggressive disease biology. Emergence of an ABC-like genotype in a proportion of RRWM indicates convergent evolution with aggressive lymphomas. Transcriptional studies are warranted to explore this further.

Abstract Background: Flow cytometry (FC) is widely utilized for the identification of measurable residual disease (MRD) in multiple myeloma (MM), providing an essential laboratory indicator for prognostic assessment and therapeutic management. Unfortunately, the high level of expertise required, long manual analysis time, and complexity of FC data files currently limit the offering of this testing to selected reference laboratories. Methods: We developed an artificial intelligence (AI) pipeline to automatically enhance unaltered FC files, minimizing the expertise and manual analysis time needed to detect MM MRD using the 2-tube Euroflow Consortium Panel (Cytognos). Raw FC files corresponding to 150 MRD-positive bone marrow samples (MRD < 5%, median = 0.0022%), and 39 MRD-negative samples were processed through our CCADDAS (Clustering and Classification of All events, Dimensionality reduction, Downsampling and Aberrancy Scaling) pipeline, deployed on a custom clinical-grade cloud environment (Biolegend). Automated processing steps included elimination of acquisition errors (FlowCut), state-of-the-art clustering (PARC), cluster-informed and density-based downsampling with preservation of low-event subsets, dimensionality reduction (UMAP), and cluster-based anomaly detection compared to negative controls (29 bone marrow aspirates). In addition, a deep neural network (DNN) trained on expert-defined subpopulations from negative controls was included for automatic gating of normal subsets (Tensorflow v2). AI-enhanced data files were analyzed using a general-purpose flow cytometry analysis software (Kaluza v 2.3, Beckman Coulter), and %MRD estimates were compared to reported results based on traditional expert analysis of original FC files. Results: Cluster-informed downsampling reduced the number of cells to be manually analyzed from 4.1 million to 196,761 cells per tube, on average (95% cellularity reduction); resulting in a smaller FC data file (from 203.2 MB to 20.5 MB; 90% data reduction). Importantly, low-level MRD events were adequately preserved after downsampling (median 88% and 77% retention, for 0.01-0.001% MRD and < 0.001% MRD, respectively). True number of events were accurately estimated on gated subsets using an “upsampling factor” parameter [observed # events x mean (upsampling factor)]. Gating of normal subsets was completely automated using a DNN classifier parameter. In 96% of positive cases, MRD was rapidly identified as immunoglobulin light chain-restricted plasma cells on a single pre-calculated UMAP plot (tube 2), after dual coloring for kappa and lambda. On the remainder 4% of positive cases, MRD was easily segregated from benign plasma cells using the AI-generated “aberrancy scale” parameter or other strategies. The use of AI-enhanced files reduced manual analysis time from 15.05 minutes (SD = 6.84) to 1.29 minutes (SD = 0.90) per case, on average (p < 0.0001) (91% reduction of manual analysis time). MM MRD was detected in all positive cases above limit of detection for our assay (≥0.0002%), with excellent quantitative correlation with conventional analysis on original FC files (linear regression slope = 0.9841, P < 0.0001). Conclusion: We introduce a largely unsupervised AI pipeline that transforms raw MM MRD FC data from the Euroflow Consortium panel into a markedly smaller, AI-annotated and software-agnostic FC file, including comparison to normal controls. This CCADDAS pipeline simplifies and accelerates detection of MM MRD in clinical diagnostics, reducing the number of cells analyzed by 95% and manual analysis time by 91%, without impacting test performance. Moreover, CCADDAS can be utilized for any laboratory-developed MM MRD assay, and its small-sized export is compatible with any clinical FC software, computing platform, and analysis strategy. Adoption of CCADDAS is likely to facilitate the implementation of MM MRD FC analysis by more clinical laboratories.

Abstract Background Extramedullary disease (EMD) in multiple myeloma (MM) represents an aggressive phenotype with poor outcomes. We performed single-cell RNA sequencing (scRNA-seq) on clonal plasma cells (CPCs) from EMD tissue, solitary plasmacytomas (SPM), and bone marrow aspirates (BMAs) from patients with MM to identify transcriptional states unique to EMD. Methods We processed FFPE tumor samples from EMD and SPMs and cryopreserved CD138+ BMAs from MM, using the 10x Genomics FLEX platform. Cells/nuclei were fixed, hybridized with Human WTA probes, pooled, and sequenced (10,000–40,000 cells/library) on NovaSeq X. After removing low quality cells, data was normalized and integrated using RPCA in Seurat. Plasma cell identity was assigned via Cellarium and canonical markers. Clonality was inferred by light chain restrictions and refined via random forest modeling. Cell cycle effects were regressed out to reduce proliferation bias. Results We analyzed 128, 908 CPCs from 16 samples [8 EMD (62, 669 CPCs), 4 SPM (20, 607 CPCs), 4 BMA (45, 632 CPCs)] and identified eight transcriptionally distinct cell clusters. Differential expression and pathway comparisons were performed between EMD/SPM/BMA within each of these clusters. Integration of gene expression and pathway analyses revealed consistent upregulation of cell cycle and E2F pathways in EMD. We identified two highly proliferative cell cycle clusters (Cluster 2 with S-phase genes and Cluster 4 with a G2/M transcriptional program) that were expanded in EMD compared to BMA and SPM. Cluster 0 comprised a quiescent CPC population enriched for RNA processing pathways alongside downregulation of ER stress and autophagy. EMD cells demonstrated suppressed interleukin signaling [fold change (FC) –33.8), and class I MHC antigen processing (FC –23.5) relative to BMA, suggestive of an immune suppressive transcriptional program. Angiogenesis-related genes [VEGFA (log2FC 2.4), ADM2 (log2FC 2.2)] and EZH2 [log2FC 2.1; q<0.0001)] were upregulated in EMD. Compared to SPM, EMD plasma cells in this cluster demonstrated marked downregulation of CD74 (log2FC –2.6; q<0.0001), a key modulator of MHC class II antigen presentation. Cluster 1 comprised of metabolically stressed secretory CPCs and displayed downregulation of early growth response pathways in EMD compared to BMA, concurrent with upregulation of E2F signaling. Suppression of IL-2/STAT5, IL-6/JAK/STAT3, and TGF-β pathways in EMD suggests reduced reliance on bone marrow–derived cytokines and impaired immunoregulatory signaling. CD44 was upregulated in EMD versus BMA (log2FC 3.2), consistent with its role in EMD spread and marrow independence. Cluster 3 was enriched for extracellular matrix–interacting plasma cells (COL1A1, MMP2) with upregulated epithelial–mesenchymal transition (EMT) pathways in EMD compared to both BMA (FC 7.5) and SPM (FC 7.07). Cluster 6 was defined by a migratory, motility-primed plasma cell population (CD44, AHNAK, LMNA and S100A10). Notably, IL32 was overexpressed in EMD compared to BMA (log2FC 5.1; adj. p = 2.02×10⁻¹⁴), a finding linked to hypoxia-driven proliferation. Compared to SPM, EMD cells also had elevated MYC/E2F activity and DNA repair, hallmarks of a proliferative phenotype. Cluster 5 displayed a strong interferon-stimulated gene signature (ISG15, IFIT1, TNFSF10), with enrichment of type I IFN, JAK–STAT, and TNF pathways. Within this inflammatory cluster, EMD exhibited marked suppression of immune signaling, metabolism (TCA, OXPHOS), and TP53 programs, consistent with an immune-evasive, metabolically suppressed, and deregulated state. Finally, within Cluster 7, a smaller subset defined by NFκB–driven anti-apoptotic transcriptional activity, EMD exhibited significant downregulation of KLF2 (log2FC –3.7; adj. p = 1.77×10⁻²⁷), likely reflecting disruption of the KDM3A–KLF2–IRF4 axis, a transcriptional network critical for bone marrow homing and plasma cell survival. We consistently observed downregulation of the MAPK pathway in EMD compared to BMA, aligning with prior bulk RNA-seq data and contrasting with the near-universal presence of MAPK mutations, highlighting a disconnect between mutational status and pathway activation. Conclusion In the largest yet reported single-cell transcriptional atlas of EMD and SPM, we demonstrate EMD to be highly proliferative with metabolic reprogramming, EMT, and downregulation of key immunologic signaling pathways, indicating immune escape and microenvironmental independence.

Abstract Background: Recent advances in computer vision have given rise to image foundation models (FMs), large-scale neural networks pre-trained on millions of unlabeled histopathology patches, which can learn generalizable representations of tissue morphology. In many cancers, these FMs have predicted genetic mutations and prognoses directly from histology, suggesting that morphology encodes “hidden” molecular information. In lymphoma, this field is nascent but growing.General-purpose histopathology FMs have limited representation of lymphoid morphology. We developed and evaluated LymphoVision, a lymphoma-specialized image foundation model pretrained with self-supervised learning, to extract clinically meaningful features from routine hematoxylin and eosin (H&E) histology. Methods: LymphoVision was pre-trained using knowledge distillation and consistency (DINOv2) on over 37 million multi-resolution patches (5x, 10x, 20x, 40x) sampled using a density-aware strategy from 31,211 H&E whole slide images (WSIs) from 9,155 archival lymphoma and reactive lymphoproliferative cases. To ensure diverse, representative samples, our patch sampling method uses an efficient graph-based clustering algorithm and prioritizes morphologically informative regions by clustering patch embeddings and performing density sampling within each cluster. The pretrained encoder was evaluated using a clustering-constrained attention multiple instance learning (MIL) framework on three diagnostic tasks: (1) cell-of-origin (COO) classification in diffuse large B-cell lymphoma (DLBCL), distinguishing germinal center B-cell (GCB) from non-GCB subtypes based on the Hans immunohistochemistry algorithm; (2) follicular lymphoma (FL) grading, differentiating low-grade (grades 1–2) from high-grade (grade 3); and (3) multiclass lymphoma subtype classification across six categories: DLBCL, FL, marginal zone lymphoma (MZL), mantle cell lymphoma (MCL), classical Hodgkin lymphoma (CHL), and benign/reactive lymphoid conditions. WSI-level labels were derived from pathology reports, and none of the evaluation cases were included in pretraining. Models used a ViT-Giant LymphoVision backbone to generate 20x patch embeddings, followed by MIL using 50% training data and 2-fold cross-validation (25% validation, 25% test). Performance was assessed using area under the receiver operating characteristic curve (AUC), accuracy, recall (sensitivity) and F1 score. Results: For DLBCL COO classification (total n=358 cases), LymphoVision achieved a mean test AUC of 0.93 (range 0.92–0.94), accuracy of 84.5% and F1 of 0.85 across two independent test folds (n=91 per fold, 50% GCB). These results outperformed general-purpose FMs on the same task and cases, including Virchow-v2 (AUC 0.85) and UNI-v2 (AUC 0.88), despite those models being pretrained on 10–100 times more images. For FL grading (total n=660 cases), the test set in each fold included 166 cases (18% grade 3), yielding a mean test AUC of 0.88 (range 0.87–0.90), accuracy of 89.2% and F1 of 0.80. In multiclass lymphoma subtype classification (total n=1,176 cases), the test set included 296 cases per fold. LymphoVision achieved a mean test AUC of 0.98 (range 0.97-0.99), mean test accuracy of 87.1% and mean F1 of 0.84. Class-wise recall was 92.0% for DLBCL (n=112), 88.2% for FL (n=72), 77.2% for MZL (n=29), 70.8% for MCL (n=35), 87.5% for CHL (n=24), and 95.8% for benign/reactive lymphoid conditions (n=24). Conclusion: LymphoVision is a lymphoma-specialized FM trained using self-supervision that achieves high diagnostic accuracy across multiple clinically meaningful tasks using weak supervision and standard H&E images, suggesting the encoder has learned fine-grained lymphoma morphology. Based on historical benchmarks from the 1997 International Lymphoma Study Group classification project, expert hematopathologists achieved diagnostic agreement rates of 55% to 84% using H&E morphology alone for the non-Hodgkin lymphoma (NHL) entities included in our multiclass task. LymphoVision approached or exceeded these benchmarks. These findings underscore the value of our approach using disease-specialized pretraining and intelligent patch sampling for data-efficient histopathology learning. LymphoVision has the potential to augment diagnostics, enable scalable AI-powered classification in real-world workflows, and serve as a foundation for prognostic and theragnostic applications. External validation and multimodal integration are ongoing.

Abstract Introduction: Only 30% of patients (pts) presenting with abnormal bleeding symptoms receive a clear diagnosis, with many classified as having bleeding disorder of unknown cause, or BDUC (Baker et al. 2021). Currently, the recommended diagnostic workup for BDUC by the International Society on Thrombosis and Haemostasis (ISTH) includes plasmatic assays for coagulation factors, von Willebrand factor (vWF) antigen/activity, and platelet function testing (PFT), specifically platelet function analyzer-100 (PFA-100) and light transmission aggregometry (LTA). However, PFTs are not widely available and cannot be mailed to reference laboratories. We studied the diagnostic yield of an expanded testing panel which includes quantitative platelet glycoprotein flow cytometry (GpFL) and platelet electron microscopy (PTEM), which we have optimized for specimen mailing. Methods: We conducted a minimal risk, retrospective study by reviewing the electronic health records of pts (18 years and older) who were seen in the Coagulation Clinic of a tertiary academic medical center for evaluation of a bleeding disorder between October 2017 and June 2025. Testing included plasma-based assays including PT, APTT, vWF assay, FVIII, FIX, FXIII screen, and thrombin time (and additional reflex assays as needed). At the discretion of the treating clinician, PFTs (PFA-100 and LTA), GpFL, PTEM, and fibrinolysis assays (plasminogen activator inhibitor-1 antigen [PAI-1 Ag] and alpha-2 plasmin inhibitor) were included. We abstracted clinical data on pt demographics, ISTH bleeding scores, and the results of coagulation testing. Pts with an established diagnosis of a bleeding disorder, underlying malignancy, and those who did not have plasma-based coagulation testing were excluded from this study. Results: 109 pts met study criteria, and 82.6% were females (n = 90) with a median age of 45 years (range = 18-88 years). The median ISTH-BAT score was 7 (range: 2-21). About half of the pts, 49.5% (n = 54), had normal coagulation profiles and PFTs. All 54 had PTEM performed, of which 14.8% (n = 8) were abnormal (alpha and/or dense granule deficiencies). 85.2% (n = 46) had GpFL testing performed, with one showing reduced Gp2b/CD42 and Gp3a/CD61 expression, four showing reduced Gp1a/CD49b expression, and three showing reduced GpVI expression. Half of the pts, 50.5% (n = 55) had an abnormal coagulation profile from plasma and/or PFTs, of which 20% (n = 11) had a clinically significant coagulation factor deficiency (FV, FVII, FX). Roughly 9.1% (n = 5) had type 1 vWD of which 1 also had PAI-1 deficiency. One pt had isolated PAI-1 deficiency. Abnormal coagulation profiles were noted in 53 pts who had PFA-100 and LTA done, and both tests were normal in 22.6% (n = 12). Within the abnormal coagulation profile group, 10.9% (n = 6) had ultrastructural abnormalities on PTEM (dense and alpha granule abnormalities). Majority of the pts (n = 46) had GpFL done, and one had abnormal reductions in Gp1b-aCD42b, three in Gp1aCD49b, and seven in GPVI. Comparison of the ISTH BAT score between the normal and abnormal coagulation profiles/PFTs groups by non-parametric (Mann-Whitney U test) and parametric (Welch's t-test) showed no significant difference (p = 0.56 and p = 0.87) respectively. Conclusion: In this cohort of pts with abnormal bleeding, standard recommended testing resulted in a diagnosis in 66.4% pts. The remainder would have been classified as BDUC. The addition of GpFL and PTEM resulted in establishing a diagnosis that explains the bleeding phenotype in 14.8% of pts with an otherwise normal coagulation profile and PFT. Although abnormal PFTs indicate a platelet function defect, GpFL and PTEM may further define the abnormalities particularly differentiating syndromic from non-syndromic inherited platelet disorders. Additional studies are needed to confirm our findings and explore the possibility of addition of GpFL and PTEM for patients with BDUC.

Abstract Measurable residual disease (MRD) assessment by flow cytometry (FC) plays an essential role in prognosis and therapy escalation of B-cell acute lymphoblastic leukemia (B-ALL). However, the high degree of expertise and manual analysis time required limits the availability of this assay. To overcome this limitation, we developed a data-enhancing artificial intelligence (AI) pipeline that accelerates and simplifies MRD analysis. Unaltered FC files from 171 B-ALL MRD-positive and 89 MRD-negative cases were processed through an AI pipeline trained with 31 expert-gated negative controls. Cluster-informed downsampling reduced FC files from 1.2 million to 155 884 cells per case, on average, (87% cellularity reduction), whereas preserving small MRD populations (median, 100% retention for MRD of <1%) and allowing for true percentage MRD estimates using a correction factor. A deep neural network cell classifier automatically identified normal hematopoietic subsets (macro-averaged F1 score of 0.86); and an AI measure of anomaly discriminated B-ALL from benign mononuclear (area under the curve [AUC] of 0.98) or B-lymphoid cells (AUC of 0.94). Manual analysis of AI-enhanced files was completed in only 1.01 minutes per case, on average (standard deviation of ±0.57); with 100% positive agreement with conventional analysis (for MRD of ≥0.01%), 100% negative agreement, and excellent quantitative correlation (R2 = 0.92). Our cloud-based AI-enhancement solution accelerates B-ALL MRD identification without compromising test performance and has the potential of facilitating B-ALL MRD analysis by more clinical laboratories.

Background: Flow cytometric detection of minimal residual disease (MRD) in chronic lymphocytic leukemia (CLL) is complex, time-consuming, and subject to inter-operator variability. Deep neural networks (DNNs) offer potential for standardization and efficiency improvement, but require rigorous validation and monitoring for safe clinical implementation. Methods: We evaluated a DNN-assisted human-in-the-loop approach for CLL MRD detection. Initial validation included method comparison against manual analysis (n = 240), precision studies, and analytical sensitivity verification. Post-implementation monitoring comprised four components: daily electronic quality control, input data drift detection, error analysis, and attribute acceptance sampling. Laboratory efficiency was assessed through a timing study of 161 cases analyzed by five technologists. Results: Method comparison demonstrated 97.5% concordance with manual analysis for qualitative classification (sensitivity 100%, specificity 95%) and excellent correlation for quantitative assessment (r = 0.99, Deming slope = 0.99). Precision studies confirmed high repeatability and within-laboratory precision across multiple operators. Analytical sensitivity was verified at 0.002% MRD. Post-implementation monitoring identified 2.97% of cases (26/874) with input data drift, primarily high-burden CLL and non-CLL neoplasms. Error analysis showed the DNN alone achieved 97% sensitivity compared to human-in-the-loop-reviewed results, with 13 missed cases (1.5%) showing atypical immunophenotypes. Attribute acceptance sampling confirmed 98.8% of reported negative cases were true negatives. The DNN-assisted workflow reduced average analysis time by 60.3% compared to manual analysis (4.2 ± 2.3 vs. 10.5 ± 5.8 min). Conclusions: The implementation of a DNN-assisted approach for CLL MRD detection in a clinical laboratory provides diagnostic performance equivalent to expert manual analysis while substantially reducing analysis time. Comprehensive performance monitoring ensures ongoing safety and effectiveness in routine clinical practice. This approach provides a model for responsible AI integration in clinical laboratories, balancing automation benefits with expert oversight.