Artificial intelligence (AI) and its subdiscipline, machine learning (ML), have the potential to revolutionize health care, including hematology. The diagnosis and treatment of hematologic disorders depend on the integration of diverse data sources, such as imaging, pathology, omics, and laboratory parameters. The increasing volume and complexity of patient data have made clinical decision-making more challenging. AI/ML hold significant potential for enhancing diagnostic accuracy, risk stratification, and treatment response prediction through advanced modeling techniques. Generative AI, a recent advancement within the broader field of AI, is poised to have a profound impact on health care and hematology. Generative AI can enhance the development of novel therapeutic strategies, improve diagnostic workflows by generating high-fidelity images or pathology reports, and facilitate more personalized approaches to patient management. Its ability to augment clinical decision-making and streamline research represents a significant leap forward in the field. However, despite this potential, few AI/ML tools have been fully implemented in clinical practice due to challenges related to data quality, equity, advanced infrastructure, and the establishment of robust evaluation metrics. Despite its promise, AI implementation in hematology faces critical challenges, including bias, data quality issues, and a lack of regulatory frameworks and safety standards that keep pace with rapid technological advancements. In this review, we provide an overview of the current state of AI/ML in hematology as of 2025, identify existing gaps, and offer insights into future developments.

Until recently, the JAK1/2 inhibitor ruxolitinib (Jakafi) was the only therapy for steroid-refractory acute graft-versus-host disease approved by the US Food and Drug Administration (FDA) for use in patients aged >12 years. The FDA has now approved a potent mesenchymal stromal cell product, remestemcel-L-rknd (Ryoncil), for children aged ≤18 years, showing 70% response rates and ∼70% 6-month survival. In this spotlight, we highlight this important advance in the field.

Two gene therapy products have been approved by the US Food and Drug Administration for sickle cell disease. Nearly all patients in the clinical trials that led to approval either were sickle hemoglobin (HbS) gene homozygotes (sickle cell anemia) or had HbS-β0 thalassemia. HbSC disease, caused by compound heterozygosity for HbS and hemoglobin C genes, is the second most common genotype of sickle cell disease. Gene therapy has not been tested in patients with HbSC disease who are severely symptomatic. We discuss the pathophysiology and clinical features of HbSC disease and how gene therapy is likely to provide a curative option for some individuals. We also discuss the mechanism through which fetal hemoglobin (HbF) and HbF-like HbA (HbAT87Q) might mitigate adverse clinical outcomes and end-organ damage in patients with HbSC disease and other compound heterozygous sickle hemoglobinopathies.

Recent advances have transformed the treatment landscape for relapsed and refractory follicular lymphoma. Although chemotherapy has long served as the backbone of treatment, the availability of novel targeted, immunomodulatory, and immunotherapeutic approaches is challenging its relevance. These approaches have focused on targeting epigenetic regulators, components of the B-cell receptor or its downstream intracellular pathways and the follicular lymphoma tumor microenvironment. The recent development of bispecific antibodies and chimeric antigen receptor T-cell therapies, which target both tumor-associated and host-specific antigens, has enabled a redirection of the immune system, enhancing the innate antitumor immune response. Rational combinations of these strategies are actively being evaluated in the relapsed and refractory setting and will inevitably move forward into earlier lines of treatment. The success of these approaches has led to numerous and parallel options for patients and clinicians. The emerging challenge now lies in how best to approach each individual patient with relapsed or refractory follicular lymphoma, addressing complex decision-making that considers a patient's previous treatment history, goals of care, clinical and biological characteristics of recurrence, and personal preferences. Understanding the implications of refractory and transformed disease, as well as the timing and biology of relapse will be critical to support a more personalized treatment approach in the modern era.

The US Food and Drug Administration recently licensed 14-day cold-stored platelets for bleeding patients. This policy change represents a reversal from the 1970s when cold-stored platelets were discontinued because of their short circulation time in healthy humans. This change will increase their availability in US hospitals with large trauma populations and in remote and rural settings in the United States. In some of these hospitals, cold-stored platelets will be the only platelets available. It is currently unclear whether patients with hypoproliferative thrombocytopenia who need platelet transfusion for prophylaxis benefit from cold-stored platelets. However, it is noteworthy that in recent clinical trials using room temperature-stored platelets, the transfusion interval in patients with hematologic and oncologic conditions can be as short as 1 transfusion per day, very similar to what one would expect to achieve with cold-stored platelets. Furthermore, the emphasis on the posttransfusion count increment and the platelet count as a transfusion trigger per se is questionable. In the PLADO trial, there was only a poor correlation between the morning platelet count and bleeding events, implicating other factors, such as red blood cells, coagulation factors, and vascular health, as possible culprits. In this perspective article, we review the history of cold platelets and the reason for their discontinuation, focus on recent clinical trial data using room temperature-stored platelets, and review the platelet count as a transfusion trigger. Overall, using cold platelets for prophylaxis may seem counterintuitive, but a closer look at the available data suggests that the indication expansion may hold more promise.

In 1974, Van den Berghe et al described a distinct hematologic disorder associated with acquired, interstitial deletion of part of the long arm of chromosome 5. This condition is now classified as myelodysplastic syndrome (MDS) with isolated deletion 5q, or MDS-del(5q). The common deletion region 5q32-5q33 contains several genes and microRNAs whose expression levels are reduced in hematopoietic cells, consistent with the loss of 1 allele. Haploinsufficiency production of multiple gene transcripts, primarily involving CSNK1A1, RPS14, MIR145, and MIR146A, results in myelodysplastic hematopoiesis. Lenalidomide can selectively suppress the del(5q)-mutant clone by promoting proteasomal degradation of casein kinase 1A1 and inducing mutant stem cell failure. However, lenalidomide is not a curative treatment, as almost all patients relapse. Molecular profiling studies have significantly improved our understanding of MDS-del(5q). Only a minority of patients have interstitial deletion 5q as their sole genetic lesion, a condition that is associated with an indolent clinical course. Most patients have co-occurring somatic mutations in myeloid genes, including DNMT3A, TET2, ASXL1, SF3B1, TP53, RUNX1, and CSNK1A1. These comutations have independent effects on leukemic transformation and survival, so genomic profiling is required for implementing a precision management approach to MDS-del(5q) in a clinical setting. Accurate assessment of the TP53 allelic state is crucial for distinguishing MDS-del(5q) from TP53-mutant MDS, a myeloid malignancy characterized by TP53 multihit state and very aggressive clinical course. Genomic profiling is also critical for therapeutic decision-making in patients with MDS-del(5q), particularly for assessing a patient's eligibility for allogeneic transplantation, which remains the only curative treatment.

Recent clinical trials in both transplant-eligible and -ineligible newly diagnosed multiple myeloma using 3- and 4-drug combinations have demonstrated unprecedented levels of response. However, 2 recently published studies redefining high risk in newly diagnosed multiple myeloma in the context of these newer and more effective treatments demonstrate that a significant minority of patients likely derive little benefit from these newer approaches. These new prognostic systems thus provide an evidence-based framework for the development of much-needed risk-stratified clinical trials.

TP53 mutations are found in 10% to 15% of myeloid neoplasms and are one of its most important prognostic factors. Emerging data show that TP53 mutational allele status is a key determinant of clinical outcomes, with multihit TP53 mutant myeloid neoplasms having a very poor prognosis. Significant differences exist among the methods used in clinical and research settings to assess TP53 mutational status, leading to variability in reported patient characteristics, response to therapy, and survival. Indeed, differences in the criteria used to define TP53 mutational states among professional societies and in landmark research studies have led to confusion, suboptimal clinical testing, and variability in therapy recommendations. We review the methods used to assess for TP53 mutational allele status and provide recommendations, based on clinically available testing, for the accurate evaluation of TP53 gene mutations in myeloid neoplasms. Hotspot mutations represent ∼35% of all TP53 missense mutations in myeloid neoplasms. There is evidence that these hotspot mutations may have dominant-negative or gain-of-function properties. Here, we review this evidence and discuss the potential impact of TP53 mutation identity on patient outcomes and clinical management.

The treatment landscape for chronic lymphocytic leukemia (CLL) has been transformed by the advent of covalent Bruton tyrosine kinase (BTK) inhibitors (cBTKis) and B-cell lymphoma 2 (BCL-2) inhibitors, leading to markedly improved outcomes and, for many, near-normal life expectancy. However, patients progressing after both classes of therapy (double-refractory) have limited options and poor prognoses. This review outlines a practical approach to managing double-exposed or double-refractory CLL, incorporating clinical cases, trial data, and expert perspectives. For cBTKi intolerance, second-generation agents may remain effective. Venetoclax retreatment is reasonable after prior fixed-duration use. In true double-refractory disease, noncovalent BTK inhibitors (eg, pirtobrutinib) and CD19-directed chimeric antigen receptor T-cell (CAR-T) therapy (lisocabtagene maraleucel) are standard-of-care options. Pirtobrutinib induces rapid responses, though often of limited duration, underscoring the need for early consolidation planning with CAR-T or allogeneic stem cell transplant. Persistent disease after CAR-T therapy warrants close monitoring and timely transplant referral in eligible patients. Phosphoinositide 3-kinase inhibitors remain available but are limited by toxicity and modest benefit. Emerging agents, including BTK degraders, bispecific antibodies, and novel cellular therapies, offer promising future directions. Optimizing outcomes in double-refractory CLL requires an individualized, nuanced strategy integrating available treatments with innovative approaches under investigation.

Hematologists have a significant role in the diagnosis and management of light-chain amyloidosis and are frequently involved in the diagnosis of transthyretin and other more rare types of amyloidosis. Recent advances in diagnostic techniques and therapies are dramatically improving patient prognosis. Radionuclide imaging methods are emerging as a highly specific and noninvasive way to diagnose, quantify, and monitor organ involvement representing a major advance in amyloidosis management.

Venous thromboembolism (VTE) affects ∼1 in 200 hospitalized children. The diagnosis of pulmonary embolism (PE), the most severe clinical presentation of VTE, has increased dramatically by ∼200% over the past 2 decades, disproportionately affecting adolescents, and is associated with adverse long-term post-PE sequelae. Nevertheless, the management of pediatric PE remains highly variable. This review focuses on significant advances in pediatric PE, with a focus on published studies within the past decade. Using a representative case, we: (1) summarize existing risk prediction tools for acute pediatric PE, and a shift in clinical practice in the management of acute PE with the implementation of pediatric PE response teams and multidisciplinary decision-making for severe pediatric PE, (2) describe recently completed clinical trials of anticoagulation in children and adolescents, and ongoing work to elucidate the appropriate duration of therapy for children and adolescents with PE, and (3) review advances in understanding post-PE syndrome, and the need for continued refinement of evaluation tools and management approaches. Many unanswered questions remain despite the significant advances in pediatric thrombosis over the past decade.

The contact system includes factor XII (FXII), FXI, prekallikrein (PK), and high-molecular-weight kininogen (HK), and has received increased interest as a potential target in immunothrombotic and inflammatory diseases. This system activates 2 distinct pathways, the intrinsic pathway of coagulation via cleavage of FIX, and inflammation via HK cleavage resulting in bradykinin (BK) generation. HK is central to the function of both arms of the system as a substrate for plasma kallikrein and critical cofactor, which forms interactions with cell receptors and activators. Both FXI and PK circulate in complex with HK and both can be cleaved by activated FXII. Reciprocal activation and continuous consumption of PK and FXII is a feature of the contact system. On endothelial cells, PK and FXII become activated but only in the presence of secreted receptor for the globular domain of C1q and Zn2+ ions. A second mechanism exists on endothelial cells whereby prolylcarboxypeptidase activates the PK-HK complex to generate BK in an FXII-independent manner. On platelets, FXI can be cleaved by thrombin, but only in the presence of secreted polyphosphate. This review explores the 3-dimensional structure of the contact factors and examines the molecular mechanisms underlying contact activation. We focus on conformational changes that expose cleavage sites and exosites in FXII, PK, and FXI. We also discuss contact factor protein-protein interactions, recognition of polyanions, and the role of HK and Zn2+ in contact system assembly.

The pathogenesis of multiple myeloma (MM) and its precursor monoclonal gammopathy of undetermined significance (MGUS) is linked to an aging immune system. Chronic activation of B/plasma cells may contribute to the origin of MGUS, which is frequent in the older individuals. However, only 1% of individuals with MGUS annually experience progression to MM. The immune system can specifically recognize MGUS lesions, and preclinical MM models provide evidence for both innate and adaptive immune surveillance. Multiomic studies have identified several systemic alterations at the MGUS stage, suggesting accelerated immune aging prior to evolution into clinical malignancy. MM is further associated with spatial alterations in patterns of tumor growth and in situ regulation of regional immunity. Both tumor and microenvironment-related factors contribute to immune paresis, which facilitates the dissemination of clonal plasma cells, and increases the risk of infections in patients with MM. Immune profiles in blood or marrow exhibit considerable heterogeneity, and have been linked to outcomes following immune therapies, including T-cell redirection. Understanding how underlying systemic immune changes impact in vivo function and durability of natural or synthetic tumor/antigen-specific immunity needs further study. Preserving or restoring immune function may be critical for long-term outcomes both in the context of prevention of clinical MM and of treating active disease. Benchmarking of immune biomarkers followed by its prospective integration into current risk models, together with improved understanding of mechanisms underlying tumor immunity in vivo, are needed to optimize immune approaches and improve outcomes in MM.

The development of multiple myeloma is typically associated with various cytogenetic abnormalities; however, these genetic changes alone do not fully account for the observed heterogeneity in patient prognosis and treatment response. Recent studies leveraging next-generation sequencing and genomic approaches have shown that epigenetic alterations are crucial in myeloma development and therapeutic resistance. These changes contribute to high levels of transcriptomic instability and enable cellular adaptation to targeted therapies and immunotherapies through diverse evolutionary trajectories. In this regard, aberrations of histone modifications and chromatin remodeling affect various cellular processes such as DNA repair, DNA damage response, cellular survival, and apoptosis signaling, which provides a strong rationale for developing epigenetic-targeted therapies for myeloma treatment. In this review, we focus on recent advances and research gaps in understanding the deregulation of histone acetylation, a widespread and versatile process of histone modification occurring at lysine residues at the N-terminus of histone tails, and its intimate interplay with chromatin remodeling complexes in orchestrating dynamic chromatin functional states and transcriptional outputs. We also provide an updated review of epigenetic modulatory drugs targeting histone deacetylases (CREB-binding protein/p300) and bromodomain and extraterminal proteins, along with a discussion of their limitations and future perspectives in myeloma treatment.

Immunological memory in adaptive and innate immune cells is well characterized, enabling enhanced responses upon secondary challenges. However, it has only been recently appreciated that the nonimmune target cells of inflammation, particularly organ-specific stem cells (SCs), also exhibit memory of previous inflammatory exposures. Previous inflammation experience imprints on the SCs and influences their regenerative potential and responses to subsequent inflammatory insults. This phenomenon has been observed in hematopoietic, intestinal, and skin epithelial SCs, with profound implications for tissue homeostasis, disease progression, and therapeutic strategies. Herein, we expand and develop the notion of inflammatory memory of SCs and explore recent insights in the field. We discuss the emerging understanding of the molecular underpinnings and their potential clinical and biological implications. Inflammatory memory is driven by spatiotemporal changes in gene loci and transcription regulated by DNA and histones' epigenetic modifications, metabolic reprogramming, and chromatin accessibility changes. Understanding these mechanisms is critical for improving the outcomes of hematologic diseases, hematopoietic SC transplantation, and cellular immunotherapies.

Although allogeneic hematopoietic stem cell transplantation is a leading treatment approach for myeloid malignancies, challenges in its immune biology and in treatment approaches remain. In the past decade, major advances in the knowledge of mechanisms of graft-versus-host disease (GVHD) has allowed development of new treatments both for GVHD prophylaxis and treatment. However, although successes did occur, failure did as well. Reasons for failure can be linked either to incomplete understanding of the pathophysiology of GVHD, or, in some cases, to errors in the design of clinical trials. Better GVHD prophylaxes and disease control have likely led to decreased nonrelapse mortality (NRM). However, although NRM rates have decreased, rates of relapse of the original malignancy have not significantly improved. Our current understanding of the biology of the graft-versus-leukemia (GVL) effect still lags behind that of GVHD, and treatment approaches to manipulate the GVL effect remain limited. The reasons for such a lag are numerous, but improved knowledge of the biology of hematological malignancies open the gate to new developments, providing that we can better understand the interplay between the immune system with leukemic clones. From a therapeutical perspective, much attention has been paid to the results from randomized clinical trials and from a biological perspective on recent discoveries, especially in the human setting. The objective of this perspective is to analyze what are the current challenges in the biology and treatment of GVHD and GVL and to provide a personal view on how some biological and therapeutic issues could be approached.

Chronic lymphocytic leukemia (CLL) is a disease of great clinical and biologic heterogeneity; some patients are observed for years without symptoms, whereas others rapidly develop progressive disease requiring treatment. With therapy, some patients eventually develop resistant CLL or transformation to an aggressive form. Across this spectrum, patients experience immune dysfunction associated with increased risk for infection and second cancers, contributing to morbidity and mortality of the disease. The ultimate therapeutic bull's-eye for CLL is to eliminate the disease and achieve immune restoration. Disease elimination can potentially be achieved for a fraction of patients treated first line with chemoimmunotherapy (fludarabine, cyclophosphamide, and rituximab), for some patients who receive time-limited combined targeted therapy, and for some patients with relapsed/refractory CLL who undergo allogeneic stem cell transplant. Long-term immune restoration for these patients is elusive. Current targeted therapies, including Bruton tyrosine kinase and B-cell lymphoma 2 inhibitors and CD20 monoclonal antibodies used in combinations, can produce exceptional therapeutic outcomes, which are improving survival for patients who need treatment. Although clear progress has been made toward highly effective CLL management, appreciation of the full impact of these advances will require time because of the chronic nature of the disease. In addition, it is imperative to ensure global access to the targeted therapies, emphasizing the need for harmonized regulatory oversight and affordable treatment options worldwide. Here, we discuss research and collaborative strategies to refine the use of targeted agents to eliminate CLL and restore immune function for all affected individuals.

Although most patients with follicular lymphoma (FL) follow an indolent disease course, some patients experience a critical inflection point when FL transforms into an aggressive lymphoma. Historically, FL transformation is marked by poor outcomes, particularly for patients with previous FL-directed treatment. Compared with FL, transformed FL (tFL) is marked by numerous additional genetic changes, upregulates novel signaling pathways, and arises from an ancestral FL clone with shared FL-initiating mutations. Prediction of tFL risk remains a high-priority area of disease research, with recent work highlighting memory-like B-cell phenotypes associated with transformation risk and implicating critical tumor-immune interactions at transformation emergence. Mechanistic studies provide insight into the role of genetic drivers in determining malignant B-cell phenotypes or reducing microenvironmental dependencies. In parallel, a shifting therapeutic landscape marked by novel immune-based therapeutics is improving outcomes for patients, yet further clinical outcome data in tFL are greatly needed. This review summarizes recent scientific and clinical studies in tFL and provides an updated understanding of the biological basis, diagnosis, and clinical management of tFL. We conclude with a proposed plan of future research aimed at the goal of increasing tFL biologic knowledge and improving outcomes for patients with tFL.

In 2013, we published a Perspective titled "The myth of the second remission of acute leukemia in the adult," which underscored the dismal outcomes of relapsed acute leukemia in adults. We emphasized that only a few patients achieved second complete remission (CR2) after relapse and were subsequently eligible to receive a potentially curative allogeneic hematopoietic stem cell transplantation (HSCT). Hence, we urged the leukemia community not to delay HSCT in first complete remission if indicated to avoid dire outcomes. Historically, poor outcomes resulted from suboptimal frontline therapy, inadequate risk stratification, and lack of effective agents to achieve CR2. In the past decade, remarkable progress has been made in the treatment paradigm of acute leukemia, most evidently in B-cell acute lymphoblastic leukemia. Key advancements include refinement of frontline treatment, incorporation of early immunotherapy, improved disease risk stratification based on molecular profiling and assessment of measurable residual disease, and discovery of highly effective salvage immunotherapies. These innovations have led to a high rate of cure by frontline therapy, precise selection for HSCT in first complete remission for high-risk patients, and the reality of HSCT for patients in CR2. Here, we re-evaluate the myth of CR2 given the progress in the field.

Thrombotic thrombocytopenic purpura (TTP) is a rare, potentially fatal thrombotic microangiopathy caused by severe ADAMTS13 (a disintegrin and metalloproteinase with thrombospondin motif 13) deficiency. Prompt treatment improves survival; however, reference standard enzyme-linked immunosorbent assay and FRETS-VWF73 ADAMTS13 activity assays have long turnaround times (TATs) that necessitate empiric treatment of many patients who ultimately are found not to have TTP. Rapid assays with analytical TATs <1 hour have recently become available. We conducted a systematic review and meta-analysis of the performance characteristics of rapid assays relative to reference standard assays for ADAMTS13 activity for patients with suspected or confirmed TTP. Nineteen studies representing 3 rapid ADAMTS13 assays and 4207 patient samples were included. The HemosIL AcuStar chemiluminescence immunoassay (CLIA) demonstrated high sensitivity (0.98; 95% confidence interval [CI], 0.94-1.00), specificity (0.99; 95% CI, 0.97-1.00), and positive (PPV) (0.96; 95% CI, 0.90-0.98) and negative predictive values (NPV) (0.99; 95% CI, 0.99-1.00). The Technofluor fluorescence resonance energy transfer (FRET) and Technoscreen assays had sensitivity of 0.93 (95% CI, 0.86-0.96) and 0.98 (95% CI, 0.42-1.00), specificity of 0.98 (95% CI, 0.95-0.99) and 0.87 (95% CI, 0.76-0.94), PPV of 0.97 (95% CI, 0.85-1.00) and 0.71 (95% CI, 0.59-0.80), and NPV of 0.96 (95% CI, 0.93-0.98) and 0.99 (95% CI, 0.72-1.00), respectively. The proportion of discrepant results (relative to reference standard assays) was 0.04 (95% CI, 0.03-0.05) for HemosIL AcuStar, 0.04 (95% CI, 0.02-0.06) for Technofluor FRET, and 0.11 (95% CI, 0.07-0.16) for the Technoscreen assay. With rapid TAT and high sensitivity, the HemosIL AcuStar CLIA seems able to reliably avert empiric plasma exchange, corticosteroids, and caplacizumab in patients without TTP.