Hemophagocytic lymphohistiocytosis (HLH) is a rare, life-threatening condition characterized by excessive immune activation, cytokine storm, and aberrant macrophage function. Although HLH is well studied in children, data on adult HLH remain limited. Our primary goal was to examine in-hospital mortality and its associated risk factors in patients with HLH in a tertiary center. From 845,846 patients seen in the hospital between 2014-2021, a cohort of 54 adult HLH patients was identified. The overall mortality rate was 40.7%. In univariate analysis, we found that deceased patients with HLH were significantly older than surviving patients (median age of 69.6 (range 22-83) versus 52.5 (24-79) years old (p= 0.002). Patients with HLH were significantly more likely to have cardiopulmonary and neurological complications, higher alkaline phosphatase levels, lower platelet counts, need for platelet transfusions, and lower response rate to the HLH therapy. In multivariate analysis, age (HR 0.94; 95% CI 0.89-0.99; p= 0.024), cardiopulmonary (HR 7.045; 95% CI 1.28-38.66, p= 0.025), neurologic complications (HR 5.55; 95% CI 1.01-30.51; p = 0.04), and the requirement of platelet transfusions (HR 6.22; 95% CI 1.16-33.20; p = 0.032) were all independently associated with in-hospital mortality. This study identifies risk factors whose early presence can be used to stratify management strategies and improve prognosis in patients with HLH.

Introduction The significance of autografts' contamination by clonal plasma cells on clinical outcome in newly diagnosed multiple myeloma remains controversial. Methods We retrospectively reviewed the clinical and laboratory data of newly diagnosed multiple myeloma (NDMM) patients who underwent autologous stem cell transplantation (ASCT) and had received graft minimal residual disease (gMRD) examination by multi-color flow cytometry. Results From January 2011 to December 2022, 250 NDMM patients with complete cytogenetic information, gMRD information, and who received autologous stem cell transplantation (ASCT) as consolidation were enrolled. Multi-flow cytometry can achieve a median detection sensitivity of 0.004%, and gMRD positivity was 12.4% at a median level of 0.0160% (IQR, 0.0049%, 0.05394%). Its presence was correlated with response to induction treatment, with percentages of 2.65%, 12.94%, 28.89%, and 57.14% of patients achieving complete response, very good partial response, partial response, and minimal response/stable disease, respectively. gMRD (+) patients had a higher risk of not achieving bone marrow MRD negativity post-ASCT. After a median follow-up of 33.5 months for the whole cohort, patients in the gMRD (+) group had significantly worse PFS than those in the gMRD (-) group did (34.8 vs. 65.0 months, P = 0.001). Multivariable analysis revealed that gMRD (-) was independently predictive of better PFS (HR 0.464, 95%CI: 0.274-0.785, P = 0.004). We found the significance of gMRD on PFS was in high-risk subgroups and in patients who achieved ≤ partial response prior to ASCT. Conclusions In conclusion, gMRD (+) was an independent risk factor for inferior progression-free survival, with the impact primarily affecting high-risk groups and patients who achieved ≤ partial response before ASCT.

Acute Basophilic Leukemia (ABL) is a rare form of acute leukemia, defined by an elevated number of immature basophils in the peripheral blood. Clinically, it presents with skin infiltration, organ enlargement, osteolytic lesions, and elevated histamine levels and symptoms, with rapid progression and poor prognosis. Due to its rarity and the lack of specialized diagnostic tests, there is no consistent diagnostic standard. This report describes a rare case of acute basophilic leukemia that not only enhances our understanding of the clinical manifestations and pathological mechanisms of ABL, but also highlights the importance of precise diagnosis and individualized treatment. The detailed analysis of this case facilitated the identification of early symptoms of ABL, especially the relevance to the clinical manifestations of hyperhistaminemia. Meanwhile, the FIP1L1::PDGFRA fusion gene detection results strongly support the use of molecular diagnosis in ABL. In addition, the patient successfully achieved a complete response to treatment and was followed for a year and a half without relapse, thus offering valuable insights for treatment strategies in similar cases.

The entity of Philadelphia (Ph)-like acute lymphoblastic leukemia (ALL) accounts for approximately 20-30% of newly diagnosed adults with B-cell ALL cases in the USA. Compared to other B-cell subtypes, Ph-like ALL is associated with overall poor prognosis and inferior outcomes with high measurable residual disease rates following induction therapy, increased risk of treatment failure and relapse, as well as short event-free and overall survival. Here we aim to highlight Ph-like ALL genetic subtypes and methods of genomic profiling for diagnosis and disease prognostication and to summarize current management approaches for frontline treatment including multiagent chemotherapy, immunotherapy, tyrosine kinase and small molecule inhibitors, and the role of allogeneic stem cell transplantation. Despite the improvement in the treatment outcomes of adult patients with newly diagnosed B-cell ALL, patients with Ph-like ALL continue to do poorly with standard therapy. Thus, tailored therapeutic studies are indeed warranted to refine frontline treatment approaches and to improve outcomes for patients with Ph-like ALL.

Introduction: This study was aimed to identify the discrepancies in treatment goals and concerns regarding disease management between patients with myeloproliferative neoplasms (MPNs) and hematologists. Methods: A study was conducted among patients with MPNs, including polycythemia vera (PV), essential thrombocythemia (ET), and myelofibrosis (MF), and hematologists in China. Results: Data from 1,645 respondents with ET, PV, and MF and 715 hematologist respondents were analyzed. Cure of disease and healthy blood counts as treatment goals were reported more by almost half of the respondents with MPNs than by hematologists. However, prevention of thrombotic events, delayed transformation of disease, improvement of symptoms and better quality of life, and reduction in spleen size were less reported by respondents with MPNs than by hematologists. In multivariate analyses, education, comorbidities, symptom burden, disease duration, and annual out-of-pocket expenses for treatment were significantly associated with the treatment goals of respondents with MPNs. However, female physicians and senior professors paid more attention to these goals. Regarding concerns on MPN-related issues, more respondents with MPNs paid more attention to disease knowledge and restrictions in daily life compared to hematologists, whereas the majority of physicians attached importance to medication-related issues. Conclusion: The perceptions of patients with MPNs and hematologists differed in terms of treatment goals and concerns of management of MPNs. Sociodemographic and clinical variables were associated with the respondents’ perspectives on MPNs. Therefore, sufficient patient-physician communication is suggested to improve treatment satisfaction and compliance.

Introduction: Research has demonstrated a potential link between lipid metabolites and multiple myeloma (MM); however, the causal relationship remains uncertain. This Mendelian randomization (MR) study aimed to explore the potential causal relationship between lipid metabolites and MM. Methods: In this study, data on lipid metabolites were obtained from a genome-wide association study of metabolites in blood samples from 7,824 Europeans. Genetic information related to MM came from the UK Biobank database, encompassing 601 patients with MM and 372,016 control samples. In this MR analysis, inverse-variance weighted method was used as the primary analysis method; MR-Egger and weighted median were employed as complementary approaches. Sensitivity analyses were conducted using the Cochran Q test, MR-Egger intercept, MR-PRESSO, and leave-one-out analysis. Results: A total of 121 human lipid metabolites were analyzed in this MR study. The analysis result revealed that 1-docosahexaenoyl-glycerophosphocholine (odds ratio [OR] = 1.0059, 95% confidence interval [CI] 1.0043–1.0076, p < 0.01, FDR = 0.12), tetradecanedioate (OR = 1.0007, 95% CI: 1–1.0013, p = 0.0498, FDR = 0.23), and X-12990-docosapentaenoic acid (OR = 1.0029, 95% CI: 1.0015–1.0044, p < 0.01, FDR = 0.15) were linked to an increased risk of MM. As for palmitoleate (OR = 0.9972, 95% CI: 0.9947–0.9997, p = 0.0299, FDR = 0.19), a nominal inverse association was observed. None of these associations reached statistical significance after FDR correction (all FDR >0.05). Sensitivity analyses verified the robustness of these nominally significant results. Conclusion: Genetic evidence demonstrated nominal associations of 1-docosahexaenoyl-sn-glycero-3-phosphocholine, tetradecanedioate, X-12990-eicosapentaenoic acid, and palmitoleate with MM risk, though these did not survive FDR correction. While these findings suggest potential metabolic pathways in MM pathogenesis, further validation is required before considering these compounds as biomarkers for clinical screening or drug target selection.

Plain Language SummaryHaematology encompasses a wide variety of disorders affecting the blood and bone marrow. A diagnosis often requires multiple – and, with increased understanding of diseases, progressively complex – tests. Doctors and laboratory professionals are therefore exploring how artificial intelligence (AI) can help improve the way disorders of the blood and bone marrow are detected and understood, particularly how AI offers new ways to make this process faster and more accurate. This article reviews how AI is being used in several key areas of haematological laboratories. These include examining cells under a microscope, analysing patterns in cells using specialized machines (a method called Flow Cytometry), studying genetic changes, and measuring the ability of blood to form clots (a process called haemostasis). Some current uses of AI include automatically sorting different types of blood cells, identifying possible causes of anaemia (a condition where there are not enough healthy red blood cells), and quickly detecting certain types of leukaemia, a cancer of the blood. AI can also help reduce human error when analysing test results and speed up the processing of large genetic datasets. However, there are challenges. Many AI tools do not work well across different laboratories or have not been tested enough to be used in everyday practice. Currently and in the near future, AI is best used to support experts by handling some of the early steps in the testing process. Final decisions about a diagnosis still need to be made by trained professionals in blood diseases.

Introduction: Diffuse large B-cell lymphoma (DLBCL) is a common type of non-Hodgkin lymphoma with clinical and genetic heterogeneity, resulting in significant differences in patient prognosis. Methods: High-throughput sequencing was performed on 155 newly diagnosed DLBCL patients, and 12 genes with high mutation rates related to DLBCL were selected. Cox regression analysis was used to determine prognostic factors associated with progression-free survival (PFS) and overall survival (OS) in patients. A new prognostic model was established based on these factors, and its performance was validated using the concordance index (C-index), receiver operating characteristic curve, and calibration curve. Clinical utility was evaluated using decision curve analysis (DCA). Results: Multivariable Cox regression analysis showed that the prognostic factors for PFS and OS in DLBCL patients were IPI, FAT4 mutation, and TP53 mutation, leading to the development of the final prognostic model (FAT4-TP53-IPI model). The FAT4-TP53-IPI model demonstrated better discriminative ability than the IPI model, as indicated by the C-index. The calibration curve showed good discriminatory ability and accuracy, and DCA confirmed the clinical value of the FAT4-TP53-IPI model. Based on the cutoff values obtained from the FAT4-TP53-IPI model, patients were divided into two different risk groups, and survival analysis for PFS and OS demonstrated significantly worse prognosis in the high-risk group compared to the low-risk group (p < 0.01). Conclusion: This study demonstrates that integrating genetic mutation status enhances the prognostic value of the IPI scoring system. Our model may serve as a valuable tool for predicting the prognosis of DLBCL patients receiving rituximab-based immunotherapy.