Shah, Surbhi

Abstract Systems-based hematology (SBH) is an emerging discipline focused on optimizing care delivery for patients with hematologic disorders within complex health care systems. Despite its growing importance, the field lacks a formal definition, standardized curriculum, and clear career pathway, hindering widespread adoption and training of future SBH experts. We sought to define core principles, competencies, challenges, and future directions of SBH to identify foundational components for a potential educational curriculum. Semistructured interviews were conducted with 8 SBH experts. Expert participants were identified through their committee involvement with the American Society of Hematology. Five overarching themes emerged: (1) The accidental expert: SBH careers have developed organically and reactively in response to pressing institutional needs, not through formal training; (2) Defining the unseen: experts define the field pragmatically by its function and roles; (3) The SBH toolkit: success requires a hybrid skillset combining clinical hematology expertise with health systems science (HSS) principles; (4) The architect's dilemma: the field's growth is hindered by interdependent challenges of building a curriculum, awareness, and securing resources; and (5) Charting the future: the discipline is poised for growth, driven by increasing health care complexity and specialization. SBH represents the nexus of HSS principles and hematology care. Formalizing SBH requires a multipronged approach, including developing a standardized curriculum that equips learners with a hybrid skillset, creating clear value for health systems and patients, and fostering a community of practice through professional societies. These findings provide a foundational framework for developing educational resources and institutional support necessary to continue advancing SBH.

Abstract Background: Autoimmune hemolytic anemia (AIHA) is a condition in which the immune system erroneously attacks and damages red blood cells. While the epidemiology of AIHA has been examined in various European and Asian nations (Hansen 2020; Kao 2025), there is limited data available from the United States (US). Using US insurance claims data, this research aimed to determine the prevalence and demographic characteristics of AIHA over a five-year span. Methods: Five databases were used: Merative MarketScan Commercial Claims and Encounters; Optum's de-identified Clinformatics® Data Mart Database; Closed and Open claims sourced through HealthVerity; Merative MarketScan Medicare Supplemental Database; Merative MarketScan Multi-State Medicaid. AIHA was defined by ICD-10 diagnosis codes D59.1x in a given calendar year and a second code within 365 days prior. Annual prevalence in years 2019 – 2023 was calculated among people with insurance enrollment for at least 6 months in the calendar year. Annual prevalence was stratified by age group, sex, and prescription fill for an AIHA-related treatment (systemic steroids, immunosuppressants [azathioprine, mycophenolate, cyclosporin, cyclophosphamide, danazol, bortezomib, rituximab], splenectomy) in the calendar year. Extrapolation of prevalence rates for commercial and government assistance insurance to the US population for 2023 was implemented by age and sex-standardization to the US Census population. For data sources that included both commercial and government assistance insurance, patients aged <65 years were analyzed as commercial insurance and aged ≥65 years as government assistance. Results: AIHA prevalence in 2023 across databases ranged from 2.7 to 4.3 per 100,000 for ages <18 years; 4.2 to 9.4 per 100,000 for ages 18 – 34 years; 8.2 to 11.8 per 100,000 for ages 35 – 49 years; 15.4 to 23.9 per 100,000 for ages 50 – 64 years; 26.6 to 37.0 per 100,000 for ages 65 – 74 years; and 30.6 to 61.2 per 100,000 for age 75 years and older. Across databases and over time, the proportion of AIHA patients that were female was 55% – 67%. Across all databases, in year 2023, half (44% - 55%) of patients received any AIHA-related treatment, out of which one third (26%-34%) received systemic steroids, less than one quarter (9% – 17%) received rituximab, and one third (25% – 31%) received a different immunosuppressant. Compared to patients with commercial insurance, Medicaid-insured patients had lower AIHA-related treatment utilization, whereas Medicare-insured patients had higher utilization. Within age group and sex and by insurance type, low and high estimates of prevalence were used to define a range of prevalence. After standardization to the 2023 US Census population, the estimated prevalence in 2023 ranged from 38,459 to 57,235 patients diagnosed with AIHA, corresponding to prevalence rates of 11.6 to 17.3 per 100,000. This marked an increase from 2019 US Census-standardized estimates of 32,777 to 48,482, with corresponding prevalence rates of 10.1 to 14.9 per 100,000. Discussion: AIHA prevalence was higher in older age groups, with a peak prevalence in ages 65 and older. Prevalence increased approximately 15% over a 5-year interval and was notably higher in patients with commercial health insurance compared to government insurance programs, which may in part be due to the older age of the Medicare population. Fewer than half of patients received an AIHA-related treatment in 2023, which could reflect heterogeneity in disease severity necessitating treatment, in addition to the relapse-remitting nature of AIHA whereby patients experience remission for periods of time. Additionally, insurance claims do not specify an indication for treatment nor a reason for discontinuation; some treatments could have been prescribed for an underlying condition contributing to AIHA. Finally, prevalence extrapolation assumes Americans without health insurance are diagnosed with AIHA at the same rate as those with insurance, which may slightly overestimate the total number of diagnosed patients.Conclusion: AIHA prevalence in this study was consistent with estimates from European countries and has been increasing in recent years, providing evidence of consistent disease epidemiology across several geographies.

Abstract Introduction: Monoclonal gammopathy of thrombotic significance (MGTS) is a recently recognized thrombophilia where monoclonal platelet factor 4 (PF4)-specific IgG antibodies initiate a chronic prothrombotic state characterized by recurrent and potentially life-threatening thrombosis. Studies from large data sets demonstrate that patients with monoclonal gammopathy of undetermined significance (MGUS) have a heightened risk of experiencing thromboembolic events without a precise explanatory mechanism. The prevalence of MGTS in MGUS-related thrombosis is unknown; however, the recent identification of MGTS antibodies as a cause of thrombosis suggests a potentially causative mechanism. Emerging evidence indicates that detecting pathogenic MGTS antibodies presents diagnostic challenges due to the insensitivity of FDA-approved diagnostic assays adopted from HIT (HIT ELISAs) and gold-standard functional tests that use heparin-treated platelets, such as the serotonin release assay (SRA). Only PF4-enhanced functional assays, including the PF4-dependent P-selectin expression assay (PEA) and PF4-SRA, have proven sensitive for detecting MGTS antibodies. In this study, the PEA was utilized to better define the incidence of pathogenic anti-PF4 antibodies among a cohort of MGUS patients with a history of thrombosis. Methods: A search of the Mayo Clinic Dysproteinemia Bank was conducted to identify two cohorts of IgG MGUS patients, one with venous thromboembolism-associated diagnosis code(s) and the other without. A diagnosis of multiple myeloma was exclusionary. Each serum sample was assessed using the PEA. Mayo Clinic's Institutional Review Board approved the research study. Results: Among the 134 patients studied in the thrombosis cohort, 17 (13%) stimulated platelet activation in the PEA (positive cut-off ≥19%), some weakly. Of these, four patients (Patients 1-4) were confirmed to have PF4-dependent platelet-activating antibodies inhibitable with high concentrations of heparin and sensitive to blockade of the IgG receptor on platelets, FcGRIIa. The thrombotic complications experienced included recurrent deep venous thrombosis (DVT) over several years (Patient 1), a single DVT (Patient 2), a thrombotic cerebrovascular accident (CVA) (Patient 3), and a series of thrombotic events, including DVTs, left atrial thrombosis, pulmonary embolism, and acute limb ischemia (Patient 4). Notably, longitudinal samples spanning twelve years (Patient 1), six years (Patient 2), three years (Patient 3), and on a single sample date (Patient 4) were confirmed positive by PEA. The prolonged persistence of platelet-activating anti-PF4 antibodies in Patients 1-3 supported a diagnosis of MGTS in these patients. Strikingly, all four patients tested negative in two FDA-approved HIT ELISAs (Immucor PF4 IgG and Zymutest HIA IgG). SRA was negative for Patients 1 and 2, weakly positive for Patient 3 (33%), and positive for Patient 4 (72%). Of the 166 patients in the MGUS minus thrombosis cohort, one patient demonstrated a non-heparin induced HIT-like antibody that was HIT ELISA (Optical Density >1.0) and PEA-positive, and inhibitable using high concentrations of heparin and FcGRIIa blockade. No follow-up samples were available to assess antibody longevity. The patient had no history of thrombosis, and testing demonstrated that the anti-PF4 antibody was not synonymous with the MGUS antibody.Conclusions: Among a cohort of 134 MGUS patients with thrombosis, the prevalence of MGTS was found to be 2.2% (3 patients with persistent platelet-activating anti-PF4 antibodies). The identification of anti-PF4 antibodies among MGUS patients provides a mechanism contributing to the increased thrombotic risk associated with monoclonal gammopathies. Antibody negativity in HIT ELISA and variable positivity in the SRA highlight the nonclassical serology associated with anti-PF4 antibodies in MGTS and support performing PF4-enhanced functional testing. In addition, some patients appear to produce coincidental (non-monoclonal) platelet-activating anti-PF4 antibodies without evidence of thrombosis, although the longevity of such antibodies could not be examined in the one patient where it was identified. These findings highlight the need for larger, prospective MGUS studies to validate these observations, further define the interplay between monoclonal gammopathies and thrombotic risk, and to better understand the long-term outcomes of patients with MGTS.

Abstract Background Thrombotic Thrombocytopenic Purpura (TTP) is a rare, typically acquired disorder due to an immune-mediated deficiency of ADAMTS13, a metalloprotease responsible for cleaving large multimers of von Willebrand Factor (vWF). In the absence of ADAMTS13 activity, ultra-large vWF multimers accumulate, promoting widespread platelet aggregation and microvascular thrombosis. This consumptive cascade leads to thrombocytopenia, microangiopathic hemolytic anemia, fever, neurological symptoms, and rarely renal dysfunction. Although therapeutic plasma exchange (TPE) has long been the cornerstone of TTP treatment, the FDA approved Caplacizumab as a first-line treatment for TTP in 2019. This drug is a bivalent single-domain antibody that binds to the A1 domain of vWF and prevents its interaction with platelets, thus effectively interrupting the pathogenic cascade, which is associated with a risk of bleeding. Methods We conducted a retrospective cohort study using the TriNetX Global Collaborative Network, comprising data from 152 healthcare organizations. Adults (≥18 years) with a diagnosis of TTP who received Caplacizumab plus TPE were compared to patients with TTP who received TPE alone. 1:1 propensity score matching (PSM) was performed to balance characteristics between cohorts. The primary outcomes assessed were all-cause mortality and major bleeding events over a period of 1 month. Results A total of 4084 patients with TTP were identified, of whom 356 received caplacizumab with TPE and 3728 received TPE. After PSM, 2 balanced cohorts were created: C+P (Caplacizumab + TPE), P only (TPE), each comprising 356 patients. The average age in C+P and P only, after PSM, was 50.4 +/- 16.2 and 50.8 +/- 16.2 years, respectively. Females formed 65.7% of both C+P and P only, and males accounted for 28.4% and 28.7% of C+P and P only, respectively. 2.8% patients in C+P and 7.06% patients in P only had all-cause mortality at the end of 1 month. One-month all-cause mortality was significantly lower in C+P (2.8%) compared to P only (7.06%), with an odds ratio (OR) of 0.385 (95% CI 0.182 to 0.814, p=0.010). Upon assessment of bleeding outcomes, no statistically significant risk difference was noted between the 2 groups, with the odds ratio being 0.974 (95% CI 0.399 to 2.379, p=0.954). Conclusion Our large database study suggests that the addition of caplacizumab to standard TPE therapy is associated with a reduction in one-month all-cause mortality in patients with TTP, without a corresponding increase in major bleeding. However, given the lack of granular clinical data, such as contraindications to caplacizumab use, it is possible that patients who received TPE alone were inherently at a higher risk. As such, while these findings support the potential benefit of early caplacizumab use, they should be interpreted with caution and warrant prospective validation to confirm causality and assess long-term outcomes.

Abstract Use of Direct oral anticoagulants (DOAC) has increased exponentially since their introduction in 2010. Over the years they have been used without any systematic oversight, there have been ongoing concerns related to inappropriate or incorrect dosing, incorrect duration, and non-adherence to the drugs leading to adverse patient outcomes. Several factors have been attributed to DOAC non-adherence such as presence or fear of adverse effects, affordability, lack of perceived benefit, or concerns for reversibility. These barriers impact both primary (first time prescription) and secondary adherence (longer term prescription). Poor adherence can lead to adverse outcomes such as stroke, death, and higher medical resource utilization hence increased cost to the patient and payors. This study was conducted as a quality improvement (QI) initiative to improve the adherence to the prescription DOACS. We included patients 18 years old, a diagnosis of Atrial fibrillation and flutter who had a primary care provider writing the prescription for chronic DOAC therapy. Patients with venous thrombosis, pregnancy or who were institutionalized were excluded. The charts were reviewed for primary and secondary adherence using Surescripts functionality which can provide baseline information about prescription fills from the participating pharmacies. In the preimplementation phase we identified 1030 patients meeting inclusion criteria associated with the pilot sites within Mayo Clinic practice in preceding 6 months. For this population, Surescripts data found that the rates of primary and secondary adherence to be 84% and 86% respectively. We then implemented a 6-month pilot where we created workflow for a pharmacy led DOAC management unit. Herein, once a DOAC prescription was made for patients meeting inclusion criteria, they would enter a registry where the pharmacy team would review the charts for primary adherence by utilizing Surescripts data. If adherent these patients would be marked for follow up in the secondary adherence phase. However, if non adherent, the team would initiate intervention including engagement with patient online services which was further escalated to phone calls by the pharmacy team. Once the barriers were identified several remediation approaches were incorporated in the process, these included education about DOACS, medicare education, social work assistance to even switching to warfarin if necessary. The same process was followed for secondary adherence management. In the intervention cohort, we had 411 patients during the pilot, for primary adherence 87.1% of patients were adherent at baseline and an additional 3.4% became adherent with interventions, while 9.5% remain nonadherent. The primary causes of nonadherence, amongst 9.5%, were cost (86%) while 1% were concerned about side effects and 3% were concerned about lack of perceived benefit. The inverntions lead to impoved adherence in most patients and very few patients were transitioned to warfarin. For secondary adherence 92% of the patients were adherent without any intervention, an additional 4% of patients benefited with the interventions of the DOAC management unit with approaches to address cost considerations, while another 4% remained nonadherent where cost remained the primary reason for nonadherence, other reasons being lack of perceived benefit and side effect concerns. In summary, with active intervention, in our patient population we were able to improve the primary adherence from 84% to 91% and secondary adherence from 86% to 96% . Cost remained the biggest barrier for adherence followed by concern for side effects and lack of perceived benefit. Our engaged team of pharmacy staff was able to manage those concerns with the education and cost mitigation strategies leading to improved adherence. Based on previously published literature even a 5% difference in adherence to DOACS can make a huge difference in population-based outcomes. Our work was able to demonstrate a positive impact of the intervention, highlighting the need for a well-managed DOAC unit especially for complex care populations.

Abstract BackgroundWarm autoimmune hemolytic anemia (wAIHA) is clinically characterized by autoantibody-mediated (typically IgG) erythrocyte destruction at body temperature. While oral corticosteroids remain first-line therapy, rituximab is frequently used as a second-line agent, mostly supported with concomitant background medications, yet real-world evidence is limited and utilization is driven by published retrospective data and expert opinion. Real-world evidence characterizing its utilization patterns and outcomes is needed to inform clinical practice. AimsThis study aimed to: 1) describe treatment patterns among patients with newly diagnosed wAIHA, 2) evaluate rescue therapy use and healthcare resource utilization (HCRU) with current standard of care (SOC), 3) characterize patterns of rituximab use during the first year. MethodsNewly diagnosed wAIHA patients were identified from multiple US databases (MarketScan® Databases, IQVIA PharMetrics, Optum Clinformatics®, and Optum PanTher EHR). Patients with ≥1 wAIHA diagnosis code in any position (ICD-10-CM D59.11) from October 2020 to December 2023 and 12-month post-index follow-up were pooled together. Secondary wAIHA was defined with evidence of hematologic malignancy, solid tumors, immunodeficiencies, or autoimmune diseases in the 6 months prior to diagnosis. Patients on treatment were defined by use of oral corticosteroids (OCS), immunosuppressants (IST) (including azathioprine, mycophenolate mofetil, cyclosporine, cyclophosphamide, or tacrolimus) or rituximab. Initial therapy was defined as the first treatment occurring ≥ 30 days prior to diagnosis. Therapies were identified as combination when initialized within 21 days of each other. The regimen is considered active until all therapies are discontinued or a new agent is initiated. Rituximab regimens were classified as standard (4 weekly doses), alternative (2 doses given 2 weeks apart), single infusion, or extended (>4 infusions). Results Among 2,028 wAIHA patients (median age 62.0 years, 59% female), 55% had secondary wAIHA. Among patients on treatment (n=1,439), 64% received OCS monotherapy; rituximab and IST, alone or in combination with OCS were received by an additional 28% and 5% of patients, respectively. Overall, 80% of treated patients received OCS in some combination during initial therapy. Within 12 months following diagnosis, 1 in 4 patients required ≥ 1 blood transfusion, 1 in 8 required either erythropoiesis-stimulating agents (ESA) or intravenous immunoglobulin (IVIG). Within 12 months of diagnosis, 35% of patients required ≥ 1 hospitalization, average length of stay 3.8 days; AIHA was the principal diagnosis for 65% cases. Greater than 1/3 of patients (38%) had ≥ 1 emergency department visit. Almost all wAIHA patients (98%) had an all-cause outpatient visit with a median of 31 visits per patient. Among patients receiving rituximab in any line (n=613), median time from diagnosis to first administration was 59.0 days (interquartile range: 42.0-171.0). For the first course of rituximab, 62.2% of patients received a complete course (standard, alternative, or extended courses). Among patients with a completed course, 39.8% had >4 infusions without a 90-day gap between administrations with a median of 8 administrations within 12 months. Of the nearly 38% of patients that did not receive a complete course of rituximab, 33% received a single infusion. The use of rescue medications in the period prior to rituximab treatment was compared to the period after treatment (defined as 60 days to 12-months post final rituximab infusion). Use of blood transfusions was observed to increase from 9.0% to 32.8%, while the use of ESA and IVIG were stable at approximately 2% and 4%, respectively.Conclusion With current wAIHA SOC, US HCRU remains high, posing a financial burden in terms of the rate of outpatient and emergency department visits, and inpatient stays. Rituximab is an early line therapy post-diagnosis, typically in combination with OCS. No changes in rescue treatments, and an increase in blood transfusions after rituximab infusion suggest treatment gaps remain. Analysis reveals heterogeneity in rituximab dosing patterns including high deviation from standard dosing practices; this variance from published expert opinion along with common use of rescue therapies suggests continued need for standardized guidelines and development of targeted therapies in wAIHA.

Abstract Gentically driven classical hematological disorders have variable prevelance in the general population, sickle cell disease (SCD) being the most common at 30 per 100,000 people, hemophilia (hem) 12 per 100,000 people for hemophilia A, 4 per 100,000 people for hemophilia B, and thalassemia (thal) at 18 per 100,000 people, with regional and racial-ethnic variation across US. There has been a dramatic change in the therapeutic landscape for these diseases in the past decade. Several new drugs and biological therapies including gene therapies have been approved for each of these disorders. Clinicians utilize different information sources to assimilate this rapid influx of knowledge into their practice; in the era of social media, the readership is not limited to original publications. As research visibility influences clinical adoption and patient awareness, it is vital to understand the modalities of engagement of gene therapy studies by clinicians and patients. An altmetric score is a simple and effective way of understanding engagement with research across platforms. It. measures digital attention received by research output, allowing us to gauge the social impact of an article. We sought to compare the impact and scientific dissemination of publications of gene therapy trials in these 3 diseases by capturing the number of traditional citations and Altmetric assessment score. Methods: A PubMed search was conducted for gene therapy trials in SCD, Hem and Thal between 2015 to 2025 and citations were captured using metrics from Google Scholar. The Altmetric assessment score for each article was determined using www.altmetric.com which includes any mention by news outlets, blogs, Twitter (X), Facebook, and Mendeley reader counts. Descriptive statistics are reported for the number of citations as well as the Altmetric assessment score in each of the disease categories. Median and range were reported due to the skewed distribution of the data. Comparison across groups for significance was done using the Kruskal-Wallis test. Results: The number of qualifying studies for SCD (n=7) were lower than hem (n=21) and thal (n=12). Median PubMed citations were highest for SCD (147, range 4-436), followed by hem (26, range 0-828) and thal (42, range 0-801), though the difference was not statistically significant (p=0.4) Altmetric scores demonstrated differences that approached statistical significance with p=0.08: SCD had the highest median (422, range 4-770), driven by exceptional Twitter engagement, while hem and thal medians were 20 (range 0-2038) and 49 (range 0-825) respectively. Twitter/X was the leading social platforms for all diseases. Scholarly interest measured by Mendeley readers was also highest for SCD (median 195, range 11-542), with thal and hem medians at 50 (range 0-773) and 64 (range 8-646) respectively. Conclusion: We observed substantial variability in both academic and social media dissemination of practice-changing SCD, Hem and thal gene therapy studies, though not statistically significant likely due to the limited number of studies overall. Higher citation count and Altmetric score in SCD studies may reflect recent robust dissemination efforts by ASH and patient advocacy groups. Intentional promotion across academic and media platforms increases awareness, helping with early dissemination of updated information that could have significant impact on patient outcomes.

Abstract Introduction: Hemophagocytic lymphohistiocytosis (HLH) is a rare and life-threatening hyperinflammatory syndrome resulting from immune dysregulation. While diagnostic and therapeutic guidelines (e.g., HLH-2004) are established for pediatric HLH, adult-onset HLH remains both a diagnostic and therapeutic challenge, with limited characterization and sparse outcome data. Here, we conduct a single-institution retrospective cohort study to evaluate clinical features, causes, and prognostic markers associated with outcomes. Methods: We performed a retrospective chart review of all adult patients (≥18 years) first diagnosed with HLH across all Mayo Clinic sites between January 1, 2006 and July 1, 2025. Diagnosis was based on hematologist adjudication at time of care. Patients with prior diagnoses of HLH or HLH related to T-cell therapy (CAR-T) were excluded. Clinical and laboratory variables were extracted from electronic health records. The primary outcome was overall survival (OS), defined as time from diagnosis of HLH to death or last follow-up. OS was estimated using the Kaplan-Meier method. Multivariable analysis to identify independent predictors of mortality was performed using a Cox proportional hazards regression model, adjusting for age at diagnosis, sex, HLH etiology, and relevant clinical and laboratory parameters. Model discrimination was assessed using Harrell's concordance index (c-index). Descriptive statistics were summarized using median and interquartile range (IQR) for continuous variables and frequency with percentages for categorical variables. Results: A total of 203 patients diagnosed with HLH were included. The median age at diagnosis was 55 years (range 18-86). The cohort was predominantly male (n=130, 64%) and White (n=163, 80%). The majority of diagnoses occurred in more recent years, with 75 patients (36.9%) diagnosed between 2021–2025, 61 (30.0%) between 2016–2020, and 43 (21.2%) between 2011–2015. Prior to HLH diagnosis, 86 patients (42.4%) had a history of malignancy, 55 (27.1%) had a history of autoimmune disease, and 28 (13.8%) had a history of transplant. The most common HLH etiologies were malignancy-associated (n=99, 49%), infectious (n=51, 25%), and autoimmune (n=24, 12%), in addition to idiopathic (n=6, 3%), post-transplant lymphoproliferative disease or graft-vs-host disease (n=3, 1%), drug-induced (n=3, 1%), or primary HLH (n=1, 0.5%). The most common malignancies included T-cell (n=35, 17.2%) and diffuse large B-cell lymphoma (n=24, 11.8%). Forty-four cases (21.7%) were associated with Epstein-Barr virus (EBV), either through overt viremia or malignancies with EBV positivity on pathology report. Five patients (2.5%) had COVID-19–associated HLH and 2 (1.0%) were attributed to immune checkpoint inhibitors. Overall, 164 patients (80.8%) met the HLH-2004 diagnostic threshold of ≥5 out of 8 criteria, while 196 (96.6%) met the HLH-2024 ≥4 of 7 modified criteria. Among patients with available information, elevated ferritin was present in 197 of 198 (99.5%), fever in 185 of 202 (91.6%), cytopenias involving ≥2 lineages in 184 of 199 (92.5%), hemophagocytosis in 156 of 182 (85.7%), elevated sCD25 in 117 of 121 (96.7%), hypertriglyceridemia in 117 of 186 (62.9%), hypofibrinogenemia in 118 of 196 (60.2%), splenomegaly in 125 of 188 (66.5%), and impaired NK cell activity in 16 of 25 (64.0%). In addition, hepatomegaly was noted in 57 of 195 patients (29.2%). After a median follow-up of 58.3 months, median OS was 62 days (95% CI 40-110) from day of diagnosis. In a multivariable Cox proportional hazards model with moderate discrimination (c-index=0.75), after adjusting for age at diagnosis and sex, T-cell lymphoma–driven HLH (HR 1.78, 95% CI 1.05–3.00, p = 0.03) and elevated blood urea nitrogen (median 51.0 mg/dL, IQR 27.9–99.1; HR 2.07, 95% CI 1.37–3.13, p < 0.005) were independently associated with shorter OS Conclusions: In this cohort of adult patients with HLH, malignancy was found to be the most common trigger for HLH. T-cell malignancies in particular were strongly correlated with poor survival. Elevated blood urea nitrogen at diagnosis also identified patients at significantly higher risk of death, consistent with other critical illnesses that can present with multi-organ failure. These findings emphasize the need for adult-specific prognostic tools incorporating disease etiology and organ dysfunction markers to better stratify risk and guide clinical management in adult-onset HLH.