Hemophagocytic Lymphohistiocytosis

Hemophagocytic lymphohistiocytosis (HLH) is a severe hyperinflammatory syndrome caused by aberrant activation of macrophages and cytotoxic T cells ^[1]A1c. The syndrome represents immune dysregulation rather than a single tumor entity, so clinicians should distinguish the HLH process from an underlying hematologic malignancy that can trigger it ^{[1]A1c[2]B2b[12]D5}.

Nomenclature

Also called: HLH; familial hemophagocytic lymphohistiocytosis (FHL); primary or genetic HLH (pHLH); secondary or acquired HLH (SHL); macrophage activation syndrome (MAS); MAS-HLH; and malignancy-associated HLH. FHL denotes genetically driven disease, whereas secondary HLH denotes an acquired syndrome associated with infection, malignancy, drugs, or autoimmune or autoinflammatory disease. When autoimmunity or autoinflammation triggers secondary HLH, the syndrome is termed MAS or MAS-HLH ^[1]A1c[12]D5.

Classification

The revised histiocytosis classification places HLH and MAS within a dedicated disease group alongside Langerhans-related, cutaneous and mucocutaneous, malignant histiocytoses, and Rosai-Dorfman disease ^[8]D5. For clinical use, the etiologic categories are more informative than a severity stage or tumor grade:

HLH-2004 refers to the pediatric diagnostic framework commonly applied in practice, while HLH-94 refers to the earlier treatment protocol; neither term denotes a disease stage ^{[1]A1c[6]B3b[9]D5}. The criteria were developed for children, are not validated for adults, and can overlap clinically with sepsis or multiple organ dysfunction syndrome ^[1]A1c.

This classification matters because inherited disease, acquired trigger-associated disease, MAS, and malignancy-associated HLH require different diagnostic interpretation, urgency, and treatment priorities ^[1]A1c[2]B2b. The next section explains how defects in immune regulation and cytotoxic function produce these overlapping phenotypes.

Pearl: In an adult with suspected HLH, label the syndrome and the trigger separately: “secondary HLH” is not an adequate final classification when MAS or an underlying hematologic malignancy is present.

The classification becomes clinically useful when traced to the shared failure of immune restraint. HLH develops when impaired cytotoxic control and an infectious, malignant, rheumatic, or other inflammatory stimulus drive immune activation beyond a critical threshold, producing a convergent hyperinflammatory phenotype ^[29]D5.

From genetic or clonal defect to immune persistence

Inherited HLH commonly reflects defects in cytotoxic lymphocyte function. Pathogenic variants affect perforin-mediated killing or the trafficking and release of cytotoxic granules, including defects involving PRF1 , UNC13D , STX11 , RAB27A , and related degranulation machinery ^{[24]C4[35]D5[37]D5}. The consequence is failure to eliminate activated or infected target cells. Persistent antigenic stimulation therefore sustains activated CD8-positive T cells and natural killer cells, while impaired negative feedback permits continued activation of macrophages and monocytes ^[11]D5[29]D5.

Acquired disease reaches the same inflammatory state through a different initiating lesion. EBV -associated immune dysregulation can arise with defects in CD27/CD70 signaling, which impair the generation of EBV-specific T cells and CD8-positive T-cell effector function ^[21]B2b. In T-cell or natural-killer-cell lymphoma, somatic alterations in the FAS pathway are associated with increased HLH risk, linking defective apoptotic regulation to a clonal trigger ^[38]D5. Germline HAVCR2 variants provide another example of host susceptibility, with the Y82C variant associated with HLH or HLH-like systemic illness in subcutaneous panniculitis -like T-cell lymphoma ^[39]D5.

Cytokine amplification and tissue injury

The central amplifier is sustained T-cell production of interferon gamma (IFN-γ), which activates macrophages and other myeloid cells. Activated myeloid cells then intensify cytokine and chemokine signaling, including interleukin-6 , interleukin-18 , tumor necrosis factor , and the IFN-γ-regulated chemokines CXCL9 , CXCL10 , and CXCL11 ^{[11]D5[29]D5[31]D5}. These mediators signal through JAK-STAT pathways, especially JAK1/JAK2-dependent STAT1 activation, while STAT3 and STAT5 signaling also contributes to secondary HLH inflammation and neutrophil activation ^[31]D5[32]D5.

Neutrophils and monocytes are active participants rather than passive bystanders. In a secondary HLH model, depletion of either population reduced cytokine levels, and neutrophil extracellular trap formation was accompanied by inflammatory cytokine production and tissue injury ^[31]D5. This myeloid amplification helps explain how a lymphocyte-control defect becomes systemic inflammation, hepatosplenomegaly, cytopenias, neurologic injury, and multiorgan failure ^[24]C4[30]D5.

This cascade directly connects mechanism to the findings that later sections must interpret: persistent fever and organ enlargement, falling blood counts, hyperinflammatory laboratory abnormalities, and progressive hepatic, neurologic, or multiorgan injury ^{[11]D5[24]C4[30]D5}. The initiating trigger and host defect determine which node dominates, so epidemiologic and etiologic context remains essential when evaluating the phenotype.

Pearl: A cytotoxicity defect is not required for the phenotype, but failure to terminate immune activation is the mechanistic common denominator; interpret IFN-γ-driven inflammation and myeloid activation together rather than treating hemophagocytosis as an isolated event ^[29]D5[30]D5.

Inherited immune-regulatory defects, acquired disease, and external triggers create distinct pretest probabilities, but the supplied evidence does not provide a population-based incidence, prevalence, ethnic distribution, sex distribution, geographic distribution, or temporal trend for hemophagocytic lymphohistiocytosis (HLH) ^{[50]B3b[51]D5}. Available cohorts are clinically selected. In an adult Alberta cohort, the median age was 46 years, but this does not define the age distribution of HLH overall ^[50]B3b.

Etiologic spectrum

The strongest heritable signal comes from defects in cytotoxic lymphocyte function and immune regulation. Classic genetic disorders include pathogenic changes in PRF1 , UNC13D , STXBP2 , and STX11 , alongside granule or pigment abnormality genes, X-linked lymphoproliferative disease genes, and other inborn errors of immunity ^[51]D5. Biallelic CD27 or CD70 mutations particularly predispose to Epstein-Barr virus -associated immune dysregulation, including chronic viremia, lymphoproliferation, malignancy, and HLH ^[21]B2b.

Acquired HLH occurs in the setting of infections, malignancies, and rheumatologic or autoinflammatory disease, where it is often termed macrophage activation syndrome ^[51]D5. In CD27/CD70 deficiency, 90% of patients were EBV-positive at diagnosis, 70% had lymphoproliferation, and 43% had lymphoma, but these proportions come from a defined inherited-immunodeficiency cohort and must not be generalized to all HLH ^[21]B2b. Chronic active EBV disease can involve infected hematopoietic stem cells and multiple lymphoid and myeloid lineages, providing a clonal-acquired disease model that differs from inherited HLH ^[44]C4.

Risk factors and measurable predictors

The available literature does not validate a general-population HLH risk score or calculator. Most reported etiologic associations lack multivariable odds ratios, risk ratios, or hazard ratios; the exception below concerns progression among patients who already have partial HLH, not initial disease onset ^[46]B3b.

Low C3 is therefore a progression marker rather than an established cause of HLH. It was also independently associated with early death, with HR 1.64 and no confidence interval reported, but this prognostic finding belongs to risk assessment after HLH has developed rather than to etiologic screening ^[46]B3b.

The next section builds on this etiologic heterogeneity by showing how fever, cytopenias, organ involvement, and inflammatory-marker abnormalities appear at the bedside ^[51]D5.

Pearl: Do not treat a negative genetic result as evidence against inherited susceptibility: the cited cohorts identify broad genetic and acquired pathways, while the available evidence does not quantify the population risk of any single trigger ^[51]D5.

The clinical phenotype follows the intensity, duration, and context of immune activation, so the bedside picture ranges from an acute febrile multisystem illness to a slower, atypical presentation that obscures recognition ^[51]D5. The clinician should obtain a focused history for persistent fever, constitutional decline, bleeding, neurologic symptoms, gastrointestinal symptoms, and features of an underlying infection, malignancy, rheumatic disease, or immune deficiency.

Symptoms

Patients commonly report the following symptoms:

• Persistent or unremitting fever, often with constitutional deterioration ^{[51]D5[69]B3b[70]B3b}
• Fatigue, poor intake, and weight loss, when reported as part of systemic inflammatory illness ^{[51]D5[70]B3b}
• Headache, altered behavior, reduced consciousness, or seizures when the central nervous system is involved; neurologic symptoms are the initial manifestation in 30% of children with CNS involvement, and seizures occur in 70% ^[68]C4
• Dyspnea or other respiratory symptoms, which occur in over half of children with secondary HLH in one cohort ^[69]B3b
• Abdominal fullness or discomfort associated with hepatosplenomegaly ^{[58]B3b[60]C4}
• Diarrhea or other colitis symptoms, particularly in familial HLH type 5 ^[60]C4
• Mucosal or other bleeding, especially in familial HLH type 5 or platelet-granule disorders ^[59]C4[60]C4

Signs and measured abnormalities

The characteristic findings are systemic inflammation, reticuloendothelial enlargement, cytopenia, coagulopathy, and progressive organ injury ^{[51]D5[69]B3b}.

Onset, progression, and phenotypic variants

Onset is heterogeneous. Familial disease can present from 2 months to 17 years, whereas hypomorphic variants in adult familial HLH are associated with later-onset symptoms and a more indolent course ^{[58]B3b[60]C4}. Secondary disease often presents as an acute febrile illness during infection, malignancy, or rheumatic inflammation; in acute myeloid leukemia treated with intensive chemotherapy, pulmonary or neurologic symptoms, liver abnormalities, prolonged pancytopenia, and hepatomegaly distinguish affected patients ^[66]D5. CNS-predominant disease can begin before the systemic phenotype, and more than half of affected children in one cohort were initially evaluated for another diagnosis ^[68]C4.

Red flags and atypical presentations

Urgent escalation is warranted when fever coexists with rapidly worsening cytopenias, bleeding, jaundice or liver dysfunction, coagulopathy, neurologic change, respiratory compromise, or multiorgan failure ^{[51]D5[64]D5[66]D5}. Atypical presentations include isolated neurologic disease, colitis, bleeding disorders, hypogammaglobulinemia, or infection-associated inflammation in a patient with profound T-cell deficiency ^{[60]C4[65]D5[68]C4}. These patterns should not be dismissed because hemophagocytosis or a complete classic phenotype is absent; the next section integrates the relevant blood, marrow, flow, biomarker, and molecular findings ^[51]D5[65]D5.

Pearl: When fever is accompanied by a rapidly evolving combination of cytopenia, splenomegaly, liver or coagulation abnormalities, and neurologic or respiratory decline, reassess repeatedly for HLH even if the trigger or presentation initially suggests sepsis, malignancy, or isolated CNS disease ^[65]D5[68]C4.

Once the inflammatory phenotype is recognized, diagnosis depends on a coordinated hematologic workup rather than on hemophagocytosis alone. The clinician should interpret the peripheral smear, marrow, immune-function studies, and molecular testing as one diagnostic engine, while assessing whether the combination, extent, progression, and otherwise unexplained nature of abnormalities support HLH. In adults, the commonly applied HLH-2004 criteria are not validated, and malignancy, infection, sepsis, and treatment toxicity can reproduce several of their features ^{[1]A1c[72]A1c}.

Unified hematology diagnostic engine

The initial laboratory and tissue evaluation should answer two questions simultaneously: is there evidence of HLH-associated immune activation, and is an occult hematologic neoplasm or inherited immune defect present? The core tests and their evidentiary limitations are summarized below.

The blood film is therefore a safety screen, not a rule-in test. In every patient with HLH, peripheral blood and marrow should be screened for blasts, while suspicious lymph nodes or cutaneous lesions should undergo biopsy; imaging is escalated according to the likelihood of an underlying malignancy ^[72]A1c. A negative marrow for hemophagocytosis does not exclude HLH, and a positive result does not establish it ^[72]A1c.

Flow cytometry and molecular confirmation

Flow cytometry has a dual role. It can identify an abnormal lymphoid or myeloid population suggesting malignancy, and it can assess inherited immune dysfunction through NK/T-cell degranulation and expression of proteins such as perforin and SAP. When functional testing identifies a defect, the next step is targeted or broader genetic testing to define the causal lesion, because the molecular result can distinguish a hereditary predisposition from an acquired inflammatory syndrome and inform recurrence-risk assessment ^{[72]A1c[6]B3b}.

The molecular differential should remain open even in adults. Pathogenic PRF1 variants have been reported in adult-onset HLH, and defects involving MAGT1 , ITK , or CD27 can associate EBV susceptibility, lymphoproliferation, malignancy, and HLH ^{[72]A1c[33]C4[34]C4}. In male patients with lymphoma and EBV-driven HLH, X-linked lymphoproliferative syndrome testing deserves particular consideration ^[72]A1c.

Interpreting inflammatory biomarkers

Soluble CD25 is the best-supported single diagnostic assay in the supplied adult evidence, but it remains a marker of T-cell activation rather than a disease-specific result. Levels of 2400 U/mL or less helped rule out HLH, whereas levels above 10 000 U/mL helped rule it in, with 100% sensitivity and 93% specificity, respectively; the assay was less specific in adults with other T-cell-activating disorders ^[75]B3b. In adults with hematologic malignancies, combined soluble CD25 above 3900 U/mL and ferritin above 1000 ng/mL identified HLH-2004-defining features with 84% sensitivity and 81% specificity ^[2]B2b. These thresholds support probability assessment, not replacement of tissue, immune-function, and molecular evaluation.

No single gold-standard test is established in the supplied evidence. The most defensible diagnostic standard is therefore a concordant clinical-inflammatory phenotype plus supportive laboratory findings, exclusion or identification of an active trigger, and targeted hematologic, cellular, and genetic testing. The resulting findings, particularly cytopenias, ferritin and soluble CD25 elevation, marrow interpretation, blasts or clonal populations, and evidence of inherited cytotoxic dysfunction, provide the inputs for the staging and prognostic assessment that follows ^{[72]A1c[75]B3b}.

Pearl: Do not use absent marrow hemophagocytosis to exclude HLH; when cytopenias or inflammation remain unexplained, repeat marrow evaluation and pair it with flow-cytometric immune testing and molecular analysis ^[72]A1c.

Risk stratification begins after the diagnostic phenotype is established, because HLH has no single anatomic stage that reliably predicts outcome across familial, malignancy-associated, infectious, and rheumatic disease. The available evidence instead supports integrating trigger, inflammatory burden, organ dysfunction, and early treatment response, with separate interpretation for familial HLH , secondary HLH , macrophage activation syndrome , and malignancy-associated HLH ^[83]D5.

Prognostic indices and dynamic assessment

The optimized HLH inflammatory (OHI) index is the best-supported malignancy-focused prognostic tool in the supplied evidence. It combines soluble CD25 >3900 U/mL with ferritin >1000 ng/mL. In adults with hematologic malignancies, the combination identified HLH-2004-defining features with sensitivity 84% and specificity 81%, and predicted mortality with a hazard ratio of 4.3 (95% CI, 3.0-6.2) ^[2]B2b. In a lymphoma cohort, OHI-positive patients had 1-year overall survival of 33% versus 81% for OHI-negative patients, and 3-year overall survival of 23% versus 80% ^[79]B3b. Traditional lymphoma IPI underestimated this excess risk, so the OHI result should supplement, not replace, disease-specific scores ^[79]B3b.

The HScore and HLH-2004 criteria remain useful classification frameworks, but their prognostic performance is context-dependent. In nonmalignancy-associated adult HLH, an 8-week treatment response was the strongest survival predictor; a model assigning low-risk, intermediate-risk, and high-risk groups reported 5-year overall survival of 92.1%, 36.8%, and 18.0%, respectively ^[80]B3b. In etoposide -treated children and young adults, day 7 markers outperformed pretreatment or later measurements. Failure of soluble CD25 to improve by 25% and the presence of at least 3 unfavorable markers strongly predicted pre-transplant mortality ^[73]C4.

Prognostic factors

The following table translates the evidence into bedside risk domains.

Recovery, relapse, and late outcome

Recovery is often measured over weeks rather than hours. In familial HLH, remission occurred in a median of 76 days, with a range of 15-705 days ^[24]C4. Malignancy-associated HLH remains particularly lethal: 1-year survival was 21% in one adult cohort, while Swedish population data showed 2-year survival of 25% ^{[23]B3b[81]D5}. Among familial HLH patients with the A91V/pLOF genotype, recurrent inflammatory episodes occurred in 7 of 39 patients, demonstrating that remission does not eliminate recurrence risk ^[82]D5.

Long-term outcome also depends on the trigger and its treatment. In malignancy-associated HLH, OHI-positive patients most often died from multiorgan failure, whereas OHI-negative patients more often died from lymphoma progression ^[79]B3b. These distinctions should guide follow-up intensity and trigger-specific surveillance, while acute deterioration, organ failure, or failure of early response requires immediate transition to the emergency pathway in the next section.

Pearl: Do not let a reassuring disease-specific score overrule rapidly worsening inflammation: reassess dynamic markers, especially soluble CD25, ferritin, C3, organ function, and early treatment response, because hyperinflammatory risk can exceed the risk predicted by the underlying disease score ^{[46]B3b[73]C4[79]B3b}.

Risk stratification now determines whether treatment can await trigger clarification or must begin immediately. Secondary HLH in the intensive care setting requires anti-inflammatory treatment before invasive organ failure, because all deaths in one ICU cohort occurred among patients receiving invasive organ support ^[88]C4.

Emergency protocol

1. Assess severity and identify the trigger immediately. Admit patients with progressive cytopenias, coagulopathy, hepatic injury, persistent fever, or evolving organ dysfunction to a setting capable of intensive monitoring. Investigate infection, malignancy, rheumatic disease, and treatment-related causes in parallel. The clinician should also consider HLH mimics such as visceral leishmaniasis , because marrow examination identified Leishmania in two children initially presenting with HLH-like features ^[93]C4.

2. Start a trigger-directed bridge without waiting for every complementary result. Treat a documented infection with appropriate antimicrobial therapy while controlling hyperinflammation. In reported VZV-associated HLH, multimodal treatment combined acyclovir, intravenous immunoglobulin , therapeutic plasma exchange , dexamethasone, and reduced-dose etoposide 75 mg/m² ^[92]C4. The ICU evidence supports symptomatic anti-inflammatory treatment as an emergency while etiological treatment is being established ^[88]C4.

3. Use the treatment of choice according to severity and response. For fulminant or refractory secondary HLH, the best-supported emergency strategy in the supplied evidence is trigger control plus corticosteroid-based immunomodulation, with early reduced-dose etoposide escalation when fever, cytopenias, or organ dysfunction persist. In the VZV case, etoposide 75 mg/m² was given for 2 doses, with dexamethasone 5 mg/m²/day, and temperature normalized within 48 hours ^[92]C4. Ruxolitinib is a non-myelosuppressive investigational option, administered orally at 15 mg twice a day in a five-patient pilot study; all patients had symptom and laboratory improvement, but this small uncontrolled study does not establish it as standard emergency therapy ^[4]C4.

4. Monitor continuously for trajectory, not isolated values. Reassess fever, blood counts, coagulation, liver injury, ferritin, inflammatory markers, renal function, neurologic status, and respiratory or circulatory support requirements. The supplied studies do not specify a validated laboratory monitoring interval. Persistent lymphocytopenia can continue after clinical remission and therefore warrants follow-up of immune recovery ^[92]C4.

5. Treat complications while controlling HLH. DIC, bleeding, acute liver failure , respiratory failure, renal failure, infection, and treatment-related cytopenias require concurrent critical-care management ^[87]D5[92]C4. After CAR T-cell therapy , distinguish HLH-like toxicity from cytokine release syndrome, immune effector cell-associated neurotoxicity syndrome, sepsis, and meningitis ^[98]D5.

Treatment failure and avoidable errors

• If the trigger remains active or fever and cytopenias persist, escalate promptly to reduced-dose etoposide rather than waiting for invasive organ support ^[92]C4.
• Do not delay anti-inflammatory treatment until microbiology or histology is complete when the patient is deteriorating ^[88]C4.
• Do not apply intensive immunochemotherapy without adapting it to an uncontrolled infection, because unmodified treatment can worsen pathogen dissemination ^[92]C4.
• Do not assume that HLH-like deterioration after CAR T-cell therapy represents uncomplicated sepsis ^[98]D5.

The evidence supplied contains no named NCCN, ASCO, or ESMO guideline recommendation or recommendation-strength classification; the treatment hierarchy above therefore reflects pilot, cohort, case-report, and review evidence rather than a formal guideline grade ^{[4]C4[88]C4[92]C4[102]D5}. Definitive trigger control and longer-term disease-specific therapy follow once the emergency phase is stabilized.

Pearl: In secondary HLH, a falling platelet count with rapidly rising ferritin should prompt treatment escalation before invasive organ support becomes necessary ^[88]C4.

Once acute hyperinflammation is controlled, treatment must remain response-adapted and directed at the underlying trigger. The 2019 Histiocyte Society adult recommendations are expert recommendations rather than trial-based treatment standards, because adult HLH evidence is limited and pediatric protocols can cause overtreatment and toxicity ^[1]A1c.

1. Reassess severity and treatment intent

The clinician should first determine whether ongoing inflammation reflects persistent HLH, an uncontrolled trigger, treatment toxicity, or a combination. This distinction determines whether to intensify immunomodulation, treat the trigger, defer cytotoxic therapy, or proceed toward definitive cellular therapy.

• Review clinical trajectory, cytopenias, organ dysfunction, ferritin, fibrinogen, soluble CD25, and LDH as response markers ^[72]A1c.
• Reassess for infection, particularly when HLH occurs during chemotherapy or after immunosuppressive treatment ^[72]A1c.
• In malignancy-triggered HLH, decide case by case whether the initial emphasis is HLH-directed therapy, malignancy-directed therapy, or a combined approach ^[72]A1c.
• In HLH during chemotherapy, consider postponing subsequent chemotherapy blocks or interrupting maintenance therapy, except when the malignancy has relapsed ^[72]A1c.

2. Initiate, advance, and tailor disease-directed therapy

The Histiocyte Society recommends individualized treatment in adults, including dose reduction and tailoring of treatment duration when pediatric-derived protocols are used ^[1]A1c. Conventional disease-directed agents include glucocorticoids , cyclosporine , etoposide , and immunoglobulin , while anti-thymocyte globulin and alemtuzumab have more limited supporting experience ^[72]A1c. Treatable infections require rigorous therapy, and anti-infective prophylaxis and surveillance for fungal infection, Pneumocystis jirovecii, EBV, and CMV should be strongly considered during active disease ^[72]A1c.

For highly replicative EBV-associated disease, rituximab may be considered, particularly when a CD20-positive B-cell malignancy is present ^[72]A1c. In autoimmune or autoinflammatory disease, interleukin-1 blockade can be used with glucocorticoids, cyclosporine , and etoposide ^[72]A1c.

The strongest targeted evidence concerns primary HLH. In a phase 2-3 study, emapalumab produced a response in 63% of previously treated patients and 65% of all patients who received an infusion; 70% of previously treated patients proceeded to transplantation, and severe infections occurred in 10 patients ^[112]B2b. Real-world treatment across 46 patients showed normalization of fibrinogen in 97.4%, platelets in 84.8%, and absolute neutrophil count in 88.9%; 12-month survival from emapalumab initiation was 73.1% ^[5]B2b.

Ruxolitinib remains investigational rather than an established universal replacement for conventional therapy. In an adult pilot study of five patients receiving 15 mg twice daily, 2-month overall survival was 100% and all patients had symptom or laboratory improvement; one serious adverse event, grade 4 febrile neutropenia, occurred ^[4]C4. In a pediatric retrospective series, a ruxolitinib-based regimen produced an objective response in all patients within the first 8 weeks, but these findings do not establish comparative efficacy ^[113]B3b.

3. Monitor response and define treatment failure

Serial response assessment is essential because early failure predicts poor outcome and should trigger an alternative strategy. In patients treated with etoposide-based therapy, day-7 measurements were more informative than pretreatment or later assessments; less than 25% improvement in soluble CD25, together with platelet count, absolute lymphocyte count, and blood urea nitrogen, identified high-risk patients ^[73]C4.

Treatment failure should prompt urgent review of the trigger, infection control, malignancy response, drug toxicity, and candidacy for salvage therapy or definitive cellular therapy. For refractory or relapsed EBV-associated HLH, compassionate-use nivolumab produced responses in 6 of 7 adults, with complete remission in 5 and eradication of plasma and cellular EBV DNA in 4 ^[105]C4. These data are retrospective and do not establish routine use.

4. What not to do

• Do not apply a fixed pediatric protocol to every adult without individualized adjustment ^[1]A1c.
• Do not assume that identifying an infection excludes malignancy as a co-trigger ^[72]A1c.
• Do not continue chemotherapy automatically during HLH arising during treatment; reassess whether postponement or interruption is safer ^[72]A1c.
• Do not treat malignancy-associated HLH with HLH-directed therapy alone when the malignancy remains uncontrolled; once HLH resolves, a neoplasm-directed protocol is required ^[72]A1c.
• Do not use tocilizumab as an assumed treatment for malignancy-associated HLH; its effectiveness in that context remains unclear ^[72]A1c.

Management flowchart

Severity and response assessment → identify and treat infection or malignancy → choose individualized HLH-directed, trigger-directed, or combined therapy → repeat clinical and laboratory response assessment, including early soluble CD25 and cytopenia trends → if refractory or relapsed, use specialist salvage therapy and proceed to definitive cellular therapy when indicated ^{[1]A1c[72]A1c[73]C4}.

Primary or relapsing genetic disease generally requires hematopoietic stem-cell transplantation for cure, while emapalumab, ruxolitinib, and other agents serve principally as disease control or bridges in the evidence provided ^{[7]B3b[5]B2b[113]B3b}. Detailed transplant selection, conditioning, and post-transplant management are addressed in Hematopoietic Cell Transplantation & Cellular Therapy .

Pearl: A falling platelet count or inadequate early soluble CD25 response should trigger a search for persistent trigger activity and an immediate response-adapted plan, not passive continuation of the same regimen ^{[72]A1c[73]C4}.

The preceding disease-control strategy becomes curative-intent hematopoietic stem-cell transplantation when the phenotype signals a durable risk of recurrence. The supplied literature does not provide a named NCCN, ASCO, or ESMO guideline recommendation with a Category or strength grade; decisions therefore rest on genetic findings, disease persistence, central nervous system inflammation, and transplant fitness ^[114]B2b.

Protocol

First, stratify transplant risk and urgency. Allogeneic HCT is indicated particularly for familial or genetic HLH, relapsing or severe and persistent disease, persistent cytolytic immune defects, or central nervous system inflammation ^[114]B2b. The diagnosis should be genetically reviewed before transplantation because children without verified familial HLH had lower post-transplant survival than those with verified disease, and the authors caution that familial HLH requires careful investigation before HCT ^[7]B3b. Whole-exome sequencing can identify inherited immune defects and unexpected donor-relevant variants, which can alter the transplant plan ^[114]B2b.

Second, control inflammation while advancing toward HCT. Emapalumab, an anti-IFN-γ antibody, provides a bridge for primary HLH when conventional therapy is inadequate. In a phase 2-3 study, 63% of previously treated patients responded and 70% proceeded to transplantation; severe infections occurred in 10 patients, and disseminated histoplasmosis led to treatment discontinuation in 1 patient ^[112]B2b. A retrospective study found that emapalumab given before reduced-intensity conditioning was associated with less mixed chimerism, 48% versus 77%, and higher intervention-free survival, 73% versus 43%; NNT is not calculable from the reported data ^[42]B3b.

Third, use allogeneic HCT as the treatment of choice when durable cure is required. Reduced-intensity conditioning with melphalan , fludarabine , and intermediate-timing alemtuzumab , followed by graft-versus-host disease prophylaxis with cyclosporine and methylprednisolone , produced 80.4% one-year overall survival in a prospective phase 2 study, although secondary graft failure or a second intervention occurred frequently ^[41]B2b. Treosulfan-based conditioning with fludarabine and alemtuzumab , with or without thiotepa , also showed favorable outcomes in a retrospective HLH cohort, but donor lymphocyte infusion or second transplantation was more frequent after HLA-mismatched transplantation ^[48]C4. Matched or single-locus mismatched related or unrelated donors were used in the prospective reduced-intensity study ^[41]B2b. Umbilical cord blood is an alternative donor source when a matched related donor is unavailable, but the cited prospective series included only 1 child with HLH ^[43]B2b.

Fourth, monitor engraftment and immune toxicity. The evidence supports serial donor-chimerism assessment because mixed chimerism, graft failure, donor lymphocyte infusion, and second HCT are clinically consequential after reduced-intensity transplantation ^[41]B2b. The retrieved studies do not report a universal laboratory schedule. Monitor for acute and chronic graft-versus-host disease , viral infection, veno-occlusive disease, neurologic sequelae, and recurrent inflammation, with the interval individualized by the transplant program ^{[25]C4[41]B2b[48]C4}.

Finally, reserve engineered cellular therapy for a defined malignancy indication, not routine HLH eradication. The retrieved evidence does not establish autologous CAR-T or bispecific antibody therapy as standard treatment for HLH itself. CAR-T can instead produce IEC-HS, a rapidly progressive HLH-like toxicity associated with CRS; pooled incidence was 1.6%, with HLH-related mortality of 0.7% ^[108]B2a. In pediatric B-ALL, HLH-like toxicities after tisagenlecleucel were associated with one-year overall survival of 25.7% versus 80.1% without these toxicities ^[47]B3b. CAR-T studies also report CRS, ICANS, infections, and HLH-like syndromes as major monitoring priorities ^{[115]B2a[117]C4}. The cited evidence does not support a standard cellular-therapy dose or treatment-failure algorithm for HLH.

What not to do

• Do not proceed to HCT solely from an HLH phenotype without reviewing familial HLH and broader inherited immune defects ^{[7]B3b[114]B2b}.
• Do not assume reduced-intensity conditioning eliminates graft failure or the need for donor lymphocyte infusion ^[41]B2b.
• Do not use CAR-T or bispecific antibodies as routine HLH-directed therapy outside a defined malignancy protocol, because the retrieved evidence does not establish that indication ^{[108]B2a[117]C4}.

The next management pillar addresses transfusion, anticoagulation, and cytoreduction, which remain distinct from transplant conditioning, graft complications, and cellular-therapy toxicities.

Pearl: Persistent inflammation is not merely a reason to intensify treatment, it is a reason to accelerate genetic review and transplant planning, because disease status and donor-derived chimerism directly determine whether curative-intent HCT succeeds ^{[7]B3b[41]B2b}.

Protocol

The preceding transplant discussion establishes the definitive pathway; immediate hematology care now depends on balancing cytopenia support, thrombosis prevention, and suppression of the inflammatory or malignant driver. The retrieved evidence does not report an NCCN, ASCO, or ESMO recommendation strength for this triad.

1. Stratify severity and competing risks first. Assess platelet count, fibrinogen, coagulation tests , renal and hepatic function, ferritin, inflammatory markers, organ support requirements, bleeding, and venous thromboembolism (VTE). Bleeding occurred in 40% of adults in one cohort, whereas VTE occurred in 11%; the 90-day cumulative incidences were 39% and 13%, respectively ^[50]B3b. This profile makes anticoagulation an individualized decision rather than an automatic component of HLH care. Lower platelet counts and mechanical ventilation independently predicted bleeding, while VTE was associated with worse overall survival ^[50]B3b.

2. Support the blood counts while treating the driver. Provide platelet and red-cell support according to active bleeding, procedures, and clinical need; the retrieved studies do not report universal transfusion thresholds. In a randomized study of adults with platelet counts below 40 × 10(9)/L, subcutaneous recombinant human thrombopoietin at 300 IU/kg Qd increased platelet recovery and reduced platelet transfusion frequency when added to conventional systemic therapy, although survival did not differ ^[118]A1b. The study therefore supports thrombopoietin as an adjunct for selected treatment-related thrombocytopenia, not as a substitute for disease control.

3. Start and advance cytoreductive or clonal-directed therapy promptly. Etoposide -based therapy with dexamethasone remains the most commonly used standard approach in the cited pediatric and young-adult literature ^[73]C4. In malignancy-associated disease, corticosteroids and/or intravenous immunoglobulins were used with adapted antimicrobial therapy, while lymphoma-directed chemotherapy followed initial anti-HLH treatment in the reported T/NK-cell lymphoma cohort ^{[66]D5[128]C4}. The appropriate sequence depends on whether infection, malignancy, or another trigger is driving the syndrome; the retrieved evidence does not establish one universal regimen or dose for all phenotypes.

4. Use response-adapted escalation. Reassess clinical status and laboratory abnormalities serially. Day-7 response markers were more predictive than pretreatment or later assessments in etoposide-treated patients; inadequate improvement in soluble CD25, platelet count, absolute lymphocyte count, and blood urea nitrogen identified patients at high risk of early death ^[73]C4. Oral ruxolitinib 15 mg twice a day on a continuous 28-day cycle produced symptom and laboratory improvement in all five treated adults in a pilot study, but the small, uncontrolled study does not establish it as universal first-line therapy ^[4]C4. A serious adverse event was grade 4 febrile neutropenia ^[4]C4.

5. Manage failure and complications explicitly. Persistent inflammation, worsening organ dysfunction, progressive cytopenias, bleeding, thrombosis, or inadequate early marker response should trigger specialist review for salvage therapy, trigger control, malignancy-directed treatment, or expedited transplant assessment. Do not regard hemoadsorption as established cytoreduction: current evidence is limited, heterogeneous, and insufficient to determine treatment effect, survival benefit, timing, device selection, anticoagulation strategy, or pharmacokinetic safety ^[123]B2a. Do not apply routine anticoagulation without weighing platelet count and active bleeding, because VTE prevention and treatment remain challenging in this population ^[50]B3b.

Management flowchart: severity and bleeding/VTE assessment → blood-product support and trigger treatment → etoposide-based or phenotype-directed cytoreduction → serial response assessment, particularly early soluble CD25 and blood-count trends → salvage or clonal-directed escalation and transplant review for inadequate response.

Pearl: A falling platelet count is both a transfusion problem and a bleeding-risk signal, but it does not by itself justify anticoagulation; reassess thrombosis, hemorrhage, and inflammatory response together before each treatment escalation ^[50]B3b.

Treatment evolved from broad immunosuppression toward pathway-directed control of hyperinflammation and, when indicated, hematopoietic stem-cell transplantation . The pivotal therapeutic sequence came from pediatric studies, while adult practice later required more individualized intensity and duration because pediatric protocols can cause overtreatment and unnecessary toxicity in adults ^[1]A1c.

From cytotoxic remission induction to HLH protocols

Early treatment of familial HLH used myelotoxic therapy, particularly etoposide , because it induced remission in most patients, but toxicity and secondary relapse from treatment resistance limited the approach ^[130]C4. A nonmyelotoxic alternative combining intravenous methylprednisolone , rabbit antithymocyte globulin , and maintenance cyclosporine A produced systemic remission rapidly in a small pilot study, but the evidence remained insufficient to replace etoposide-based therapy ^[130]C4.

The HLH-94 protocol established the therapeutic backbone of etoposide and dexamethasone , with cyclosporine A added later, while hematopoietic stem-cell transplantation remained the route to cure for familial/genetic, relapsing, or severe and persistent disease ^[9]D5. HLH-2004 moved cyclosporine A to the beginning of treatment in an attempt to reduce pretransplant mortality and morbidity. Although the later study confirmed the efficacy of etoposide and dexamethasone, upfront cyclosporine A, corticosteroids added to intrathecal therapy, and shortening the interval to transplantation did not significantly improve outcome ^[9]D5. These findings explain why intensification of the basic protocol did not become the central advance in treatment.

Targeted therapy and adaptation to phenotype

Emapalumab marked the transition to mechanism-directed treatment for primary HLH. In a phase two-three study of previously treated patients, the response rate was 63%, and 70% proceeded to transplantation; severe infections occurred in 10 patients, demonstrating both therapeutic activity and the need for infection surveillance ^[112]B2b. Real-world use subsequently showed laboratory improvement and enabled many transplant-eligible patients to reach transplantation, supporting emapalumab as a practical second-line strategy rather than a replacement for definitive cellular therapy ^[5]B2b.

JAK inhibition extended pathway-directed treatment to secondary HLH. Ruxolitinib was administered at 15 mg twice a day in an adult pilot study, in which all five patients had clinical or laboratory improvement and 2-month overall survival was 100%; the study remained preliminary because of its small size ^[4]C4. In previously untreated children with secondary HLH, weight-based ruxolitinib at 2.5 mg, 5 mg or 10 mg twice daily produced an overall response rate of 83.3%, but nonresponders still responded to subsequent HLH-1994 therapy, supporting ruxolitinib as a phenotype-adapted option rather than a universal substitute for conventional treatment ^[49]C4.

Contemporary treatment logic

The historical record therefore supports a layered strategy: suppress life-threatening inflammation, tailor treatment to inherited versus acquired disease and trigger, use targeted agents when conventional therapy fails or causes unacceptable toxicity, and proceed to transplantation when the underlying disorder requires definitive immune replacement ^[1]A1c[7]B3b. The next section addresses the complications that arise from both the syndrome and this immunochemotherapy-based treatment history.

Pearl: Do not interpret failure of one treatment layer as failure of the overall strategy: targeted therapy can stabilize selected patients, but definitive transplantation remains necessary when the underlying familial or otherwise transplant-requiring disease persists ^[7]B3b.

As treatment has evolved, attention to complications remains inseparable from controlling the inflammatory syndrome. The disease itself can progress from cytopenias and hepatosplenomegaly to hepatitis, coagulopathy, liver failure, central nervous system involvement, and multiorgan failure ^[51]D5. In malignancy-associated disease, cytokine-driven inflammation can also enhance replication of the underlying hematologic malignancy while disrupting host immunity ^[12]D5.

Disease- and treatment-related complications

The complication profile changes with disease activity, immune suppression, transplantation, and cellular therapy. The evidence supplied here supports the following monitoring and mitigation priorities:

These complications explain why disease control must be balanced against immunosuppression, myelosuppression, and transplant toxicity. Adult guidance specifically supports individualized treatment duration and dose reduction when pediatric protocols would create unnecessary toxicity ^[1]A1c. The next section builds on these complications by examining the survival consequences and natural history of persistent or relapsed disease.

Pearl: A new infection, falling blood counts, neurologic change, or evolving liver, coagulation, or multiorgan dysfunction should be treated as a possible change in HLH activity or treatment toxicity, not as an isolated abnormality ^[51]D5.

Once organ dysfunction develops, the untreated trajectory can be rapid and fatal, particularly in malignancy-associated hemophagocytic lymphohistiocytosis (HLH), where excess mortality is predominantly attributable to multiorgan dysfunction rather than lymphoma progression alone ^[79]B3b. Prognosis therefore depends on the trigger, inflammatory burden, early treatment response, and the prognostic scores described in Section 6.

Risk-stratified outcomes

The optimized HLH inflammatory (OHI) index identifies a high-risk inflammatory phenotype in patients with hematologic malignancy. OHI-positive patients had 1-year overall survival of 33%, compared with 81% in OHI-negative patients; median survival was 190 days in the OHI-positive group and was not reached in the OHI-negative group ^[79]B3b. At 3 years, overall survival was 23% versus 80%, respectively ^[79]B3b. The OHI index remained strongly associated with mortality after adjustment, with a hazard ratio of 5.2 (95% CI, 3.0-8.8); NNT not calculable from the reported data ^[79]B3b.

Adult secondary HLH without malignancy also shows marked heterogeneity. In one cohort, 5-year overall survival was 25.1% for Epstein-Barr virus-associated HLH, compared with 82.4% for autoimmune disease, 78.7% for other infections, and 55.5% for unknown causes ^[80]B3b. A model incorporating the reported clinical risk factors separated patients into low-risk, intermediate-risk, and high-risk groups with 5-year survival of 92.1%, 36.8%, and 18.0%, respectively ^[80]B3b.

Recovery, relapse, and late sequelae

Response often declares itself early. In etoposide -based therapy, day-7 laboratory changes predicted outcome more effectively than pretreatment or later measurements; fewer than 25% improvement in soluble CD25 was the strongest unfavorable marker ^[73]C4. In familial HLH, remission occurred at a median of 76 days, with a range of 15-705 days ^[24]C4. Neurological sequelae remain an important long-term burden after severe disease or transplantation, and were reported in 7 patients in one transplant cohort, 6 of whom had central nervous system disease at diagnosis ^[25]C4.

Recurrence risk is trigger- and genotype-dependent, and general recurrence rates are not reported in these studies. In a pediatric secondary-HLH cohort, one patient with EBV-associated HLH relapsed after complete response ^[49]C4. By contrast, among patients with CD27 or CD70 deficiency who underwent transplantation, 95% survived without disease recurrence ^[21]B2b.

Treated survival and transplant trajectory

Disease control before definitive therapy changes the outlook. In a phase 2 trial of children with primary HLH, 74% of previously treated patients were alive at last observation and 70% proceeded to transplantation ^[112]B2b. In real-world use, the 12-month survival probability after emapalumab initiation was 73.1%, and pretransplant survival among transplant-eligible patients was 90.5% ^[5]B2b. Transplant remains the curative pathway for familial disease, although historical transplant cohorts reported 8-year overall survival of 58.6% and transplantation-related mortality of 25.7% ^[25]C4.

The next section addresses how these risk differences are modified in special populations, including pregnancy, where organ reserve, trigger profile, and treatment constraints require separate interpretation.

Pearl: A falling inflammatory burden by the first week and a clear treatment response by 8 weeks should guide prognostic reassessment, because persistent inflammation identifies patients who require an alternative or escalated strategy even when the initial diagnostic score is incomplete ^[73]C4.

The age and organ context behind an HLH phenotype changes both diagnostic interpretation and the safety margin of treatment, so the standard pathway requires host-specific judgment rather than automatic protocol transfer.

Paediatric HLH

Children require early assessment for inherited or immune-dysregulation disorders, even when a trigger appears infectious, rheumatologic, or malignant. In a cohort of children meeting HLH-2004 criteria, biallelic familial HLH defects were enriched in those diagnosed younger than 1 year, while broader whole-exome sequencing identified likely molecular explanations in 28 of 48 (58%) research-tested patients ^[114]B2b. These findings support pairing functional immune studies with broad genetic evaluation when results will influence hematopoietic stem cell transplantation or long-term recurrence planning, rather than relying only on a familial-HLH gene panel ^[114]B2b.

Paediatric treatment evidence remains protocol-specific. A prospective trial used ruxolitinib 25 mg/m2 twice daily with dexamethasone , with or without etoposide , for 8 weeks; all 5 patients with newly diagnosed HLH achieved complete response by week 8, and all were alive at 1 year ^[109]B2b. This regimen should therefore be understood as trial evidence, not a universal substitute for response-adapted specialist care.

Adults and the elderly

Adults need an age-adjusted diagnostic approach because HLH-2004 criteria were developed for children and are not validated in adults; pediatric-derived treatment can also cause overtreatment and unnecessary toxicity, prompting individualized treatment duration and dose reduction ^[1]A1c. Marked hyperferritinemia is less specific in adults: among 113 patients with ferritin higher than 50,000 µg/L, common associated conditions included renal failure, hepatocellular injury, infection, and hematologic malignancy ^[142]B3b.

Age does not exclude inherited disease. A 62-year-old man with recurrent HLH carried compound heterozygous PRF1 mutations, supporting consideration of genetic HLH in older adults with recurrent or unexplained episodes ^[33]C4.

Renal or hepatic impairment

Renal failure and hepatocellular injury can accompany profound hyperferritinemia without establishing HLH in adults, so organ injury should not be treated as diagnostic proof ^[142]B3b. The supplied evidence does not report organ-specific dosing, contraindications, or validated treatment modifications for renal or hepatic impairment. Any adjustment must therefore follow the prescribing information for the selected agent and specialist pharmacologic review, rather than an HLH-specific dose stated here.

Pregnancy

The supplied HLH evidence does not report pregnancy-specific teratogenicity, delivery-planning, or breastfeeding data. It therefore does not establish a pregnancy-safe regimen, timing for delivery, or when breastfeeding can resume. Management requires coordinated maternal-fetal medicine, hematology, neonatology, and pharmacy review, with treatment selection and delivery timing individualized to maternal inflammatory control and fetal exposure; the evidence base for these decisions is not provided in the cited abstracts.

Pearl: In an older adult or a patient with renal or hepatic injury, neither age nor ferritin alone establishes secondary HLH; reassess the trigger, pursue genetic evaluation when the phenotype is recurrent or unexplained, and avoid importing paediatric criteria or dosing without adult-specific review ^{[1]A1c[142]B3b}.

After pregnancy-specific considerations, prevention focuses on identifying hereditary risk before the first inflammatory episode and recognising trigger-associated disease early. The supplied evidence does not identify an NCCN, ASCO, or ESMO population-screening program, nor does it provide a guideline-graded surveillance interval for the general population.

Hereditary-risk screening and cascade evaluation

Prospective newborn screening for genetic HLH is biologically plausible, but genotype and functional assays do not reliably predict individual disease risk in hypomorphic PRF1 states ^[82]D5. Therefore, a positive familial result should prompt structured genetic counseling , confirmation of the variant, and testing of relatives rather than automatic pre-emptive hematopoietic stem cell transplantation . In A91V/pLOF carriers, family screening identified many asymptomatic individuals, and the study supports regular monitoring while arguing against transplantation in every asymptomatic carrier ^[82]D5.

When hereditary HLH is suspected, screening should prioritise germline testing and cytotoxicity-pathway evaluation. Combined perforin expression and CD107a upregulation testing is more sensitive than NK-cell function testing for biallelic HLH-associated mutations, with reported sensitivities of 96.6% and 93.8%, respectively, compared with 59.5% for NK-cell function testing ^[144]B3b. These tests support risk assessment, but they do not replace clinical follow-up or molecular interpretation, particularly in adults with apparently secondary HLH, where no major intrinsic cytotoxicity defect was found compared with disease controls ^[145]B3b.

Surveillance of at-risk patients and survivors

Surveillance should be symptom-led and trigger-aware rather than calendar-based when no interval is specified. For asymptomatic hereditary-risk carriers, education and monitoring should include:

• New or recurrent fever, cytopenia, or constitutional deterioration ^[82]D5
• Abdominal enlargement or suspected hepatosplenomegaly, assessed clinically and, when indicated, by ultrasound ^[82]D5
• New neurological symptoms, assessed with serial neurological examination ^[82]D5
• Infectious or malignant triggers, particularly when the clinical course is unexplained ^[82]D5

For patients with newly diagnosed malignancy-associated HLH risk, routine inflammatory surveillance has reported diagnostic and prognostic value ^[2]B2b. After CAR T-cell therapy , surveillance should be intensified for patients with B-ALL, recipients of tisagenlecleucel or ciltacabtagene autoleucel, and those who develop severe cytokine release syndrome, because HLH incidence varies by disease, product, and CRS severity ^[108]B2a. The evidence does not establish a universal post-treatment interval; local cellular-therapy protocols should define assessment timing.

Prevention limits and patient education

No supplied study demonstrates that vaccination, antimicrobial prophylaxis, or another primary-prevention intervention prevents HLH across inherited or secondary forms. Education should instead emphasise urgent assessment for fever, falling blood counts, organ enlargement, neurological change, or a new infection or malignancy, because early recognition is the actionable preventive strategy supported by the available evidence ^[82]D5.

Pearl: Do not interpret a reassuring genotype or cytotoxicity assay as permission to discharge an at-risk carrier from follow-up; phenotype can diverge within families, so education and ongoing clinical surveillance remain essential ^[82]D5.