https://orcid.org/0000-0002-6758-7809
In this issue of Blood, Henter et al report on the updated hemophagocytic lymphohistiocytosis (HLH)-2004 diagnostic criteria for primary or familial hemophagocytic lymphohistiocytosis (FHL), developed through case-control analysis of international pediatric trial databases.1 The remarkable bottom line result is genetic and statistical confirmation of expert clinical judgment, which was the basis of the Histiocyte Society’s trial entry criteria HLH-94 and HLH-2004. The clinical pathway to diagnose FHL is confirmation of 5 of 8 diagnostic criteria, including (1) fever, (2) bicytopenia, (3) splenomegaly, (4) hyperferritinemia, (5) hypertriglyceridemia or hypofibrinogenemia, (6) hemophagocytosis, (7) elevated soluble interleukin-2 receptor, and (8) low/absent natural killer (NK)-cell activity.2 An affected sibling or detection of a mutated FHL gene is equally valid for diagnosis of FHL.
Primary HLH prior to HLH-94 was a nearly 100% deadly hyperinflammatory syndrome caused by inherited impaired lymphocyte cytotoxicity and subsequent organ-damaging lymphocyte and macrophage proliferation. Stringent immunosuppression to stop the detrimental cytokine storm using timely application of dexamethasone, etoposide, cyclosporin A, and intrathecal methotrexate as induction and bridging therapy toward allogeneic hematopoietic stem cell transplantation achieved a >50% cure rate, which improved to 79% in 2024 (see table).3-5 This diagnostic and therapeutic success story depended on the vigilant clinical observation of sudden and extreme multiorgan hyperinflammation followed by trial-structured development of diagnostic variables with concomitant, courageous application of single-arm induction/consolidation therapy. Primary HLH is a truly rare syndrome, with an estimated incidence of 1:50 000 births, so a randomized comparison of treatment strategies has not been an option.
Evolution of diagnostic accuracy and survival rates for FHL in the HLH-94 to HLH-2024 study generations
| . | HLH-943 . | HLH-20044 HLH-2016-20215 . | HLH-20241 . | OHI8 . |
|---|---|---|---|---|
| Clinical criteria | Fever, splenomegaly | Fever, splenomegaly | Fever, splenomegaly | Fever, splenomegaly |
| Laboratory criteria | Cytopenia in ≥2 of 3 cell lines Hemoglobin <90 g/L Platelets <100 × 109/L Neutrophils <1.0 × 109/L Hypertriglyceridemia and/or hypofibrinogenemia Fasting triglycerides ≥2 mmol/L Fibrinogen ≤1.5 g/L | Cytopenia in ≥2 of 3 cell lines Hemoglobin <90 g/L Platelets <100 × 109 /L Neutrophils <1.0 × 109/L Hypertriglyceridemia and/or hypofibrinogenemia Fasting triglycerides ≥3 mmol/L Fibrinogen ≤1.5 g/L | Cytopenia in ≥2 of 3 cell lines Hemoglobin <90 g/L Platelets <100 × 109/L Neutrophils <1.0 × 109/L Hypertriglyceridemia and/or hypofibrinogenemia Fasting triglycerides ≥3 mmol/L Fibrinogen ≤1.5 g/L | Cytopenia in ≥2 of 3 cell lines Hemoglobin <90 g/L Platelets <100 × 109/L Neutrophils <1.0 × 109/L Hypertriglyceridemia and/or hypofibrinogenemia Fasting triglycerides ≥3 mmol/L Fibrinogen ≤1.5 g/L |
| Histopathologic criteria | Hemophagocytosis in spleen/bone marrow/lymph node, no evidence of malignancy | Hemophagocytosis in spleen/bone marrow/lymph node, no evidence of malignancy | Hemophagocytosis in spleen/bone marrow/lymph node, no evidence of malignancy | Hemophagocytosis in spleen/bone marrow/lymph node |
| Inflammatory criteria | Low or absent NK-cell activity, ferritin ≥500 μg/L, soluble CD25 ≥2400 U/mL | Ferritin ≥500 μg/L, soluble CD25 ≥2400 U/mL | Ferritin ≥1000 μg/L, soluble CD25 ≥3900 U/mL | |
| Overall survival | 5 years, 50% | 2004: 5 years, 59%; 2016-2021: 3 years, 79% | NA | 5 years, 10%-30%∗ |
| . | HLH-943 . | HLH-20044 HLH-2016-20215 . | HLH-20241 . | OHI8 . |
|---|---|---|---|---|
| Clinical criteria | Fever, splenomegaly | Fever, splenomegaly | Fever, splenomegaly | Fever, splenomegaly |
| Laboratory criteria | Cytopenia in ≥2 of 3 cell lines Hemoglobin <90 g/L Platelets <100 × 109/L Neutrophils <1.0 × 109/L Hypertriglyceridemia and/or hypofibrinogenemia Fasting triglycerides ≥2 mmol/L Fibrinogen ≤1.5 g/L | Cytopenia in ≥2 of 3 cell lines Hemoglobin <90 g/L Platelets <100 × 109 /L Neutrophils <1.0 × 109/L Hypertriglyceridemia and/or hypofibrinogenemia Fasting triglycerides ≥3 mmol/L Fibrinogen ≤1.5 g/L | Cytopenia in ≥2 of 3 cell lines Hemoglobin <90 g/L Platelets <100 × 109/L Neutrophils <1.0 × 109/L Hypertriglyceridemia and/or hypofibrinogenemia Fasting triglycerides ≥3 mmol/L Fibrinogen ≤1.5 g/L | Cytopenia in ≥2 of 3 cell lines Hemoglobin <90 g/L Platelets <100 × 109/L Neutrophils <1.0 × 109/L Hypertriglyceridemia and/or hypofibrinogenemia Fasting triglycerides ≥3 mmol/L Fibrinogen ≤1.5 g/L |
| Histopathologic criteria | Hemophagocytosis in spleen/bone marrow/lymph node, no evidence of malignancy | Hemophagocytosis in spleen/bone marrow/lymph node, no evidence of malignancy | Hemophagocytosis in spleen/bone marrow/lymph node, no evidence of malignancy | Hemophagocytosis in spleen/bone marrow/lymph node |
| Inflammatory criteria | Low or absent NK-cell activity, ferritin ≥500 μg/L, soluble CD25 ≥2400 U/mL | Ferritin ≥500 μg/L, soluble CD25 ≥2400 U/mL | Ferritin ≥1000 μg/L, soluble CD25 ≥3900 U/mL | |
| Overall survival | 5 years, 50% | 2004: 5 years, 59%; 2016-2021: 3 years, 79% | NA | 5 years, 10%-30%∗ |
The OHI for secondary, malignancy-associated HLH in adults relates to HLH-2024 through 5 of 7 positive HLH-2024 criteria. HLH-2024 is a structured case-control study to validate the diagnostic criteria.
FHL, familial hemophagocytic lymphohistiocytosis; HLH, hemophagocytic lymphohistiocytosis; NA, not applicable; NK, natural killer; OLH, optimized HLH index.
This survival estimate relates strictly to malignancy-associated HLH in adults. Other forms of the secondary type in adult patients with HLH (triggered by autoinflammatory disorders, autoimmune disorders, or infections that come with HLH) have better outcome data.
Henter et al set out to interrogate the accuracy of the clinically developed diagnostic HLH-2004 score for the specificity against phenotypically overlapping hyperferritinemic syndromes, such as sepsis, infections, and juvenile idiopathic arthritis, by comparing genetically or pedigree-confirmed primary HLH from the HLH-94 and HLH-2004 and the Italian HLH registry with pediatric sepsis and “rheuma” cohorts. High sensitivity was ascertained through well-characterized HLH cohorts with genetic confirmation of HLH.
Certainly, the retrospective nature of the study with missing variables, multiple imputations, and control groups without malignancies or atypical infections represent potential scientific weaknesses. In HLH, such conditions need to be excluded to confirm FHL and to safeguard disease-specific treatment for malignancies or infections. Yet, by computing several predicting models, the authors achieved improved clinical practicability by reducing the original HLH-2004 set from 8 to 7 variables, excluding NK-cell activity. Further, the study allows for the conclusion that the number of criteria fulfilled (5 of 7) is more accurate than a graded threshold score based on continuous variables, as it is used by the HScore algorithm for HLH in adults.6 Besides providing us with a validated and updated HLH-2004 score, which I named HLH-2024 (see table), the authors also provide guidance with regard to companion genetic FHL diagnostics and cellular studies, which are recommended as confirmatory diagnostic steps in helping to fine-tune treatment and transplantation planning.
When applicable to only such a small fraction of affected children, why and how does this HLH-2004 remodeling project impact hematology in a broader sense?
First, this historic approach to convert a formerly deadly syndrome into a curable condition demonstrates how clinical pioneers become experts, consolidate their gut feeling accuracy through international network collaboration, and finally---by incorporation of genetic and functional assays---are able to elevate the formerly expert opinion score to a best possible evidence level that helps determine the correct diagnosis independent of HLH expert status. This type of clinical science was frequently used (and often criticized) during the COVID-19 pandemic, where fast-track clinical pioneering was called into action to define the hyperinflammatory COVID-19 subtype for targeted immunosuppression in a subgroup of patients with viral hyperinflammation.7 Second, this validation in the pediatric HLH setting impacts the evidence level of the optimized HLH index (OHI) for adult patients with secondary, acquired HLH triggered by hematologic malignancies.8 OHI combines HLH-2004 with adapted threshold values for ferritin (1000 μg/mL) and soluble interleukin-2 receptor (3900 U/mL) for rapid identification of high-risk patients with hematologic malignancies in need of adapted fast-track diagnostic interventions and modified disease-specific therapy to overcome organ-damaging hyperinflammation. The international OHI consortium modeled the adapted criteria without NK-cell activity (because of unavailability). The HLH-2024 computations support this modification. It is hoped that the life-saving scientific collaboration of the Histiocyte Society´s consortium along HLH-94, HLH-2004, HLH-2016-2021, and HLH-2024 having achieved a drop in the mortality rate from nearly 100% to about 20% (see table) role models for secondary HLH in adults. In contrast to the progress in pediatric patients, delayed diagnosis together with missed underlying trigger conditions and difficult-to-standardize treatment protocols in adult patients with HLH still result in a 40% to 80% mortality rate. Structured OHI application for timely diagnosis, a stringent and potentially repetitive biopsy-driven search for the underlying malignancy, careful suppression and monitoring of hyperinflammation along the diagnostic pathway, and disease-specific treatment with individualized control of hyperinflammation are of utmost importance!9 While politics loses the vision of a better world in urgent need of international collaboration, the international collaboration within scientific societies keeps this vision alive!
Conflict-of-interest disclosure: The authors declare no competing financial interests.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal