Hemophagocytic lymphohistiocytosis (HLH) is a rare and deadly disease in which unregulated proliferation of histiocytes and T-cell infiltration takes place. It is known as a pediatric disease in which gene defects result in impaired cytotoxic NK- and T-cell function. It has been associated with autosomal recessive inheritance pattern. Without therapy, survival for these patients with active familial HLH is approximately 2 months.
Recognition of the disease has increased over the years, and as a result the diagnosis of HLH in adults also has increased. An acquired form can be triggered by viruses like Epstein-Barr virus, influenza, HIV, lymphoid malignancies, rheumatologic disorders, or immunodeficiency disorders. Survival rates for untreated HLH have been reported at < 5%.1 Despite early recognition and adequate treatment, HLH carries an overall mortality of 50% in the initial presentation, 90% die in the first 8 weeks of treatment due to uncontrolled disease.2
Case Presentation
A 56-year-old man with no active medical issues except for a remote history of non-Hodgkin lymphoma treated with chemotherapy and splenectomy in 1990 presented to the Veterans Affairs Caribbean Healthcare System in San Juan, Puerto Rico. He was admitted to the medicine ward due to community acquired pneumonia. Three days into admission his clinical status deteriorated, and the patient was transferred to the intensive care unit (ICU) due to acute respiratory failure and sepsis secondary to worsening pneumonia. Chest imaging demonstrated rapidly progressing diffuse bilateral infiltrates. Due to the severity of the chest imaging, a diagnostic bronchoscopy was performed.
The patient’s antibiotics regimen was empirically escalated to vancomycin 1500 mg IV every 12 hours and meropenem 2 g IV every 8 hours. Despite optimization of therapy, the patient did not show clinical signs of improvement. Febrile episodes persisted, pulmonary infiltrates and hypoxemia worsened, and the patient required a neuromuscular blockade. Since the bronchoscopy was nondiagnostic and deterioration persistent, the differential diagnosis was broadened. This led to the ordering of inflammatory markers. Laboratory testing showed ferritin levels > 16,000 ng/mL, pointing to HLH as a possible diagnosis. Further workup was remarkable for triglycerides of 1234 mg/dL and a fibrinogen of 0.77 g/L. In the setting of bicytopenia and persistent fever, HLH-94 regimen was started with dexamethasone 40 mg daily and etoposide 100 mg/m2. CD25 levels of 154,701 pg/mL were demonstrated as well as a decreased immunoglobulin (Ig) G levels with absent IgM and IgA. Bone marrow biopsy was consistent with hemophagocytosis. The patient eventually was extubated and sent to the oncology ward to continue chemotherapy.
Discussion
A high clinical suspicion is warranted for rapid diagnosis and treatment as HLH evolves in most cases to multiorgan failure and death. The diagnostic criteria for HLH was developed by the Histiocyte Society in 1991 and then restructured in 2004.3,4 In the first diagnostic tool developed in 1991, diagnosis was based on 5 criteria (fever, splenomegaly, bicytopenia, hypertriglyceridemia and/or hypofibrinogenemia, and hemophagocytosis). Three additional laboratory findings were also described as part of HLH diagnosis since 2004: low or absent NK-cell-activity, hyperferritinemia of > 500 ng/dL, and high-soluble interleukin-2-receptor levels (CD25) > 2400 U/mL. Overall, 5 of 8 criteria are needed for the HLH diagnosis.
Despite the common use of these diagnostic criteria, they were developed for the pediatric population but have not been validated for adult patients.5 For adult patients, the HScore was developed in 2014. It has 9 variables: 3 are based on clinical findings (known underlying immunosuppression, high temperature, and organomegaly; 5 are based on laboratory values (ferritin, serum glutamic oxaloacetic transaminase, cytopenia, triglycerides, and fibrinogen levels); the last variable uses cytologic findings in the bone marrow. In the initial study, probability of having HLH ranged from < 1% with an HScore of ≤ 90% to > 99% with an HScore of ≥ 250 in noncritically ill adults.5 A recently published retrospective study demonstrated the diagnostic reliability of both the HLH-2004 criteria and HScore in critically ill adult patients. This study concluded that the best prediction accuracy of HLH diagnosis for a cutoff of 4 fulfilled HLH-2004 criteria had a 95.0% sensitivity and 93.6% specificity and HScore cutoff of 168 reached a 100% sensitivity and 94.1% specificity.6