Persistent fever, extreme ferritin, and a hidden malignant trigger
A 38-year-old previously healthy woman arrived with two weeks of pharyngitis, diarrhea, and fever, but the biologic signal was far more ominous than the symptom list suggested. Her temperature reached 39.5°C, heart rate climbed to 115 beats/min, and laboratory testing showed normocytic anemia, liver enzyme elevation, and hyperferritinemia of 48,975 μg/L. In a patient with this degree of inflammatory activation, ferritin is not just an acute-phase reactant; it is often a marker of unchecked macrophage and T-cell activation, the same immune circuit that drives hemophagocytic lymphohistiocytosis.
Initial computed tomography of the chest, abdomen, and pelvis revealed splenomegaly but no lymphadenopathy. Cultures of blood, urine, and cerebrospinal fluid were negative, as were respiratory viral polymerase chain reaction testing and studies for Epstein-Barr virus, cytomegalovirus, and hepatitis A, B, and C. Empiric broad-spectrum antibiotics were reasonable at presentation, yet the workup quickly shifted away from routine infection once the laboratory pattern became harder to ignore.
One sentence captured the bedside problem: this was not behaving like a standard febrile illness.
Hemophagocytosis confirmed the syndrome, not the cause
Further testing showed hypertriglyceridemia of 2.77 mmol/L and an elevated soluble interleukin-2 receptor level of 4741 U/mL. Bone marrow aspirate demonstrated hemophagocytosis, with histiocytes engulfing erythroid precursors and other hematopoietic elements, but no lymphoma was visible in the marrow sample. Together, the fever, splenomegaly, cytopenia, triglyceride elevation, ferritin surge, soluble interleukin-2 receptor elevation, and marrow hemophagocytosis fit HLH using the familiar clinical and pathologic framework.
In adults, HLH is most often secondary to infection, autoimmune disease, medications, or malignancy. Lymphoma is the most common malignancy-associated trigger, and the immunobiology makes sense: an occult neoplasm can sustain antigenic stimulation, cytokine release, and macrophage activation until systemic inflammation becomes self-amplifying. Bone marrow can show the consequence of the storm without revealing the spark that started it.
The patient was started on dexamethasone at 24 mg/day, calculated as 10 mg/m2/day, together with anakinra at 2 mg/kg/day. That choice addressed the immediate danger. Steroids blunt the cytokine cascade, and IL-1 blockade can dampen the hyperinflammatory loop that makes HLH rapidly fatal if untreated.
PET-CT exposed the lesion conventional imaging missed
The diagnostic turning point came with positron emission tomography CT, ordered to look for an occult malignancy after standard investigations and conventional imaging failed to identify a trigger. PET-CT showed fluorodeoxyglucose-avid bilateral palatine tonsils. In HLH, that finding carries particular weight because metabolically active lymphoid tissue may be the only accessible clue to an otherwise hidden lymphoma.
Tonsillar biopsy confirmed the diagnosis: an atypical CD3- and CD8-positive T-cell infiltrate with a high proliferative index, consistent with peripheral T-cell lymphoma, not otherwise specified. The marrow had shown the inflammatory aftermath. The tonsils revealed the underlying malignancy.
This sequence matters at the bedside because HLH treatment cannot stop at immune suppression alone. A patient may stabilize briefly with dexamethasone or other immunomodulators, yet persistent antigenic drive from an untreated lymphoma can keep the syndrome smoldering or recur once steroids taper. PET-CT can change management by identifying a metabolically active biopsy target when anatomic imaging remains unrevealing.
Treating both the cytokine storm and the lymphoma
After diagnosis, the patient received six cycles of CEOP, using cyclophosphamide, etoposide, vincristine, and prednisone. She then underwent autologous hematopoietic stem-cell transplantation with BEAM conditioning, consisting of carmustine, etoposide, cytarabine, and melphalan. At 24-month follow-up, she remained in complete remission.
The treatment arc reflects the dual obligation in malignancy-associated HLH: suppress the inflammatory syndrome quickly, then eradicate the driver. Etoposide-based therapy has long held an important place in lymphoma-associated HLH because it can act on both malignant lymphoid cells and the immune activation cycle. Here, the clinicians used immunosuppression first to stabilize the cytokine storm, then definitive lymphoma therapy once pathology identified the trigger.
The favorable remission at two years is encouraging, but the case is best read as a diagnostic lesson rather than a claim of generalizable success. Peripheral T-cell lymphoma can be elusive, especially when it presents in extranodal tissue without obvious nodal enlargement. Tonsillar tissue, despite being easy to overlook on conventional CT, can become the decisive biopsy site when PET metabolism points the way.
Why this case should change how HLH is investigated
HLH in adults is rare, life-threatening, and diagnostically treacherous. The syndrome is defined by inflammation, but the clinical task is to identify the upstream cause before organ failure advances. In a patient with persistent fever, splenomegaly, marked ferritin elevation, hypertriglyceridemia, and marrow hemophagocytosis, the absence of infection should prompt renewed suspicion for malignancy even when CT does not show lymphadenopathy.
PET-CT deserves a place in that escalation pathway when standard tests fail to identify a trigger and clinical suspicion remains high. Its value lies in metabolic detection rather than gross anatomy, which is why it can reveal occult lymphoma in sites that appear unremarkable on standard cross-sectional imaging. In this case, bilateral palatine tonsillar uptake was not incidental; it was the only radiographic clue that led to tissue diagnosis.
The key limitation is also the case’s central warning: marrow hemophagocytosis confirms HLH physiology, but it does not exclude an occult lymphoma elsewhere. When the inflammatory phenotype is strong and the usual infectious studies are negative, the search must continue until a driver is found. Delay costs time, and in HLH, time is measured against organ dysfunction, not convenience.