Case Record | Ten Years of Splenomegaly, a Hidden Lesion: DLBCL With Secondary HLH, From Diagnosis to Autologous Transplant

General Condition and Chief Complaint

A middle-aged man presented in April 2025 with recurrent fever at another hospital, with a peak temperature of 40.1 °C. Blood counts showed a white cell count of 7.5 × 10⁹/L, hemoglobin 157 g/L, and platelet count 90 × 10⁹/L; high-sensitivity C-reactive protein was 87.1 mg/L. Biochemistry showed gamma-glutamyl transferase 88 U/L, total bilirubin 43.5 μmol/L, and indirect bilirubin 34 μmol/L. Splenomegaly had been noted on a previous health check about 10 years earlier and had been followed without specific treatment.

Physical Examination

No superficial lymph nodes were palpable. The spleen was enlarged. No other abnormalities were found on physical examination.

Investigations

Abdominal ultrasound at the outside hospital showed an enlarged spleen measuring approximately 152 × 58 mm. PET-CT suggested an enlarged spleen with diffusely increased FDG uptake in the bone marrow throughout the body, considered most likely a reactive proliferative change, with specialist follow-up and pathological correlation if necessary; no enlarged superficial lymph nodes or other biopsy-accessible lesions were identified. Bone marrow morphology at the outside hospital showed hypercellularity with active hematopoiesis, a myeloid-to-erythroid ratio of 2.377:1, and 7% unclassified cells — large cells with irregular shape, abundant sky-blue cytoplasm containing vacuoles, and irregular nuclei with indentations and relatively fine chromatin. The myeloid, erythroid, megakaryocytic, and lymphoid lineages showed no obvious abnormality. Hemophagocytosis was observed. Bone marrow flow cytometry showed 19.4% mature lymphocytes with approximately 0.9% CD5-CD10- clonal mature large B cells, with an immunophenotype of CD19+CD38-CD20+CKAPPA+CLAMBDA-CD5-CD10-FS elevated, requiring pathological correlation and BCL-2 and BCL-6 studies to confirm the diagnosis. Bone marrow biopsy at the outside hospital suggested neoplastic proliferation of B lymphocytes, with possible B-cell lymphoma involving the marrow.

Repeat assessment at our hospital showed a white cell count of 3.63 × 10⁹/L, hemoglobin 85 g/L, and platelet count 94 × 10⁹/L; triglycerides 3.55 mmol/L and fibrinogen 5.53 g/L; ferritin 3284 ng/mL; soluble CD25 6378 pg/mL; and NK cell activity 13.67%. Bone marrow aspirate was hypercellular with a reduced myeloid-to-erythroid ratio, active proliferation of both myeloid and erythroid lineages with mild dysplasia, and markedly active megakaryopoiesis with platelets visible singly or in clusters; occasional hemophagocytes were seen. Marrow morphology showed no obvious abnormal hematopoietic cells. Karyotype was 46,XY. Iliac crest bone marrow biopsy showed mild to moderate active proliferation of all three lineages, with myeloid and erythroid cells predominantly at the mid-to-late and later stages, normal megakaryocyte morphology and maturation, scattered small lymphocytes and a few plasma cells, no atypical elements, and no abnormal immunophenotype on immunohistochemistry — no evidence of tumor in the tissue at that time.

Clinical Course

The H-Score was 205, giving a probability of HLH greater than 90%. The diagnosis was hemophagocytic lymphohistiocytosis.

Etiological screening for HLH showed variants in NLRP1, MEFV, CARD14, and TTC37, all reported as variants of uncertain significance, which was insufficient to diagnose hereditary HLH. EBV-DNA was negative, and rheumatological and immunological markers including ANA and anti-dsDNA were normal, excluding EBV-associated and rheumatological HLH. Bone marrow biopsy at the outside hospital had suggested neoplastic B-lymphocyte proliferation with possible B-cell lymphoma involving the marrow, whereas biopsy at our hospital showed no atypical elements, no abnormal immunophenotype, and no evidence of tumor.

From June 4, 2025, the patient received the HLH-94 regimen: etoposide 200 mg on days 1 and 4, dexamethasone 20 mg/day on days 1 to 14 followed by tapering, and oral ruxolitinib 10 mg twice daily. Temperature returned to normal and general condition improved.

Repeat PET-CT showed no superficial biopsy site and an enlarged spleen with infarction. A multidisciplinary team meeting was held to establish the diagnosis, and total splenectomy was performed on July 14, 2025. A lesion measuring 2.8 × 2.2 × 1.5 cm with extensive coagulative necrosis was found on the splenic cut surface. Pathology showed necrotic cell ghosts expressing CD20, CD79a, BCL-2 (weakly positive, more than 50%), and Ki-67 (approximately 80%), consistent with diffuse large B-cell lymphoma. The final clinical diagnosis was DLBCL, Ann Arbor stage IVB, involving the spleen and bone marrow, complicated by hemophagocytic lymphohistiocytosis.

From August 2025 the patient received four cycles of POLA-R-CHP (polatuzumab vedotin, rituximab, cyclophosphamide, liposomal doxorubicin, and prednisone). During chemotherapy there was transient hepatic impairment (ALT 181 U/L), grade II peripheral neuropathy (transient weakness of both lower limbs), and infectious fever, with blood next-generation sequencing identifying Pseudomonas aeruginosa, Candida auris, and Aspergillus niger. All improved with hepatoprotective treatment, anti-infective therapy, and stepwise reduction of the polatuzumab dose from 1.8 to 1.4 to 1.0 mg/kg.

PET-CT on October 24, 2025 showed no enlarged or hypermetabolic lymph nodes within the imaging field, with a Deauville score of 1, assessed as complete remission.

The patient was admitted in November 2025 for stem cell mobilization with rituximab and teniposide. On December 6 the peripheral blood CD34+ count was 6305.31/μL, with a total collection volume of 250 mL and a total CD34+ cell dose of 19.7 × 10⁶/kg. Conditioning used a modified BEAM regimen (carmustine, teniposide, cytarabine, and melphalan). Autologous hematopoietic stem cells were reinfused on January 16, 2026, with platelet engraftment on day +9 and neutrophil engraftment on day +11. The patient was discharged in February 2026 and received antiviral prophylaxis with entecavir and acyclovir plus regular immunoglobulin replacement.

At follow-up in May 2026, PET-CT again showed no enlarged or hypermetabolic lymph nodes within the imaging field, with a Deauville score of 1, assessed as complete remission.

Initial Diagnostic Reasoning and Differential Diagnosis

The patient presented with fever of unknown origin, splenomegaly, and cytopenia, and the differential diagnosis had to consider several directions.

First, infectious disease is a common cause of fever with splenomegaly, including Epstein-Barr virus, cytomegalovirus, and Leishmania. However, EBV-DNA was negative, and the markedly elevated inflammatory markers together with multilineage cytopenia could not be explained by infection alone.

Second, autoimmune disease such as adult-onset Still disease or systemic lupus erythematosus can also produce high fever, splenomegaly, and cytopenia, but rheumatological and immunological markers were all normal, which does not support this direction.

Third, hematologic malignancy, and lymphoma in particular, had to be a priority.

The ten-year history of splenomegaly was deceptive and easy to dismiss as a benign lesion, yet long-standing splenomegaly may be precisely the manifestation of an indolent or occult lymphoma.

On that basis, the clinical picture pointed strongly toward hemophagocytic lymphohistiocytosis — a fatal hyperinflammatory syndrome in which excessive activation of lymphocytes and macrophages produces a cytokine storm and organ damage, and in which untreated patients survive only weeks. Applying the HLH-2004 criteria, the patient met multiple criteria including high fever, splenomegaly, two-lineage cytopenia, hypertriglyceridemia (3.55 mmol/L), markedly elevated ferritin (3284 ng/mL), hemophagocytosis in the marrow, and elevated soluble CD25 (6378 pg/mL). The H-Score reached 205, where a threshold of 169 has 93% sensitivity and 86% specificity for HLH, confirming the diagnosis.

Once HLH was confirmed, tracing the cause became essential. HLH is classified as primary (hereditary) or secondary (infection, malignancy, autoimmune disease). In this patient genetic testing showed no clearly pathogenic mutation, and screening for EBV and rheumatological disease was negative, whereas bone marrow flow cytometry identified 0.9% monoclonal large B cells and outside biopsy suggested possible B-cell lymphoma involving the marrow, so lymphoma-associated HLH was highly suspected. Lymphoma-associated HLH is the subtype with the worst prognosis, with a median survival of only about two months if not treated promptly and effectively, and it requires early initiation of therapy directed at the underlying tumor alongside control of inflammation.

Turning Points in the Diagnostic and Treatment Pathway

Bone marrow aspiration: the first light through the fog. With the cause of HLH still unknown, bone marrow aspiration was the key investigation. Marrow morphology found 7% unclassified cells with hemophagocytosis, and flow cytometry further detected 0.9% CD5-CD10- monoclonal large B cells. Although the proportion was very low, the presence of monoclonal B cells strongly suggested a B-cell lymphoproliferative disorder and pointed the way to the eventual diagnosis. The finding is also a reminder that in patients with HLH, the possibility of lymphoma must not be discounted even when the proportion of tumor cells in the marrow is very low.

Splenectomy: the key decision when no superficial biopsy site exists. A major difficulty in this case was that PET-CT showed no enlarged superficial lymph nodes or other biopsy-accessible lesion, and bone marrow biopsy at our hospital showed no evidence of tumor, leaving pathological diagnosis at an impasse. In that situation the spleen was the only organ from which pathology could be obtained, and with splenomegaly and infarction the patient met the indication for surgery. Splenectomy had dual value here: on one hand, postoperative pathology confirmed DLBCL, with an immunophenotype of CD20, CD79a, BCL-2 (weakly positive, more than 50%), and Ki-67 of approximately 80% providing firm evidence of aggressive DLBCL; on the other, removing the spleen — a large tumor-burden organ — also helped control the inflammatory state of HLH. The decision reflects a clinical approach in which diagnosis and treatment proceed together.

From HLH to DLBCL: establishing causality. Once splenectomy established the diagnosis of DLBCL, the causal relationship became clear: the patient's HLH was lymphoma-associated, a secondary HLH driven by DLBCL. This was a decisive step, because the core of treatment shifted from simply controlling HLH to a two-pronged strategy of controlling HLH and curing the lymphoma, laying the foundation for subsequent chemotherapy and autologous transplantation.

Treatment Strategy

Initial control of HLH. Once HLH was diagnosed, the HLH-94 regimen was started immediately, comprising etoposide and dexamethasone together with ruxolitinib. Etoposide controls activated T cells by inhibiting topoisomerase II, dexamethasone suppresses the cytokine storm, and ruxolitinib blocks the JAK-STAT pathway synergistically. This combination effectively controlled the patient's hyperinflammatory state and bought time for definitive diagnosis and targeted treatment.

Lymphoma chemotherapy and toxicity management. After DLBCL was confirmed, chemotherapy was given with the POLA-R-CHP regimen. The POLARIX study confirmed that, in previously untreated DLBCL, POLA-R-CHP significantly improves progression-free survival compared with R-CHOP. This patient achieved complete remission after four cycles, with a Deauville score of 1, confirming the efficacy of the regimen.

Toxicity management during chemotherapy was central to the success of treatment. The patient developed grade II peripheral neuropathy with transient weakness of both lower limbs. As an antibody-drug conjugate, polatuzumab vedotin carries an MMAE payload with microtubule-inhibiting toxicity, and peripheral neuropathy is a common adverse event. A stepwise dose-reduction strategy was adopted, lowering the polatuzumab dose from 1.8 mg/kg in cycles 1 and 2 to 1.4 mg/kg in cycle 3 and 1.0 mg/kg in cycle 4. Notably, dose reduction did not prevent the eventual achievement of complete remission, suggesting that timely reduction when neuropathy appears is a safe and effective management strategy.

Autologous hematopoietic stem cell transplantation. After complete remission, the patient proceeded to consolidation with autologous transplantation. Mobilization with rituximab plus teniposide was highly effective, yielding a total CD34+ cell dose of 19.7 × 10⁶/kg, far above the collection target. This efficient mobilization owed much to two factors: deep response to chemotherapy, already in complete remission, and a rational, effective mobilization regimen.

Conditioning used a modified BEAM regimen (carmustine, teniposide, cytarabine, and melphalan). BEAM is a classic conditioning regimen for autologous transplantation in non-Hodgkin lymphoma. Our center substituted teniposide, which is more lipophilic, for etoposide in order to preserve antitumor efficacy while exploiting its better penetration of the blood-brain barrier and thereby reducing the risk of central nervous system relapse.

After reinfusion on January 16, 2026, platelet engraftment occurred on day +9 and neutrophil engraftment on day +11. Engraftment was rapid, closely related to the high quality of the stem cell product. The patient was discharged smoothly in February 2026.

Lessons and Literature Review

First, in patients with fever of unknown origin accompanied by splenomegaly and cytopenia, HLH should be suspected, and once HLH is confirmed the cause must be actively pursued, in particular to rule out occult lymphoma.

Second, lymphoma-associated HLH has an extremely poor prognosis, and the core of treatment is to start therapy directed at the underlying tumor as early as possible while controlling the inflammatory state. In this case HLH-94 was used first to control inflammation, and once DLBCL was confirmed treatment switched immediately to targeted chemotherapy — a strategy of stabilizing first and then advancing.

Third, in patients with suspected lymphoma and no superficial biopsy site, splenectomy can in selected circumstances serve both diagnosis and treatment. Here splenectomy not only provided pathological diagnosis but also reduced tumor burden, creating the conditions for subsequent treatment. Surgical indications and perioperative management must, however, be strictly controlled.

Fourth, the management of ADC-related peripheral neuropathy deserves attention. Polatuzumab vedotin has a relatively high incidence of peripheral neuropathy, and timely recognition with stepwise dose reduction is key to maintaining treatment continuity. The experience here shows that reducing the polatuzumab dose progressively to 1.0 mg/kg at grade II neuropathy did not prevent the achievement of complete remission.

Fifth, for patients with lymphoma-associated HLH, autologous hematopoietic stem cell transplantation as consolidation after chemotherapy-induced complete remission is a feasible treatment option. This patient engrafted smoothly and had a good short-term outcome.

Zhu Jun

Chief physician and Medical Director of Shanghai GOBROAD Zhaxin Hospital and Shanghai Liquan Hospital.

This is a typical case of lymphoma-associated HLH. The diagnostic and treatment course was tortuous but logically clear, and it carries considerable clinical reference value.

The strengths of the case are as follows. First, diagnostic thinking was rigorous: the team did not let the deceptive ten-year history of splenomegaly obscure the abnormal clues, and detected a low proportion of monoclonal B cells by bone marrow flow cytometry, thereby locking onto the direction of lymphoma. Second, faced with the difficulty of having no superficial biopsy site, the team resolutely chose splenectomy to achieve both diagnosis and tumor debulking, demonstrating individualized clinical decision-making. Third, management across the whole course was standardized, with HLH control, lymphoma chemotherapy, toxicity management — in particular the stepwise reduction for ADC-related peripheral neuropathy — and autologous transplantation linked in orderly sequence, so that the patient ultimately achieved complete remission and completed transplantation.

The lesson of this case is that lymphoma-associated HLH has an insidious onset and rapid progression. Clinicians should remain alert and actively screen for the cause of unexplained HLH. When pathology is difficult to obtain, thinking should be broadened and invasive approaches such as surgery used when necessary to establish the diagnosis. In treatment, the principle of controlling HLH while curing the lymphoma should be maintained, and autologous hematopoietic stem cell transplantation can be considered for consolidation in eligible patients.

Wang Chun

Chief physician, discipline leader, and Medical Director of Shanghai GOBROAD Zhaxin Hospital and Shanghai Liquan Hospital.

This is a highly instructive case of lymphoma-associated HLH, and its course fully reflects the modern hematology principle of precise diagnosis, timely decision-making, and whole-course management. Faced with the difficulty of no enlarged superficial lymph nodes, the team resolutely chose splenectomy, which not only established the diagnosis of DLBCL but also laid the foundation for subsequent treatment. That clinical courage to break with convention when the diagnosis is unclear deserves recognition. After HLH was controlled, treatment switched promptly to POLA-R-CHP chemotherapy and achieved complete remission, followed by a smooth bridge to autologous transplantation. The strategy was coherent throughout and shows a deep understanding of the biological behavior of the disease.

With regard to transplant timing, this patient received modified BEAM conditioning after deep remission and engrafted rapidly, which confirms our view: for lymphoma-associated HLH, once the inflammatory storm has been contained, the window should be seized to proceed actively to consolidative transplantation rather than adopting a watch-and-wait approach. High-quality stem cell mobilization and the choice of conditioning regimen are the keys to transplant success, and this case offers a model that can be referenced for this very high-risk group of patients.

The path of medicine is full of challenges, but every difficult case that is overcome is the best tribute to life.