When Bone Marrow Macrophages Deceive: Six Cases That Blur the Line Between Reactive and Cancerous

Deep inside the spongy cavities of bone marrow, a quiet population of scavenger cells called histiocytes performs one of the body’s most underappreciated jobs. These resident tissue macrophages patrol the hematopoietic landscape, clearing cellular debris, presenting antigens to the immune system, and keeping the delicate factory of blood production running smoothly. Usually they are inconspicuous, appearing in small numbers on biopsy slides. But under the right circumstances, they can multiply dramatically, swell with stored material, or even devour neighboring blood cells, producing microscopic scenes that look alarmingly like cancer. A new case series from pathologists at the University of California, Davis, published in Frontiers in Hematology, documents six such cases and delivers a sobering message: in the bone marrow, the line between a benign reactive process and a malignant proliferation can be vanishingly thin.

The study, led by Ananya Datta Mitra and colleagues in the Department of Pathology and Laboratory Medicine, assembles a collection of bone marrow specimens in which non-neoplastic histiocytic proliferations created genuine diagnostic dilemmas. These reactive expansions of macrophages can be driven by an extraordinarily diverse set of triggers, including mycobacterial, fungal, and viral infections, autoimmune and inflammatory disorders, lysosomal storage diseases such as Gaucher disease, hemophagocytic syndromes, drug effects, and cytokine activation associated with underlying malignancies. The morphologic consequences are equally varied: increased histiocytes with or without hemophagocytosis, granuloma formation, or accumulations of pigmented macrophages known as sea-blue histiocytes, often set against a backdrop of altered marrow cellularity. The danger, the authors emphasize, is that any of these patterns can be mistaken for histiocytic sarcoma, Langerhans cell histiocytosis, or marrow involvement by lymphoma or leukemia, with potentially serious consequences for the patient.

The first case in the series is perhaps the most technically striking. An elderly man in his seventies with metastatic non-small cell lung cancer, previously treated with multiple lines of chemotherapy, immune checkpoint inhibitors, and surgery, developed pancytopenia, a dangerous drop in all three blood cell lineages. His marrow showed features consistent with a myelodysplastic neoplasm with low blasts arising after cytotoxic therapy. But the core biopsy also revealed a focal aggregate of epithelioid cells with enlarged, pleomorphic nuclei and abundant cytoplasm. Given his history of metastatic carcinoma, the immediate concern was tumor spread to the marrow. Immunohistochemistry complicated matters further: the cells stained strongly for pan-cytokeratin with a characteristic dot-like, paranuclear pattern, a staining signature usually associated with epithelial cells. Yet the same cells expressed CD163, a robust histiocytic marker, and were negative for TTF-1, CAM5.2, Napsin A, and other carcinoma markers, as well as for myeloid and stem cell markers.

The conclusion was remarkable: aberrant cytokeratin expression in bone marrow macrophages. Cytokeratin positivity with a dot-like pattern has been described before in hematopoietic neoplasms, including myelodysplastic syndrome, acute myeloid leukemia, and myeloid sarcoma, where it is a recognized pitfall. Similar aberrant expression has been documented in tissue macrophages in prior studies, but never before in bone marrow macrophages. The practical lesson is stark. In a patient with a known carcinoma undergoing marrow evaluation for cytopenias, reliance on a single cytokeratin stain could easily lead to a false diagnosis of metastatic disease, triggering inappropriate oncologic treatment. The authors argue that a comprehensive immunohistochemical panel, including histiocytic markers such as CD163 and CD68 alongside lineage-specific epithelial and hematopoietic markers, is essential before any such conclusion is drawn.

The second case shifts the focus to metabolic disease. A teenage boy with a congenital neurodevelopmental disorder presented with acute hypoxemic respiratory failure from a viral infection complicated by bacterial pneumonia, and was found to have profound pancytopenia. CT imaging revealed massive splenomegaly measuring over thirty centimeters with hypoenhancing lesions, raising fears of lymphoma. The bone marrow, however, was nearly one hundred percent cellular and packed with large histiocytes whose cytoplasm had a distinctive crinkled tissue-paper appearance, the morphologic hallmark of Gaucher cells. Rapid trio whole genome sequencing identified biallelic pathogenic variants in the GBA1 gene, one inherited from each parent, and biochemical testing showed deficient beta-glucocerebrosidase activity alongside markedly elevated glucosylsphingosine. The diagnosis was type 1 Gaucher disease, and the ominous splenic lesions were in fact Gaucheromas, collections of storage-laden macrophages characteristic of the disease rather than malignancy.

This case carries an important subtlety: the pitfall of pseudo-Gaucher cells. In hematologic malignancies such as chronic myeloid leukemia, myelodysplastic syndromes, Hodgkin lymphoma, and multiple myeloma, macrophages overwhelmed by debris from rapid cell turnover can acquire a nearly identical tissue-paper morphology at light microscopy. Distinguishing true from pseudo-Gaucher cells cannot rest on enzyme assays alone, since demonstrating deficient beta-glucocerebrosidase activity confirms Gaucher disease, but iron staining offers a useful clue, often positive in true Gaucher cells and typically absent or faint in their mimics. The clinical background, a known storage disorder versus a state of high cell turnover, remains decisive. The authors also caution against diagnostic anchoring: the patient’s neurological findings, including congenital hypotonia and developmental delay, could not be explained by type 1 Gaucher disease and reflect a separate, still-unresolved genetic condition, a reminder that not every finding in a complex patient belongs to a single diagnosis.

The third case introduces crystal-storing histiocytosis, a rare phenomenon in which histiocytes accumulate crystallized immunoglobulin within their lysosomes. A man in his sixties with progressive fatigue and a femoral neck stress fracture was found to have monoclonal gammopathy of undetermined significance, with a small IgG lambda protein in his serum. His hypocellular marrow contained numerous histiocytes filled with irregular, non-polarizing crystalline inclusions, and remarkably, similar crystals were visible within the plasma cells themselves. Immunohistochemistry confirmed that the crystals stained for IgG and lambda, matching the monoclonal protein, while IgD and kappa stains were negative and Congo red excluded amyloid. The pathogenesis is thought to involve overproduction or structural alteration of immunoglobulins that impairs lysosomal degradation, leading to intracellular crystallization. Prior mass spectrometric analysis of stored proteins in one reported case revealed amino acid substitutions hypothesized to disrupt hydrophobic packing of the light chain, suggesting that conformational changes derail normal protein degradation. Because the crystal-laden histiocytes can dominate the marrow and conceal the underlying plasma cell or lymphoplasmacytic disorder, recognition of this entity is critical for directing the appropriate workup.

The fourth case tackles one of hematology’s most contested morphologic signs: hemophagocytosis, the engulfment of blood cells by macrophages. A man in his fifties with severe aplastic anemia and tiny paroxysmal nocturnal hemoglobinuria clones, awaiting an allogeneic stem cell transplant after failing immunosuppressive therapy, was hospitalized for gangrenous cholecystitis, during which a markedly elevated CMV viral load was detected and later resolved spontaneously. His pre-transplant marrow biopsy unexpectedly showed prominent hemophagocytic histiocytes ingesting granulocytes and nucleated erythroid precursors. Expert review excluded Langerhans cell histiocytosis, Rosai-Dorfman disease, and Erdheim-Chester disease through a negative immunophenotypic profile, including absent CD1a, S100, langerin, Factor XIIIa, and BRAF V600E expression. Laboratory results, including a ferritin of 7,757 ng/mL and elevated soluble CD25, raised concern for hemophagocytic lymphohistiocytosis under the HLH-2004 and 2024 revised criteria, yet the patient had no fever or organomegaly, and a repeat marrow two months later showed no hemophagocytosis at all.

The ambiguity is the point. As the authors note, hemophagocytosis is a hallmark of HLH but a nonspecific finding that appears in infectious, inflammatory, autoimmune, and neoplastic conditions. Isolated marrow hemophagocytosis has limited sensitivity and specificity for HLH in adult studies, and the North American Consortium for Histiocytosis guidelines explicitly state that it is not specific to, or essential for, the diagnosis, though certain subtypes, particularly phagocytosis of multiple nucleated cells including granulocytes, are more strongly associated with true HLH than simple erythrophagocytosis. The overlap between aplastic anemia and HLH compounds the difficulty, since both can produce profound cytopenias and hypocellular marrows through immune-mediated mechanisms, and reactive hemophagocytosis can appear in marrow failure syndromes without fulfilling formal diagnostic criteria.

The final two cases round out the spectrum. A man in his forties with advanced HIV and a CD4 count below twenty cells per cubic millimeter underwent marrow biopsy for unexplained anemia and imaging findings suggestive of disseminated disease. The marrow was nearly acellular, containing poorly formed granulomas without necrosis, and special stains revealed histiocytes stuffed with rod-shaped organisms that stained with both PAS and acid-fast stains, confirming disseminated Mycobacterium avium complex infection. In severely immunocompromised patients lacking robust CD4 T-cell responses, well-formed granulomas are uncommon; instead, macrophages distended with organisms and minimal inflammation are typical, and the mild reticulin fibrosis seen here is a recognized reactive finding rather than primary myelofibrosis. In the sixth case, a sixteen-year-old boy with a white blood cell count exceeding 425 billion per liter and the t(9;22) translocation diagnostic of chronic myeloid leukemia also harbored scattered sea-blue histiocytes and a pseudo-Gaucher cell, secondary manifestations of massive myeloid turnover rather than a storage disorder. Together, the series argues that morphology alone is rarely sufficient: only the systematic integration of clinical history, laboratory data, immunohistochemistry, microbiology, and molecular testing, often within a multidisciplinary team, allows pathologists to separate these reactive mimics from true histiocytic neoplasia and spare patients the consequences of overdiagnosis.

Subject of Research: Reactive non-neoplastic histiocytic proliferations in bone marrow and their distinction from histiocytic neoplasms

Article Title: Non-neoplastic histiocytic proliferations in bone marrow: a case series of uncommon morphologic findings

Article References: Ha, M., Parsazad, S., Dang, L. T., Vali Betts, E., Dwyre, D. M., & Datta Mitra, A. (2026). Non-neoplastic histiocytic proliferations in bone marrow: a case series of uncommon morphologic findings. Frontiers in Hematology, 5, Article 1939019. https://doi.org/10.3389/frhem.2026.1939019

Image Credits: AI Generated

DOI: 10.3389/frhem.2026.1939019

Keywords: bone marrow, histiocytes, macrophages, Gaucher disease, hemophagocytic lymphohistiocytosis, crystal-storing histiocytosis, cytokeratin, Mycobacterium avium complex, chronic myeloid leukemia, sea-blue histiocytes, pathology, differential diagnosis