Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
The term histiocyte has come to encompass two cell lines, the monocyte-macrophages and the specialized antigen-presenting dendritic cells (DCs). Although a long-held belief that myeloid-derived monocytes were the immediate precursors of all macrophages and dendritic cells, the understanding of the mononuclear phagocytic system (MPS), which includes DC, monocytes, and macrophages, is undergoing continual refinement, including the nomenclature, cellular origin, and overlapping functional capabilities. , proposed a unified nomenclature based on ontogeny in which the MPS could be better understood as a dynamic monocyte-macrophage system, with both differences and redundancy in origin, function, and morphology, that includes three broad cell families—namely, embryonic-derived macrophages, common DC precursor (CDP)-derived DCs, and monocyte-derived cells (including both monocyte-derived macrophages and monocyte-derived DCs), based first on their ontogeny and secondarily on their functional and phenotypic properties ( Fig. 19.1 ).
Under the influence of cytokine macrophage colony-stimulating factor (M-CSF or CSF1) embryonic precursors of the yolk sac and fetal liver migrate to the fetal peripheral tissues where they maintain a self-renewing population of tissue-resident macrophages throughout adult life (e.g., splenic red pulp macrophages, bone osteoclasts, brain microglia, liver Kupffer cells, renal mesangial cells, pulmonary alveolar macrophages, placental Hofbauer cells, serosal macrophages, the interstitial histiocytes of the skin and connective tissues). Within this nomenclature, the epidermal Langerhans cells, derived from embryonic precursors and possessing self-renewing properties, would also be classified as resident tissue “macrophages” (see Fig. 19.1 ). The lymph nodes have their own resident macrophages that reside within the sinuses, the follicular germinal centers, and the nodal paracortex, and there is functional heterogeneity of the various macrophages. The spleen has follicular, sinus, and cordal macrophages.
DCs, on the other hand, are believed to be derived from hematopoietic stem cells of the bone marrow, distinct from classic monocytes (see Fig. 19.1 ). The common dendritic cell precursor (CDP) gives rise to either the classic/conventional DCs (cDC) or plasmacytoid DCs (pDC). The marrow, under the influence of the cytokine FMS-like tyrosine kinase 3 ligand (FLT3), can give rise to the CDP that can either circulate as a preclassic DC or function as an intermediate stage within the bone marrow as a pre-plasmacytoid DC. The cDC may be further divided into cDC1 or cDC2 based on developmental pathways with differentiation achieved in tissues, as opposed to the pDC (plasmacytoid monocyte/interferon-producing cell) in which terminal differentiation occurs in the bone marrow and then populates the lymph node in steady-state or other tissues in inflammatory states. Studies have also suggested that some of the pDC may also be derived from lymphoid bone marrow precursors. Tissue DCs are found in the tissues during steady state (e.g., lymph node interdigitating and follicular, spleen, thymus). Only the brain appears to lack an intrinsic DC pool. Of note, despite their dendritic morphology, antigen-presenting capabilities, and phenotypic expression as DCs, epidermal Langerhans cells (LCs) may be better classified under the embryonic-derived “tissue macrophage” grouping, given their shared embryonic ontogeny and self-renewal properties (vide supra). The germinal center follicular DCs are the only members of the DC family not to have a myeloid origin and are most likely derived from fibroblast-like mesenchymal cells in various tissues. Their nonmyeloid origin now serves as the basis for the exclusion of follicular DC tumors from current histiocytic disorders/neoplasia classifications ( Box 19.1 ).
a Emile JF, Abla O, Fraitag S, et al:. Revised classification of histiocytoses and neoplasms of the macrophage-dendritic cell lineages, Blood 127(22):2672–2681. Note: Follicular dendritic cell tumor/sarcoma is no longer considered a part of histiocytic/dendritic cell classification.
Langerhans cell histiocytosis (LCH)
Indeterminate dendritic cell tumor (ICH)
Erdheim-Chester disease (ECD)
Mixed LCH/ECD
Cutaneous non-Langerhans cell histiocytosis
Xanthogranuloma family
Juvenile xanthogranuloma (JXG)
Adult xanthogranuloma (AXG)
Solitary reticulohistiocytoma (SRH)
Benign cephalic histiocytosis (BCH)
Generalized eruptive histiocytosis (GEH)
Progressive nodular histiocytosis (PNH)
Non-xanthogranuloma family
Cutaneous Rosai-Dorfman disease
Necrobiotic xanthogranuloma (NXG)
Cutaneous histiocytosis not otherwise specified
Cutaneous non-Langerhans cell histiocytosis with a major systemic component
Xanthogranuloma family Xanthoma disseminatum
Non-xanthogranuloma family: Multicentric reticulohistiocytosis (MRH)
Familial Rosai-Dorfman disease
Sporadic Rosai-Dorfman disease
Classic (nodal) Rosai-Dorfman Disease
Extranodal Rosai-Dorfman disease
Neoplasia-associated Rosai-Dorfman disease
Immune disease–associated Rosai-Dorfman disease
Primary malignant histiocytoses: localization and subtype (histiocytic, Langerhans cell, interdigitating, indeterminate cell, or not specified)
Secondary malignant histiocytoses (following or associated with another hematologic neoplasia)
Primary HLH: Monogenic, mendelian inherited conditions leading to HLH
Secondary HLH (apparently non-mendelian HLH)
HLH of unknown/uncertain origin
b Swerdlow SH, Campo E, Pileri SA, et al: The 2016 revision of the World Health Organization (WHO) classification of lymphoid neoplasms, Blood 127:2375–2390, 2016.
Histiocytic sarcoma
Langerhans cell histiocytosis
Langerhans cell sarcoma
Indeterminate dendritic cell tumor
Interdigitating dendritic cell sarcoma
Follicular dendritic cell sarcoma
Fibroblastic reticular cell tumor
Disseminated juvenile xanthogranuloma
Erdheim-Chester disease c
c Changes from the 2008 classification.
The monocyte-derived cells are the third group of the dynamic MPS that serve as a pool of trafficking and inflammatory-mediated cells, derived from so-called common monocyte progenitor cells (cMoPs) of the bone marrow, which can give rise to a variety of functional cell types dependent on the local microenvironment. This includes both the monocyte-derived macrophages and monocyte-derived DCs (see Fig. 19.1 ). In human tissues, there is still debate as to what cells are actually “monocyte” derived. However, a subset of tissue macrophages of the gastrointestinal tract, dermis, heart, and pancreas appear to be derived from circulating adult cMoPs rather than embryonic precursors. In mouse models these bone marrow–derived tissue macrophages do not exhibit the same self-renewal properties as embryonic-derived tissue macrophages. Human blood monocytes are typically classified as CD14 + /CD16− classic type, CD14 + /CD16 + intermediate type, and CD14 − /CD16 nonclassic type. Studies have demonstrated that the plasticity of the CD14 + /CD16 + monocytes is dependent on the local tissue environment, which, when stimulated with macrophage cytokine M-CSF (CSF1), induces high expression of CD163. Stimulation with dendritic cell cytokines, granulocyte-macrophage colony-stimulating factor (GM-CSF or CSF2), and IL4 results in much lower CD163 expression.
In general, the “macrophage,” both of the resident tissue cell (embryonic-derived) and monocyte-derived bone marrow (BM) cell, is an active cell that is at the front line of the innate immune system. Both tissue-resident macrophages and monocyte-derived macrophages play a role during tissue injury and repair and are often hard to differentiate in active inflammatory states since they share tissue markers. For example, only recently has a microglia-specific marker, anti-Transmembrane 199 (Tmem199), made claims to selectively differentiate microglia from BM-derived monocytes in inflammatory CNS lesions. In basic terms the “macrophage” has an extensive repertoire of pattern recognition motifs, and can phagocytose, digest, and destroy pathogens; become immune activated; and interact with DCs, B cells, and others to induce specific and acquired immunity, both cellular and humoral. The “macrophage” can modulate responses to intrinsic and extrinsic stimuli. Individuals, by virtue of their genetic heterogeneity, can vary in their ability to handle similar insults. Polymorphisms in interleukin (IL)-12R, for example, will determine a variety of granulomatous responses to mycobacterial infections. Macrophage production from monocyte-derived marrow precursors is driven by IL-1, IL-3, M-CSF, and GM-CSF. Similarly, DC replenishment can occur during periods of need from adult stem cell precursors and/or monocyte-derived DC, regulated, in part, by IL-4, GM-CSF, and tumor necrosis factor alpha (TNF-α). DCs exist in the periphery as sentinel cells, sampling the internal or external environment. When they encounter a danger signal through their cytokine or toll-like receptors, the cells become activated and proceed to mature. DCs can change their phenotype from one adapted to antigen acquisition (pinocytic, phagocytic) to one more suited to antigen presentation, which occurs in the central lymphoid tissues, lymph nodes, or spleen. DCs present antigen in a major histocompatibility complex (MHC)-restricted manner to a variety of cell types, T lymphocytes, B cells, and natural killer (NK) or NKT cells to induce immune or tolerizing effects. There is plasticity in the system, and both macrophages and DCs can be driven in the direction of the other by various stimuli; cells with dendritic phenotype can be converted to macrophage phenotype by IL-10 or M-CSF. It is therefore not surprising that there are gray zones in which the distinction of macrophages from DCs is not absolute and that histiocytic lesions commonly contain a mix of monocytes, macrophages, and DCs, even if one cell type dominates. Macrophage fusion is a common phenomenon resulting in multinucleated giant cells. Under certain circumstances, such as incubation with IL-4, macrophages can assume an epithelioid appearance.
Histiocytes, DCs, and macrophages can be followed along their life cycle from precursors, through activation to mature functional cells by a wide panel of markers, surface and cytoplasmic, and by some of their histochemical characteristics. Table 19.1 lists some of these molecules, many of which are detected by flow cytometry, in situ by immunohistiology, or on frozen sections. A significantly limited panel is available for the recognition of DC and macrophages in fixed embedded tissues, and the panels can also be informative regarding maturation of the histiocytes. For example, PU.1, an Ets family transcription factor, stains nuclei of both macrophages and DCs but also myeloid cells and their leukemias. Few histiocyte-related molecules are informative on their own and are best used as panels of markers for tissue histiocytes. Table 19.2 shows antibodies that react almost exclusively with macrophages, those that are common to both macrophages and DCs (though the amount and pattern can differ), and the antibodies that are more exclusive to DCs. Some are more informative regarding subsets, such as CD1a and Langerin for Langerhans cells and factor XIIIa for dermal and interstitial macrophages. Although CD163 is a more specific tissue marker for the tissue macrophage than CD68, which is a lysosomal marker, epithelioid histiocytes, as seen in epithelioid granulomas and follicular tingible body macrophages, have low to negative CD163 expression. The Th1-induced proinflammatory markers interferon-γ (INF-γ) and TNF-α downregulate CD163 expression on the epithelioid histiocytes, which have high lysozyme expression and variably strong CD68 expression. CD163, however, is not a reliable marker of M2 versus M1 macrophages in tissues, as IL4 and IL13 cytokines in the Th2 response can also downregulate CD163, whereas CD163 + macrophages are invariably present in both Th1- and Th2-associated diseases. One should also be aware of the clone type used for identifying CD68, especially in the evaluation of marrow macrophages as the CD68 KP1 clone also labels myeloid/granulocyte cells unlike the CD68 PGM1 clone. A strong diffuse cytoplasmic staining for CD68 KP1 in extramedullary lesions with absent to weak CD68 PGM1 expression serves as a diagnostic clue for myeloid sarcoma. CD123, the interleukin 3 receptor α, identifies the plasmacytoid DCs. Other molecules are more informative about DC maturation: CD83, DC-LAMP, and hi-fascin indicating an activated and mature phenotype. S100 has long been used as a DC marker when present in high amounts, but it is variably expressed in subpopulations of activated macrophages and is thus of limited discretionary power.
Cluster | Cell Function | Predominant Histiocyte |
---|---|---|
CD1a, b, c | T-cell response to nonpeptide lipids and glycolipids | DC |
CD4 | MHC class II/HIV receptor | M, DC |
CD11b | Complement C3b receptor | M |
CD11c | CD11/CD18 receptor | M, DC |
CD14 | Lipopolysaccharide receptor | M |
CD25 | IL-2 receptor | M |
CD31 | PECAM-1 | M |
CD32 | Fc IgG receptor, low affinity | M |
CD33 | Sialoadhesin | M |
CD49 | Integrin receptors | M |
CD64 | Fc lgG receptor, high affinity | M, DC |
CD68 | Macrosialin, lysosome-associated membrane glycoproteins | M, DC |
CD83 | Ig superfamily | DC |
CD86 | CD28/CD152 ligand | DC |
CD91 | Low-density lipoprotein related protein-1 | M |
CD103 | Integrin α E | DC |
CD116 | GM-CSF receptor α | M, DC |
CD123 | IL-3 receptor α | PDC |
CD163 | Hemoglobin/haptoglobin scavenger receptor | M |
CD169 | Sialoadhesin | M |
CD204 | Macrophage scavenger receptor | M |
CD205 | DEC 205 | M, DC |
CD206 | Macrophage-mannose receptor | M, DC |
CD207 | Langerin | DC |
CD208 | DC-LAMP | DC |
CD209 | DC-SIGN | DC, M |
CD254 | RANK-L | Tumor necrosis factor ligand M |
CD283 | TLR3 | Toll-like receptor D |
CD284 | TLR4 | Toll-like receptor M |
CD303 | CLEC4 | C-type receptor PDC |
Cell Type | Marker |
---|---|
Macrophages | CD14, CD163, acid phosphatase, nonspecific esterase |
Macrophages and dendritic cells | CD68 (PGM1 in bone marrow), HLA-DR, S100 |
DCs | CD1a, Langerin |
Mature DCs | CD83, DC-LAMP, hi-fascin |
Dermal macrophages | Factor XIIIa, CD163 |
Plasmacytoid DCs | CD123 |
Follicular DCs | CD21, CD35, clusterin |
Macrophages derived from recruited monocytes will accumulate at sites of infection and inflammation wherever tissue damage and destruction occur. Scavenging, phagocytosis, and digestion of debris are major functions, although macrophages are also active participants in the subsequent repair and immune reactions. Lymph node reticular sinus histiocytes accumulate largely in response to local draining stimuli, inflammation, or tissue damage. Follicular tingible body macrophages are characterized by the presence of apoptotic debris within the cytoplasm and increase in the presence of apoptotic activity within the dark portion of the germinal center. Normal macrophages can be overloaded by increasing the amount of substrate they are expected to handle. Collections of xanthomatous (yellow) macrophages are also part of the repair and scavenging process following tissue injury, such as perforation of the appendix or ruptured cyst ( Fig. 19.2 ). Excessive turnover of cells with uptake of membrane-derived complex lipids leads to accumulation of ceroid–lipochrome-rich “sea-blue histiocytes,” a term derived from their appearance on Wright- or Wright-Giemsas–type stains. Sea-blue histiocytes are an accompaniment of myeloproliferative disorders in the bone marrow or high platelet phagocytosis in the spleen. Ceroid histiocytosis is also the explanation for most pigmented macrophages in the gastrointestinal mucosa and in chronic granulomatous disease. Intravenous alimentation, especially when lipid is added, can lead to the accumulation of ceroid-filled macrophages in the liver, spleen, and lungs. The atheromatous plaque in large arteries has lipoproteins deposited in excess, and the accumulation of foamy lipid-filled macrophages is characteristic.
Germinal center responses require binding of antigen to follicular DCs that produce IL-6, leading to germinal center expansion, a process inhibited by corticosteroids. Follicular hyperplasia is common in childhood infections and autoimmune and rheumatic disorders. Cytologically atypical follicular DCs are seen in multicentric hyaline-vascular type Castleman disease, probably driven by human herpesvirus 8–derived viral IL-6.
Accumulation of normal sentinel DCs can be seen along their migratory pathway when their numbers are increased in inflammatory processes. The best described is the paracortical DC hyperplasia known as dermatopathic lymphadenopathy but is not unique to that condition. Increased numbers of Langerhans cells from the periphery accumulate in the nodal paracortex in which they can form large, confluent, and nodular aggregates. Because these DCs are maturing or fully mature, they will have the phenotype of interdigitating DCs, S100 + , fascin-hi, CD83 + , DC-LAMP + with interspersed CD1a + /Langerin + Langerhans cells in the paracortex, in contrast to LCH, which is a sinus disease (see Fig. 19.3 ). A diffuse paracortical DC hyperplasia in tonsils or lymph nodes can be an accompaniment of other infiltrating processes and can be misleading by diverting attention from the infiltrate, usually a leukemia or lymphoproliferative disorder.
Macrophages can accumulate at sites of infection or tissue destruction as part of their physiologic role in repair. Osteomyelitis is an example in which the presence of excess macrophages can sometimes mimic that of a neoplastic overgrowth such as Langerhans cell disease or other histiocytosis ( Fig. 19.4 ). Infection in the immune-suppressed individual can also be characterized by unconventional organisms and unusual responses. Atypical mycobacterial disease with mycobacterium avium intracellulare in particular has sheets of macrophages that harbor the organisms. Mycobacterium tuberculosis not only evades phagocytosis but inhibits apoptosis of infected macrophages, presumably by preventing mitochondrial damage and initiating plasma membrane repair. Malakoplakia with the typical Michaelis-Guttmann bodies is a disordered macrophage response to Escherichia coli . There are organisms that find safe haven in macrophages, shielded from other elements of the immune system, best characterized by mycobacterial spindle cell diseases including histoid leprosy and Whipple disease ( Tropheryma whippelii ), leishmania, and rhinoscleroma (Mikulicz cells containing Klebsiella rhinoscleromatis ).
Many organisms can induce epithelioid transformation of macrophages and a granulomatous response, some with a more acute inflammatory response, such as Yersinia, Tularemia, Bartonella, and Brucella spp. Other organisms such as Mycobacterium, Pasteurella, and Burkholderia spp. (the last causing glanders and melioidosis) have a more chronic course and innate epithelioid transformation, often with a giant cell component. Sarcoidosis is the prototype of the noninfectious epithelioid granuloma, but nonsarcoidal conditions such as common variable immune deficiency (CVID) and Blau syndrome ( NOD2 mutations) can also manifest as granulomatous disease. Giant cell formation is also seen in a number of noninfectious circumstances with collections of epithelioid histiocytes at various tissue sites, with sarcoidosis and Crohn disease as common examples. Beryllium exposure is less frequent. Foreign bodies of many kinds can lead to macrophage accumulation, often with a giant cell component, and some aggregates may be epithelioid. Talc, silicone, starch, and detritic granulomas from prostheses are examples of exogenous material, but crystal-storing histiocytes may accumulate endogenous crystals that are derived from immunoglobulin ( Fig. 19.5A ). Drugs such as paclitaxel (Taxol) can lead to accumulation of histiocytes with periodic acid–Schiff (PAS)-positive inclusions, whereas clofazimine produces histiocytes containing red crystals (see Fig. 19.5B and C ). Some, but not all, foreign bodies are birefringent, and polarization is mandatory when confronted with a histiocytic aggregate of unknown type.
The xanthomatous (yellow) macrophage response in tissue repair is a feature when inflammatory cells and tissue necrosis are being scavenged. There are sites where the xanthomatous inflammation can be excessive and even simulate tumor in its extent. Xanthogranulomatous pyelonephritis is a chronic pyelonephritis of adults in which an exuberant xanthomatous process can replace much of the kidney and even extend into the retroperitoneum. It is rare in children. A similar process can affect the gallbladder, and bacteria can be isolated from most. There are instances of so-called necrobiotic xanthogranulomas in soft tissue, orbit, and rarely viscera, sometimes in association with paraproteinemia. These xanthogranulomas are not infectious in nature. A xanthoma is a mass or nodule composed almost entirely of lipid-rich foamy macrophages. Most are associated with hyperlipidemias, both primary inherited forms and secondary hyperlipidemic states such as diabetes mellitus, cholestatic liver disease, or nephrotic syndrome. Xanthomas occur at various sites on the skin in crops as eruptive xanthoma, xanthelasma around the eyes, and tendinous xanthoma around the ankles, knees, hands, and elbows. The larger xanthomas may be associated with a center that contains fat necrosis, cholesterol crystals, and a surrounding foreign-body type giant cell reaction. Atherosclerotic lesions are characterized by xanthomatous macrophages. A history of hyperlipidemia is the rule in true xanthomas.
Because of their specialized peculiar functions, macrophages are rich in lysosomes. In a number of inherited defects of the lysosomal apparatus, there is a deficiency of enzymes needed to degrade macromolecules or to transport the degraded substance out of the lysosome, leading to accumulation of naturally occurring metabolites within lysosomes (lysosomal storage disorders). The classes of molecules involved include the mucopolysaccharidoses (MPS I-VII), glycoproteinoses, glycogenosis (type II), sphingolipidoses, lipidoses, the multiple enzyme deficiency disorders, and the lysosomal transport defects. Each of the defects included within these categories will have a peculiar tissue distribution depending on the substrate that fails to be catabolized or transported. The accumulation of the various substances will produce highly different phenotypic expression when viewed by light or electron microscopy, and both of these modalities have been combined with histochemistry to categorize the various disorders ( Fig. 19.6 ). On occasion, discovery of the storage cells in tissues might be the first clue to the presence of a metabolic disorder, and the physical characteristics of the storage material can lead to the biochemical or genetic testing needed for confirmation.
Considering the numerous functions that macrophages have in health and disease, it seems remarkable that so few defects have been documented that are restricted to these cells. Apart from the inherited lysosomal storage disorders, some examples often share the deficiency with other cells. Autosomal-recessive osteopetrosis affects osteoclasts from the same hematopoietic precursors as macrophages. When the responsible gene is TCIRG1, it affects the vacuolar proton pump of osteoclasts and gastric parietal cells. Fewer instances are caused by the CLCN7 gene mutation that encodes a chloride channel. Hematopoietic stem cell transplant in the TCIRG1 but not the CLCN7 form can arrest the disease by providing functional monocytes and osteoclasts. The effect of these mutations on other macrophages, if any, is not known.
Leukocyte adhesion deficiency due to defects in β-integrins hinders the ability of granulocytes and monocyte-derived macrophages to accumulate at infection sites. Chronic granulomatous diseases involve the inability to provide an oxidative burst important in microbial killing. The responsible defects resulting in abnormal electron transfer from cytoplasmic nicotinamide adenine dinucleotide phosphate (NADPH) to molecular oxygen are common to neutrophils and macrophage abnormalities of gp91 phox and include a macrophage-specific component of NADPH oxidase.
Chronic granulomatous disease (CGD) can develop for the first time as lymphadenopathy, and nodal involvement is seen in 50% of patients. Active chronic inflammation and granulomas occur, and occasionally pigmented macrophages might provide a clue to the diagnosis. The pigment is lightly PAS positive and has the features of ceroid, the ultrastructural equivalent being lysosomal debris. The most frequent organisms in current experience are Staphylococcus aureus, followed by Burkholderia cepacia, Serratia marcescens, Nocardia species, Aspergillus species, Salmonella species, and bacille Calmette-Guérin. Confirmation of the diagnosis of CGD is by direct measurement of superoxide production, ferricytochrome c reduction or dihydrorhodamine oxidation.
Hermansky-Pudlak syndromes with oculocutaneous albinism and platelet storage disease are caused by defects in intracellular protein trafficking that can affect lysosomes in some macrophages and melanocytes. Alveolar macrophages may be targeted in Hermansky-Pudlak syndrome type 1, in which progressive pulmonary fibrosis occurs.
The emerging field of toll-like receptors (TLR), their normal variation and defects, is of importance because of their widespread expression on macrophages. Recent reviews provide insights into research investigating the expression of TLR in macrophages of patients with autoimmune disease, which may be implicated as potential targets of therapy. For example, increased TLR5 was found in the synovium and synovial-associated macrophages in patients with rheumatoid arthritis, as compared to control subjects. MicroRNAs, small non-coding RNAs that are a class of gene expression regulators, may also play a role via their induction and regulation by TLR stimulation in macrophages, which appears altered in some autoimmune disease.
Kikuchi-Fujimoto disease (histiocytic necrotizing lymphadenitis) is a self-limiting condition characterized by proliferation and accumulation of histiocytes of unclear origin. The benign lymphadenopathy is associated with systemic symptoms and fever. Paracortical areas are filled with histiocytes, plasmacytoid DCs, CD8 + T cells, and karyorrhectic debris. Later changes are more xanthomatous. The diagnosis is histopathologic by exclusion of other causes of inflammatory lymphadenopathy.
Tangier disease owing to high-density lipoprotein deficiency leads to the accumulation of cholesterol esters in lymphoid tissues, tonsils, lymph nodes, liver, and spleen. Functional effects on macrophage physiology have been documented for Gaucher disease. Hemophagocytic syndromes are also a serious consequence of the Griscelli syndrome, Chediak-Higashi syndrome, and Hermansky-Pudlak syndrome diseases (see below under Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndromes).
DC dysfunction has been identified in the X-linked Wiskott-Aldrich syndrome in which Wiskott-Aldrich syndrome protein is functionally defective. Macrophages, DCs, and osteoclasts are affected and have disordered motility because of their inability to form podosomes. Common variable immune deficiency and CD40 defective hyperimmunoglobulin M disease may have diminished DC function, but it is not established if this is the primary defect. Because DCs can act as the portal for viral infection, they may be increased or selectively depleted in some viral infections, notably human immunodeficiency virus (HIV).
Hemophagocytic lymphohistiocytosis (HLH) is a severe, potentially fatal systemic inflammatory activation disorder (not a primary histiocytic disorder) characterized by a regulatory disorder of T cells leading to unbridled macrophage activation and inflammatory cytokine upregulation. Despite the name, the pathologic features of hemophagocytosis are neither specific nor sensitive in the overall diagnosis and constitute only one of the eight diagnostic criteria. Historically HLH has been classified as primary or familial based on an underlying genetic defect in lymphocyte cytotoxic defects. However, in the recent revised classification of histiocytoses, an expansion of this group to include other mendelian inherited conditions associated with HLH and immune impairment was proposed. Secondary or acquired HLH is now known as apparently non-mendelian HLH, which includes infection- related HLH, malignancy-associated HLH, transplant-related HLH, HLH associated with a defined rheumatologic condition, and HLH of unknown/uncertain origin. Historically macrophage activation syndrome (MAS) has been noted as a symptom of HLH occurring in a patient with an underlying rheumatic disorder (e.g., systemic juvenile idiopathic arthritis and its adult equivalent, adult-onset Still disease, systemic lupus erythematosus, and Kawasaki disease). Recognition of the clinical features of MAS belonging to the spectrum of HLH has led to the proposal of the term MAS-HLH for this “secondary” or so-called acquired HLH.
Secondary HLH, including rheumatic-driven MAS-HLH, is thought to be incited by an acquired (i.e., infectious, malignancy, drug) or non-mendelian source, unlike primary HLH, in which the etiology is derived from a genetically programmed defect in immune regulation, most notably of the cytotoxic lymphoid system. However, secondary HLH is now best considered as “apparently non-mendelian HLH” as an increasing number of cases in this group are revealing genetic mutations that lead to at least partial immune/cytokine function.
Although macrophage functional patterns have been described as type 1 (Th1-driven) or type 2 (Th2-driven), a wide variety of microenvironmental factors result in a large number of different and overlapping functional and phenotypic subtypes that are constantly being modulated. Macrophages can be immune activated by soluble factors, such as cytokines, with interferon-gamma (INF-γ) being the most classic example, but TNF, IL-1, IL-2, and macrophage migratory inhibitory factor (MIF) can also cause proinflammatory activation, as can lipopolysaccharide. These classically activated macrophages are referred to as M1 macrophages . Activated M1 macrophages participate in both antimicrobial immunity, by enhancing opsonization and phagocytosis, as well as in cellular immunity by stimulating or inactivating lymphoid receptors, largely through their active cytokine production and other proinflammatory mediators. Deactivation of M1 can be mediated by IL-4, IL-10, IL-13, and TGF-β. IL-4 and IL-13 can also induce an alternative anti-inflammatory activation, mediated by M2 macrophages, that enhances major histocompatibility complex class II expression and mannose receptor-mediated endocytosis. Alternatively activated (i.e., M2) macrophages are heterogeneous in terms of the cytokines that induce their differentiation; they play a role in tissue remodeling and repair, resistance to parasites, immunoregulation, and tumor promotion. Macrophages in HLH and MAS-HLH exhibit many characteristics of M2 macrophages, including the membrane and cytoplasmic expression of scavenger receptor CD163, but, as stated previously, CD163 is not a reliable M2 marker as tissue stains cannot reliably differentiate between M1 and M2 macrophages
In response to a number of different stimuli, a systemic inflammatory or macrophage activation syndrome can develop. Viral infections are the best known, but bacterial infections, rheumatologic disorders, especially the systemic form of juvenile idiopathic arthritis, cancers, lymphoproliferative disorders, intravenous alimentation, and multiple organ failure are among other causes. Although the pathogenesis is unclear, in general persistently activated macrophages appear to be caused by a defect in granule-mediated cell cytotoxicity, leading to an unrestrained lymphocyte–NK cell–driven macrophage stimulation that leads to disseminated overactivity (e.g., enhanced antigen presentation) of the macrophages throughout the body, possibly mediated by TNF-α, ineffective deactivation through depressed NK and cytotoxic CD8 T cells, and repeated INF-γ–dependent stimulation of TLR. Also, defects in perforin and FAS systems lead to unbridled immune activation via their loss in maintaining homeostasis of DC cells and T-cell activation by antigen presentation. An increase in the number and size of endogenous macrophages, with or without hemophagocytosis, is a feature of this condition that is best seen in the bone marrow but also in spleen, liver, and lymph nodes ( Figs. 19.7, 19.8, and 19.9 ). The exaggerated inflammatory response leads to a hypersecretion of proinflammatory cytokines (INF-γ, TNF-α, IL-1, IL-4, IL-6, IL-8, IL-10, IL-18) (e.g., “cytokine storm”). In its most severe expression, there are significant functional effects of the cytokine storm, with bone marrow depression, hepatomegaly with hepatocellular damage and raised hepatocellular enzymes, and effects on the clotting cascade. Treatment is centered on halting the cycle of inflammation that is best abated when the inciting condition is treated or disappears. In severe instances, without resolution, it can lead to progressive organ failure and can be fatal. Thus treatment measures include supportive care, directing the underlying inciting event, and variable success with high doses of steroids, cyclosporine (CSA), or agents aimed at counteracting the effects of TNF-α, such as infliximab and etanercept. The combination of etoposide (VP-16), a cancer chemotherapeutic drug with potent cytotoxicity against activated T cells, and dexamethasone is standard of care based on HLH-94, the international prospective clinical trial sponsored by the Histiocyte Society. In the subsequent HLH-2004 trial, preliminary results did not see additional benefit in adding CSA to the VP-16/dexamethasone backbone during the first 8 weeks of therapy.
Genetic forms of HLH (formerly “familial HLH”) refer to a group of disorders with clinical and pathologic manifestations of an acute and often fatal form of uncontrolled macrophage activation owing to defective apoptosis, cytotoxic T or NK function, or inappropriate inflammasome function. Traditionally, familial hemophagocytic lymphohistiocytosis (fHLH) has been referred to a small group of inherited disorders that have the clinical and pathologic manifestations of an acute and usually fatal form of uncontrolled macrophage activation owing to defective apoptosis or cytotoxic function of cytotoxic cells (e.g., cytotoxic T cells, T-regulatory cells, or NK cells). This group has recently expanded to include select genetic disorders that also display features of macrophage activation via inappropriate inflammasome or inflammation activation and are now collectively referred to as genetic (mendelian inherited) conditions leading to HLH.
The clinical features are similar to secondary HLH, dominated by the effects of a hypercytokine state with a wide range of clinical severity. Diagnostic guidelines were originally developed in the context of a pediatric cohort by the Histiocyte Society that include clinical, laboratory, and histopathologic findings because the constellation was important, but definitive molecular diagnosis is now available for some. A family history of parental consanguinity or prior sibling death may be elicited. Prolonged fever and cytopenias, most commonly anemia and thrombocytopenia within the first 2 years of life, may be associated with neurologic and meningeal signs and symptoms, hepatomegaly with evidence of hepatocellular dysfunction, splenomegaly, and sometimes a skin rash. Laboratory features include hypofibrinogenemia alone or with hypertriglyceridemia, hyperferritinemia, and high levels of circulating soluble CD25 (IL-2 receptor). NK cell function is low or absent.
A group of inherited conditions resulting in uncontrolled macrophage activation caused by defective apoptosis or cytotoxic function of lymphocytes (e.g., T or NK cells), with newer cases now including genetic abnormalities of inappropriate inflammasome activation.
0.12 to 0.15/100,000 children per year (Sweden systemic)
Mortality is 100% untreated, around 50% overall 5-year survival with bone marrow transplantation protocols (HLH-94/HLH-2004)
Male-to-female ratio = 1, although a slight male predominance may be related to X-linked lymphoproliferative disorders
Race distribution unknown but appears to occur in all races and ethnic groups
Median age 2 to 3 months; 90% <2 years
Rare in older children or adolescents, but late occurrences likely attributed to one or more HLH-associated gene sequence variants
A molecular diagnosis consistent with HLH diagnosis
Abnormalities of cytotoxic defects
FHL1: Gene unknown 9q21.3-22
FHL2: PRF1 gene 10q22.1
FHL3: Munc 13-4/ UNC13D gene 17q25
FHL4: Syntaxin-11/ STX11 gene 6q24
FHL5: Munc18-2/ STXBP2 gene 19p13
X-linked lymphoproliferative syndrome type 1 (XLP-1): SAP/SH2D1A Xq25
Griscelli syndrome type 2: RAB27A gene 15q21
Chediak-Higashi syndrome: LYST1 gene 1q42.1-42.2
Abnormalities of inflammasome overactivation: XLP2/ BIRC4 and NLRC4 .
Abnormalities affecting inflammation SLCA7 gene
Flow cytometry with elevated granzyme B and either undetectable perforin, diminished signaling lymphocyte activation molecule (SLAM)-associated protein (SAP), or diminished X-linked inhibitor of apoptosis (XIAP) protein expression ( SAP and XIAP only in male patients )
Diminished or absent CD107a degranulation assay (normal in FHL-2, XLP, and secondary HLH)
In the absence of genetic information, diagnosis of HLH requires five of the following eight features a
a See also Jordan et al: Blood 2011;118(15):4041–4052. When considering “familial” or genetic HLH without direct evidence of a genetic mutation, presence of malignancy or other known trigger must also be excluded. If clinical criteria are met, a provisional diagnosis of “familial (apparently mendelian) HLH” of unknown/uncertain origin can be made.
Fluctuating fever
Hepatosplenomegaly
Bicytopenia, especially thrombocytopenia (hemoglobin <9 g/L, neutrophil <1.0 × 10 e9 /L, platelet <100 ×10 e9 /L)
Hyperferritinemia (>500 µg/L; although >3000 µg/L concerning and >10,000 µg/L higher sensitivity and specificity) a
Hypertriglyceridemia (≥3 SD fasting values) and/or hypofibrinogenemia (≤3 SD of normal for age)
Low or absent NK-cell function
Increased soluble CD25 (>2400 U/mL or >2 SD)
Hemophagocytic histiocytosis (bone marrow, liver, spleen, cerebrospinal fluid)
Gross findings are nonspecific and are limited to hepatomegaly and splenomegaly
Hemophagocytic histiocytosis in bone marrow, spleen, lymph nodes, liver, cerebrospinal fluid
Liver with features of chronic hepatitis, portal lymphohistiocytosis, and Kupffer cell activation with hemophagocytosis, noted in fHLH abnormalities of cytotoxic defects
Early: Bone marrow possibly cellular and active in the face of cytopenias; Late: progressive depletion of hematopoiesis and increasing content of large phagocytic histiocytes
Cerebral spinal fluid with pleocytosis (mononuclear cells and activated macrophages)
Histiocytosis with hemophagocytosis on bone marrow or spleen
Increased numbers of marrow histiocytes CD68 (PGM-1) and CD163
Increased macrophages in spleen and liver CD68, CD163
Secondary HLH will show similar pathologic features and cannot be distinguished by pathologic features alone. Viral infection, especially EBV, rheumatologic conditions (MAS-HLH), and T-cell malignancies, should always be excluded
Hemophagocytosis and erythrophagocytosis in minor transfusion reactions
Gross findings are nonspecific and are limited to hepatomegaly and splenomegaly.
An excess of activated macrophages is the hallmark, usually with prominent hemophagocytosis. These cells can be found in the bone marrow, spleen, and lymph nodes. Cerebral spinal fluid with pleocytosis, including mononuclear cells and activated macrophages, with variable hemophagocytic cells may be found in central nervous system (CNS) involvement (see Fig. 19.8 ). The macrophage activation of hemophagocytosis can be cyclical, and intervening biopsy specimens (e.g., marrow) may be negative at first examination or during troughs of activity with gradual replacement of the marrow by increasing contents of large phagocytic histiocytes and progressive depletion of hematopoiesis. In the liver, the infiltrate is largely portal with the accompaniment of T lymphocytes (i.e., lymphohistiocytosis). Liver pathology in cases of fHLH associated with cytotoxic defects includes endothelialitis of the portal and central veins, often with free-floating macrophages in lumen of the veins. Lymphocyte-mediated bile duct injury is also a typical feature. In addition to this chronic hepatitis-like appearance, three other histopathologic patterns have been described in fHLH with cytotoxic defects. These are a leukemia-like pattern, a histiocyte storage-like pattern, and a neonatal giant cell hepatitis-like pattern. Other genetic forms of HLH and secondary HLH may or may not display these same hepatic features.
The genetic forms of HLH associated with cytotoxic defects (fHLH) include a handful of known genetic disorders. FHL1 (Online Mendelian Inheritance in Man [OMIM] no. 267700) is mapped to 9q21.3-22, but the genetic defect is not yet determined. Approximately 30% to 50% of patients have mutations in the perforin/ PRF1 gene FHL2 (FHL2, OMIM no. 603553), which is mapped to 10q22.1 and whose function is to induce apoptosis. Most defects lead to reduced or absent perforin expression on peripheral blood cells, a feature that can be documented by flow cytometry. FHL3, mapped to 17q25, is the form characterized by mutations in the Munc 13-4/ UNC13D gene (FHL3, OMIM no. 608898), which is important for vesicle priming for perforin transport as well as docking and fusion of cytotoxic granules with the cytoplasmic membrane, but perforin content is normal. Together FHL2 and FHL3 tend to make up over half of fHLH cases. A smaller frequency will have FHL4, mapped to 6q24 (OMIM no. 603552), which is caused by mutations in the Syntaxin-11/ STX11 gene, important for vesicle transport. FHL5, mapped to 19p13 (OMIM no. 613101) is caused by mutations in the Munc18-2/syntaxin-binding protein-2 gene ( STXBP2 ), which is important for vesicle trafficking and release, which can be associated with colitis, bleeding tendency, and hearing loss. Three additional fHLH resulting in lymphocyte cytotoxic defects include Griscelli syndrome type 2, mapped to 15q21, caused by a mutation in the RAS-associated protein/ RAB27A gene, one of the MUNC 13-4 effector molecules important for signal transduction and that results in partial albinism. The second is Chediak-Higashi syndrome mapped to 1q42.1-42.2, caused by a mutation in the lysosomal trafficking regulator/ LYST1 gene, which is important for vesicle transport, resulting in partial albinism, susceptibility to infection, and neuropathy. The third is X-linked lymphoproliferative syndrome (XLP) type 1 (XLP-1) mapped to Xq25, which is caused by a mutation in the SAP / SH2D1A gene, which is important in signal transduction and which results in hypogammaglobulinemia, with a heightened sensitivity for Epstein-Barr virus (EBV) infection and lymphoma.
A small but steadily increasing number of additional mutations not related to lymphocyte cytotoxic defects may also lead to primary HLH-like disorders causing hyperinflammation through an overstimulated, innate immune system and macrophage activation. These include two mutations with abnormalities of inflammasome overactivation (XLP2/ BIRC4 and NLRC4 ). Gain-of-function mutations discovered in the inflammasome gene, NLRC4 , have also demonstrated HLH-like findings, including macrophage activation, with some showning an earlier (e.g., neonatal) onset including bowel inflammation. The inflammasome is a complex of innate immune system receptors that integrate danger recognition signals and caspase-1–induced inflammation to stimulate the release of proinflammatory cytokines IL-1β and IL-18. X-linked inhibitor of apoptosis protein/X-linked lymphoproliferative (XIAP/XLP-2), mapped to Xq25, results in mutations of the XLP2/ BIRC4 gene. This mutation has been known to be a genetic variant of HLH, often EBV driven, causing mild and recurrent HLH and colitis symptoms with splenomegaly and hypogammaglobulinemia in some.
Other genetic forms of HLH that result in mutations affecting inflammatory pathways include abnormalities in the lysinuric protein intolerance/ SLCA7 gene, detected in infants with elevated plasma ammonia after a protein-rich meal with failure to thrive and muscle hypotonia.
Whereas there are reports of HLH-like features in other primary immunodeficiencies, including Hermansky-Pudlak-2 syndrome (OMIM no. 608233) and AP3 complex/ AP3B1 gene, which results in congenital neutropenia and immunodeficiency, it is generally considered that these HLH-like syndromes are not true primary HLH because they are often triggered by EBV with B-cell lymphoproliferation.
The demonstration of an excess of activated macrophages is key to the recognition of the condition, but a definitive diagnosis requires genetic documentation and/or appropriate clinical/laboratory criteria. Increased and activated marrow macrophages in HLH are demonstrated using CD163 or anti-CD68 PGM-1 antibody that has little cross-reaction against hematopoietic precursors, allowing identification of hemophagocytosis. At all other sites, CD68 (KP-1 or PGM-1) and CD163 will demonstrate the macrophages effectively, including in the portal areas and sinusoids of the liver. In patients who have a perforin mutation, the CD3/CD8 cells do not stain for perforin by immunohistochemistry, if available (see Fig. 19.9 ).
Marrow is the most commonly examined aspirate, although splenic puncture is more commonly done in Europe. In both instances, the presence of great numbers of large cytoplasm-rich but bland macrophages that have hemophagocytosis, especially erythrophagocytosis, is typical during active episodes.
The most important differential diagnosis is distinguishing a genetic primary form of HLH from the secondary HLH syndromes that are treated by targeting the inciting condition: infection (e.g., viral, bacterial, fungal, parasite), rheumatologic condition (HLH-MAS), malignancy (hematologic, solid tumor), treatment related (chemotherapy, hyperalimentation with fat overload syndrome), or transplant related. An established molecular diagnosis, when available, is definitive for primary HLH, but some cases of secondary HLH may be related to partial impairments of known mutations in cytolytic function. This has made the distinction between primary and secondary HLH less definitive. Furthermore, the clinicopathologic features in individual instances are so variable, and infection can serve as the trigger for a crisis in the primary genetic forms of HLH, mimicking a secondary infection-associated HLH. It should also be noted that the most common cause of erythrophagocytosis is not hemophagocytic syndrome but a minor transfusion reaction.
The genetic forms of HLH are rapidly fatal, with a mean survival of 2 months unless treated effectively, including bone marrow transplantation. The HLH-94 protocol (now HLH-2004), which is based on fHLH of cytotoxic defects, stabilizes patients using immunosuppression and chemotherapy (steroids, etoposide, and anti-T–cell agents) and follows that with bone marrow transplantation, the 5-year post-transplant survival probability is approximately 54% ± 6%. In genetic forms of HLH, the standard of care is hematopoietic stem cell transplantation, as it remains the only curative therapy in these patients. More recently, reduced intensity chemotherapy with alemtuzumab (a monoclonal antibody targeting CD56 on mature lymphocytes but sparing stem cells) is given 2 weeks prior to bone marrow transplant and has resulted in improved survival. Treatment of HLH-like conditions with mutations outside of classic lymphocyte cytotoxic defect group (i.e., inflammasome disorders) may require further evaluation of IL-1 and IL-18 blockage.
Cytophagic histiocytic panniculitis appears to be a localized expression of HLH. Some cases of cytophagic histiocytic panniculitis are associated with instances of HLH, including a recent association with STX11 (FHL4), but the local effect may predominate. Other instances appear to be examples of infection-associated secondary HLH, most commonly harboring EBV in the panniculitis, confirmed by in situ hybridization staining of EBV-encoded small RNA (EBER) probe and sometimes also associated with a clonal T-cell proliferation. However, both primary HLH and infection-associated HLH must be separated from subcutaneous panniculitis-like T-cell lymphoma (SPTCL) in which the lymphoma drives the histiocytic component and atypical lymphocytes encircle the fat vacuoles. Most are examples of the systemic SPTCL-αβ with clonal αβ rearranged T cells that are CD3 + CD8 + cytotoxic cells. There is loss of CD2, CD5, and CD7 in up to 50% of cases.
Box 19.1 lists the most updated classification of these disorders by the Histiocyte Society and a more recent but limited classification by the World Health Organization. Of note, although the 2016 revised classification of histiocytosis and neoplasms of the macrophage-dendritic cell lineages includes HLH, as noted above, HLH is not a primary histiocytic disorder but rather a regulatory disorder leading to unbridled macrophage activation and hyperinflammatory cytokine upregulation.
Langerhans cell histiocytosis is a disorder characterized by clonal proliferation of abnormal histiocytic cells that while having the phenotypic expression of Langerhans cells (LC) (CD1a + /Langerin + ) share no common precursor with the similarly named epidermal cells. Gene expression profiling comparing LCH cells to epidermal LCs and immature myeloid DCs has suggested that LCH arises from myeloid dendritic cell precursors. LCH cells have high expression of markers associated with myeloid DC at various stages of differentiation (CD33, CD13, CD11c, CD11b, CD66c, CD200LF), and less epithelial marker expression (EpCAM, E-cadherin, and CD36) as compared to normal epidermal LCs. A bone marrow-derived myeloid/monocyte precursor has been implicated in LCH based on mouse models and human data by . This has led the field in the Histiocyte Society to recognize LCH as an inflammatory myeloid-derived neoplasm with demonstration of BRAF -V600E mutations in bone marrow hematopoietic precursors and circulating blood myeloid/monocytes. LCH pathogenesis is now believed to be based on timing of the mitogen-activated protein kinase/extracellular signal-regulated kinases (MAPK/ERK) pathway activation in myelopoiesis, which may help explain the different clinical phenotypes in LCH ( Fig. 19.10 ). More detailed reviews on the cell of origin, pathogenesis, and molecular aspects of LCH are beyond the scope of this chapter and available elsewhere.
Langerhans cell histiocytosis affects children primarily, but there is an increasing incidence in adults. In children, the disease can be limited to a single site unifocal (e.g., bone, soft tissue, skin lesions) or involve multiple foci in a single system (multifocal bone or multiple lymph node lesions), or there may be multisystem (MS) disease that involves two or more organ systems. Involvement of some organs confers a higher risk of morbidity and mortality (MS-LCH high risk); these are the liver, bone marrow, and spleen. In children, pulmonary disease occurs in 25% of cases with multisystem LCH; however, it is no longer considered an independent risk organ. Penetrating skull lesions of the craniofacial bones and temporal vault carry an increased risk of future CNS sequelae, including diabetes insipidus (e.g., “CNS-risk” sites). Mortality in these higher risk children with multisystem disease, often younger than 2 years, has improved, with 3- and 5-year survival of 84% following the LCH III protocol, with therapy intensification after 6 weeks if active disease persists, along with prolonged therapy of 1 year rather than 6 months.
In adults, the disease often affects the lungs of smokers, although bone and skin lesions occur. Common but nonspecific B-type symptoms can include fever, weight loss, and fatigue. The specific clinical manifestations depend on the site of involvement. Bone involvement, monoostotic or polyostotic, is the most frequent presentation; it is commonly associated with pain, pathologic fractures, vertebral collapse (e.g., vertebra plana), and local soft-tissue extension. Skin involvement often involves flexures and the scalp with a seborrheic rash that may be petechial. Newborns have more papular lesions that commonly regress. Liver disease may present with abnormal liver function tests (e.g., gGTP, bilirubin), hepatomegaly, or jaundice. Clinically a biopsy may not be necessary to prove liver involvement if these features are present in the setting of known LCH disease. The liver involvement is typified by infiltration of the major bile ducts that causes a sclerosing cholangitis and biliary strictures and ends in a biliary cirrhosis with eventual need for liver transplantation. The lungs are more commonly involved as part of multifocal disease in children but may often be the only site in smoking adults. Interestingly, some children who have recovered from early childhood disease have developed pulmonary LCH when they began smoking as adolescents. Pulmonary involvement may lead to dry cough, dyspnea, pneumothorax, interstitial and peribronchial fibrosis, and respiratory compromise. CNS involvement by active LCH involves the hypothalamic-posterior pituitary axis, causing irreversible diabetes insipidus, or growth retardation, and it can involve the choroid plexus and meninges. Late CNS involvement, years later, has historically been described as indirect without active LCH; however, newer data may be challenging that notion with the possibility of a BRAF mutant precursor cell driving the late demyelinating progressive and debilitating neurodegenerative disease of LCH (ND-LCH). Clinicoradiographically ND-LCH is characterized by bilateral, usually cerebellar demyelinating foci resulting in progressive ataxia, dysarthria, nystagmus, hyperreflexia, dysdiadochokinesia, dysphagia, and blurred vision. Hematopoietic involvement is seen with bone marrow and spleen infiltration, but the cytopenias that are common are functional (possibly cytokine mediated) because the hematopoietic marrow is preserved in many sampled cases. Lymph nodes and the gastrointestinal tract can be involved, resulting in lymphadenopathy and sometimes life-threatening malabsorption, anemia, bloody diarrhea, and protein-losing enteropathy, whereas it may be asymptomatic or present as a polyp-type lesion in adults. The kidney and gonads are almost never affected. Because of the widespread and protean possibilities, LCH when first diagnosed at any site is evaluated by staging, a clinical survey that maps the extent of involvement. Referral centers also advocate mutational studies in the MAPK/ERK pathway, including sensitive polymerase chain reaction (PCR) BRAF studies on the lesion and, if BRAF -V600E mutation positive, upfront testing of peripheral blood for circulating BRAF -mutated precursor cells to help follow disease progression.
LCH is a heterogenous group of disorders, which is now considered by the Histiocyte Society as a myeloid inflammatory neoplasia with accumulation of bone marrow–derived Langerhans cell histiocytosis (LCH cells)
Multisystem (MS-LCH): Two or more organs involved (e.g., bone and skin)
Multifocal LCH: Multiple sites within one organ system (e.g., multiple bone lesion)
Single system (SS-LCH): Single site (e.g., bone)
0.54 to 0.9 per 100,000 children (Denmark, Sweden) and unknown in adults; probably underestimates the true incidence since many are not reported
Rare familial cases
Bone and soft tissue (most common), skin, lymph node, liver, bone marrow; lung in smoking adults
Highest mortality in those younger than 2 years (30% to 50%) who have “risk organ” involvement (e.g., liver, marrow, spleen), but treatment according to LCH III protocol has shown improved survival
High risk of diabetes insipidus and neurodegenerative sequelae with “CNS risk” involvement (e.g., craniofacial bones, temporal skull bone)
Morbidity high for multiple bone lesions and adult multisystem or lung disease (91% 5-year survival)
Unifocal bone disease and some cases of isolated congenital skin lesions ( BRAF -V600E wild type) usually self-limiting if staging reveals no other involvement
Variable male predominance 1–2 : 1 in children, 1.4 : 1.2 in adults
Age range: from neonates to adults
Neonatal skin-only papular disease, may be self-limited (“Hashimoto-Pritzker disease”)
Systemic disease (MS-LCH), <2 years, often with risk organ involvement (“Letterer-Siwe disease”)
Multiple bone, soft tissue disease (MS-LCH), 2 to 10 years (“Hand-Schüller-Christian disease”)
Solitary bone or soft tissue disease (SS-LCH), 5 to 15 years (“eosinophilic granuloma”)
Adult disease, mean age 33 years (SD = 15 years), 30% single system, 70% multisystem
Symptoms vary according to site and organs involved
Skeletal pain, soft tissue mass in bone lesions
Extraosseous manifestations: diabetes insipidus, seborrheicskin rash, pulmonary abnormalities, lymphadenopathy,hepatosplenomegaly, and pancytopenia
Osteolytic intramedullary bone lesions, poorly demarcated in early and active lesions
Osteolytic intramedullary bone lesions, well demarcated by sclerotic rim in older and regressing lesions
Peribronchial and interstitial lung changes progressing to honeycombing in some
Loss of normal pituitary signal and thickening of the pituitary stalk
Variable symmetric magnetic resonance imaging (MRI) signal intensity changes of the cerebellum, basal ganglia, and/or pons along with dilated Virchow-Robin spaces in neurodegenerative LCH disease
Localized skin and bone lesions can be observed or resected, treated intralesionally or topically with steroids, or given low-dose radiation to bone
Multisystem LCH given systemic chemotherapy according to LCH III protocol
Recurrent or refractory disease; 2 chlorodeoxyadenosine (2CDA) or bone marrow transplantation; newer studies investigating clofarabine and BRAF inhibitor therapy
Adult lung disease: cessation of smoking, corticosteroids
Once a diagnosis is established, an age-appropriate survey of the skeleton to look for other lesions is often performed. More sensitive modalities such as whole-body fludeoxyglucose-positron emission tomography (FDG-PET) scans can localize new lesions and are more informative of disease activity. In the early stages, expanding new bone lesions may have a rapidly growing, lytic appearance with poorly defined margins that mimic malignant disease. The presence of pathologic fractures can confound the picture. Older and involuting lesions have an osteolytic center and sclerotic, sharply defined borders, and the differential diagnosis includes low-grade lesions such as chronic osteomyelitis. Cranial bone involvement with adjacent soft-tissue disease is characteristic ( Fig. 19.11 ). Computed tomography (CT) or magnetic resonance imaging (MRI) may help to ascertain the extent and character of the bone lesion before biopsy. Early lung involvement is best demonstrated by high-resolution CT that reveals a delicate interstitial and cystic change that is bronchocentric and spares the intervening lung. Late disease is nodular, fibrotic, scarring, and honeycombing on imaging studies. Pleural bullae are responsible for the bouts of pneumothorax. Imaging of the liver may demonstrate the features of a sclerosing cholangitis. Central nervous system disease in the early infiltrative phase is characterized by posterior pituitary stalk involvement, and there can be space-occupying choroid or meningeal masses. Late disease is seen best by MRI that reveals the symmetric neurodegenerative foci of the cerebellum and basal ganglia.
Bone lesions may be sharply demarcated or gray to yellow, depending on lipid content
Lung and thymic lesions may be cystic
Cells are oval and measure 15 to 25 µm, usually nested in sheets or clusters
Nuclei are oval or grooved or have complex foldings
Nucleoli are inconspicuous
Cytoplasm is pale
Binucleated and multinucleated forms are common, with osteoclast-like cells in bone and contiguous soft-tissue lesions
Mitoses are variable, never atypical; necrosis may be present
Eosinophils are variable and may be absent to overwhelming, with microabscesses filled with Charcot-Leyden crystals
Plasma cells are rare within in the histiocytic aggregates
Oval histiocytes with CD1a/Langerin and high sensitivity and specificity
S100 nuclear/cytoplasmic, sensitive not specific; vimentin
CD68 and HLA-DR-low, paranuclear
VE1 dark cytoplasmic granular staining in BRAF -V600E mutated cases (~50% of cases)
Mutations in BRAF -V600E (50-64%), MAP2K1 (10% to 20%) , case reports of BRAF in frame deletions, ERBB3, RAS, ARAF , and PIK3CA-AKT, with about 15% of cases still unknown
BRAF gene fusions now being discovered
BRAF -V600E detection should include sensitive molecular methods for accurate detection, especially given the variable low content of histiocytes in some cases
The LCH cells have sparse organelles with Birbeck granules required for ultrastructural confirmation; however, this has been replaced by Langerin (CD207) immunohistochemistry
Classic dendritic cells and intervening macrophages
Bone, lymph node, thyroid, and hypercellular lesions with histiocytes having moderately abundant pale cytoplasm and a grooved or folded nucleus
Eosinophils possibly not obvious
Bone osteomyelitis, especially chronic recurrent culture-negative
Soft-tissue, inflammatory, and granulomatous processes
Hodgkin lymphoma with eosinophils
Foci of CD1a + DC hyperplasia can occur in lymph nodes containing lymphoma, simulating LCH, but extranodal disease never seen in these instances
Bone lesions may be sharply demarcated or gray to yellow depending on lipid content. Lung and thymic lesions if rarely resected may be cystic, but there are virtually no early or late changes in any of the organs that will be suggestive of LCH.
Diagnosis of LCH is made by identifying the lesional cells as LCH cells with the Langerhans cell phenotype and the correct pattern of tissue involvement. The cells are moderately large (20 to 25 µm), oval, and not dendritic in shape, with a grooved (coffee bean) or complex folded nuclear profile. Nuclei are commonly single or two to three per cell, and there are osteoclast-type multinucleated cells in many sites that may harbor the folded LCH nucleus ( Fig. 19.12 ), which are phenotypically different from the intermixed CD1a−/Langerin-/CD68 + osteoclast giant cells resorbing the bone. Cytoplasm is abundant and pale and may have few fine granules. The gold standard for the diagnosis remains the phenotypic confirmation of cells that look cytologically appropriate for LCH. In some sites, most notably bone, eosinophils can be interspersed often in large numbers, albeit unevenly. Although the lesions were formerly called eosinophilic granulomas, eosinophils are not required for the diagnosis, which rests on identifying the LCH cell. In bone and soft tissue, early lesions contain sheets of LCH cells with inflammatory cells, mostly T cells largely restricted to the periphery. Phagocytic macrophages and osteoclast-type giant cells can be present or even dominate the picture. The LCH cells disappear as lesions regress, and it is possible to biopsy a late, healing lesion from which the LCH cells have disappeared and therefore be unable to confirm the diagnosis. In the skin, the process is usually epidermotropic, hugging the epidermis and filling the papillary dermis. The cells are characteristically large and oval. Lymph node involvement is strictly sinus in distribution at first, spilling over into the paracortex and thus following the normal migratory pathway for sentinel DCs reaching the lymph node ( Fig. 19.13 ). Liver involvement is almost uniquely biliary, with LCH cells in the epithelium of large-caliber bile ducts where they cause a sclerosing cholangitis ( Fig. 19.14 ). The cells can migrate peripherally within the biliary tree, usually between the basement membrane and the biliary epithelial cells. When there is extensive involvement, parenchymal nodules can occur. Lung involvement is peribronchial with extension of LCH cells into the peribronchial alveolar walls. Bone marrow involvement is often subtle; first because there may only be a few LCH cells in small clusters and second because there may be a reactive macrophage histiocytosis that obscures the picture. In these cases demonstration of BRAF -mutated cells by PCR studies (e.g., aspirate) would further support the diagnosis, especially if the patient has a known BRAF -mutated lesion. Thymic involvement, although rare, shows discrete morphologic patterns including solitary and/or cystic LCH lesions with thymic gland disruption; more variable medullary or diffuse thymic disease in multisystemic LCH; and rarely, a mixed histiocytic lesion with a concurrent LCH and juvenile xanthogranuloma-like proliferations. Posterior pituitary, leptomeningeal, and choroid involvement has the characteristic large, oval LCH cells. CNS-LCH lesions have been described based on small case series that include (1) circumscribed granulomatous lesions including variable numbers of CD1a + cells with a robust surrounding xanthogranulomatous or macrophage-rich inflammatory response (most commonly in circumventricular sites, leptomeninges, and the pituitary); (2) CD1a + cell–rich lesions in infundibular sites extending to the hypothalamus and adjacent CNS parenchyma with variable inflammation and parenchymal injury including demyelination and gliosis; and (3) CD1a cell poor wherein lesions, most often in the cerebellum, are associated with ill-defined leukoencephalopathy and neurodegenerative features with a fibrotic/gliotic response. Few neurodegenerative (ND-LCH) cases have undergone pathologic investigation. Current work is in progress to better delineate the pathophysiology of late ND-LCH disease. Splenic involvement can also be difficult to define except in those patients who form LCH nodules in the spleen. Gastrointestinal involvement fills the lamina propria and can be subtle until one recognizes that the usual cell mix is replaced by a single population of LCH cells. Thyroid and thymus can be involved with sheets of oval LCH cells that have the diagnostic phenotype.
The cells of LCH have a stable phenotype: CD1a + , Langerin + , S100 + , and vimentin + . They have small, paranuclear intracytoplasmic accumulations of CD68 and HLA-DR. In sites such as the paracortex of lymph nodes, the LCH cells can show some evidence of limited maturation in which CD1a is focally lost and membrane HLA-DR expression is increased. The VE1 antibody will show a dark, cytoplasmic granular staining pattern in the clonal histiocytes with the BRAF -V600E mutation (see Fig. 19.17B ).
In the past (early 2000s), loss of heterozygosity (LOH) of possible tumor-suppressor genes (e.g., 9p and 22q adult/pulmonary and 1p and 7 in children/bone) was the only described molecular aberration in LCH. However, since the last edition, a molecular transformation has commenced in LCH starting with the identification of BRAF -V600E in 2010. Since that time, advances in genetic techniques have led to new findings and better understanding of the molecular changes in LCH. Frequency of BRAF -V600E mutations has been described in 50% to 64% of adult and pediatric patients with LCH in several studies, highly dependent of the method of testing. Highly sensitive qPCR methods are recommended in order to detect small allelic fractions (e.g., down to ~1% mutant), which are common in LCH, as cells are often diluted with a mixed inflammatory milieu. Activated BRAF stimulates constitutive transcription and cellular proliferation through the RAS/RAF/MEK/ERK pathway. Case reports of BRAF -V600E wild-type LCH have also shown MAP2K1 (10% to 20%) and fewer reports of ERBB3 , RAS , ARAF, and MAP3K1 activating mutations that also work on the same signaling pathway with universal ERK pathway activation (see Fig. 19.10 ). New reports of BRAF exon 12 in-frame deletions and FAM73A-BRAF gene fusions have been issued. Also point mutations in the phosphoinositide 3-kinases-Akt murine thymoma ( PIK3CA-AKT ) pathway has also been implicated in LCH in rare cases. In particular, BRAF mutations have clinical significance as they are therapeutic targets and appear to drive the varied pathogenesis of LCH. A bone marrow–derived myeloid/monocyte precursor has been implicated in LCH based on mouse models and human data. Although the identification of BRAF -V600E mutations have been noted in both single lesion (e.g., low-risk) and multisystem (e.g., high-risk) LCH, the landmark study by showed for the first time that MS-LCH also harbored the BRAF -V600E mutation in CD34 + bone marrow hematopoietic precursors and CD11c + /CD14 + circulating blood myeloid/monocyte fractions, unlike low-risk LCH. The human data was further supported by mouse models showing differential LCH-like ‘high risk’ and ‘low risk’ phenotypes dependent on the enforced expression of BRAF -V600E, driven either by CD11c-Cre (high-risk phenotype) or Langerin-Cre promotors (low-risk phenotype). These findings have provided initial data for a model of LCH pathogenesis. The timing of MAPK/ERK activation in myelopoiesis may help explain the clinically distinct LCH risk groups (see Fig. 19.10 ) and support a new classification of LCH as a myeloproliferative neoplasm or rather an inflammatory myeloid neoplasia , given the likely contribution of the inflammatory milieu to its pathogenesis. For the first time, a recent large French registry study has correlated BRAF -V600E mutations with high-risk LCH and poor short-term response to first-line chemotherapy. Although the Histiocyte Society has not officially recommended upfront molecular testing for all LCH, still under the prevue of “experimental” testing, data are accumulating in support of mutational testing. Furthermore, if BRAF -V600E positive, reflex qPCR BRAF -V600E on blood or bone marrow is also recommended to help follow disease course when applicable.
The Birbeck granule was originally identified in Langerhans cells of the skin, then in adult pulmonary LCH, and later in cutaneous and bone lesions of childhood. The structure forms where the C-type lectin Langerin (CD207) accumulates and the Birbeck granule appears to be related to endosomal trafficking. The use of Langerin immunostain has largely supplanted the need for ultrastructural analysis. The Birbeck granule is a rod-shaped bilaminar disk with an internal zipper-like pattern of striations, often with a bulbous dilatation at one end, like a tennis racquet. The granules are found wherever Langerhans cells are seen in the body but are not unique to Langerhans cells and have been described in other cells ( Fig. 19.15 ). LCH cells have few cytoplasmic organelles and filopodia at the cell surface.
Lesions that are aspirated are commonly mixed in nature, and the LCH cell must be identified. Presumptive diagnosis rests on the cytologic features of large, bland, oval histiocytes that are nonphagocytic and have pale cytoplasm and folded nuclei. Immunocytology for CD1a or Langerin is definitive but must be interpreted in the correct clinicopathology setting ( Fig. 19.16 ). S100 is soluble and does not survive without fixation. Pulmonary involvement can be inferred when the imaging is compatible and more than 12% of the larger bronchoalveolar cells are CD1a + or display Langerin.
The differential diagnosis depends on the site of involvement. Skin lesions of mastocytosis (urticaria pigmentosa), early type juvenile xanthogranuloma, and melanocytic nevi (S100 + ) can simulate LCH. An increase in perivascular CD1a + Langerhans cells that are spindled or dendritic in shape is a feature of many chronic dermatoses, most notably chronic scabies. The thymus and thyroid can contain microscopic collections of hyperplastic LCH-like cells in thymectomies and thyroidectomies of patients without LCH disease. Skin infiltrates with follicular hyperplasia (i.e., cutaneous pseudolymphoma) may also harbor small hyperplastic LCH-like collections without LCH disease.
Applying immunohistochemistry to a reactive lymph node with dermatopathic-like changes will reveal an increased number of dendritic cells in the paracortex that can form large, confluent, and nodular aggregates of T cells and DCs, including Langerhans cells. Because the DCs are maturing or fully mature, they will have the phenotype of interdigitating DCs (S100 + , fascin-hi) with interspersed Langerhans cells (CD1a + , Langerin + , S100 + ) in the paracortex. In some cases the Langerhans cells may be exuberant and cluster but should not be diagnosed as LCH, which is a sinus disease (see Fig. 19.3 ). A diffuse paracortical DC hyperplasia in tonsils or lymph nodes can be an accompaniment of other infiltrating processes and can be misleading by diverting attention from the infiltrate, usually a leukemia or lymphoproliferative disorder.
Bone lesions are most likely to be confused with osteomyelitis, especially chronic recurrent multifocal culture-negative disease. LCH has few plasma cells in between the lesional LCH cells but may have more in the surrounding inflammatory reaction. Hodgkin lymphoma can contain eosinophils and scattered CD1a + cells, and Rosai-Dorfman disease can also produce bone lesions. In late and healing lesions where LCH cells are few, fibrotic or fibrohistiocytic lesions enter the differential diagnosis. In these cases, demonstration of mutant BRAF can support LCH involvement, but its absence cannot exclude it.
Lung lesions that are fibrotic may have few LCH cells and can resemble fibrosing interstitial diseases with honeycombing in the most severe instances. Alveolar macrophages can simulate LCH cells but are not CD1a + ; however, smokers may have increased numbers of LCs without LCH. Spontaneous pneumothorax can be associated with an eosinophilic pleural reaction.
Bone marrow involvement by LCH is obscured by the common finding of a marrow macrophage histiocytosis. Only CD1a + /Langerin + large cells can be considered informative, but PCR (on aspirates) for BRAF -V600E has identified a mutation in the absence of CD1/Langerin staining, and this more sensitive molecular technique should supersede immunohistochemistry in such instances. S100 can be confounding in the bone marrow by staining stromal cells, fat cells, and some T cells, and activated macrophages.
Liver involvement is simulated by macrophage activation in children with widespread disease who may have hepatomegaly and hypoalbuminemia. An identifiable peribiliary infiltrate, CD1a + /Langerin + , is diagnostic, but biliary obstructive changes on a biopsy specimen are presumptive evidence of LCH involvement downstream and may not be demonstrated in a biopsy. Juvenile xanthogranuloma lesions also have a portal distribution but have a different phenotype and spare the bile ducts.
Pituitary lesions must be distinguished from germ cell tumors. The late demyelinating CNS lesions do not have a typical LCH cell presence. Choroid plexus and dural lesions of LCH acquire large numbers of xanthoma cells and can simulate reactive histiocytosis or even juvenile xanthogranuloma.
The prognosis depends on the site and extent of disease, with the high-risk organ sites having a poorer outcome. Diabetes insipidus (DI) is often permanent. The greater the number of organ systems involved in small children, the worse the outcome, with the highest mortality for those with high-risk organ involvement (e.g., bone marrow, liver, spleen) and for those who have disease progression during the first 12 weeks of therapy. High risk of long-term CNS effects (i.e., DI, endocrinopathies, ND-LCH) is associated with certain “CNS-risk sites” including the temporal bone, maxillofacial bones, and orbital bone involvement. Poor initial response to chemotherapy in the systemic form also portends a poor outcome and was the impetus for therapy intensification in LCH III protocol after 6 weeks.
Adults with smoking-related lung disease have variable outcomes; about half have disease regression with smoking cessation and with steroids, but others require chemotherapy or progress. Monoostotic bone lesions often regress no matter what modality is used but can leave impairment at the site (e.g., vertebral collapse). From a management standpoint, the pathologist should realize that excising an LCH lesion with “negative” margins is not required, and attempts to do so may result in more extensive surgery than necessary. Indomethacin or intralesional steroids are used with polyostotic bone lesions and low-dose radiation or chemotherapy with vinblastine-prednisone have been used. In the LCH-III study 66% of patients with risk organ involvement and 86% of patients without risk organ involvement demonstrated a response to corticosteroids and vinblastine by 6 weeks. Therapy intensification (addition of etoposide or mercaptopurine) was added for nonresponders and high-risk MS-LCH in the initial 12 weeks with improved outcomes. Second-line, salvage therapy includes immune suppression with cyclosporine, low-dose cladribine, intermediate-dose cytosine-arabinoside (cytarabine), or clofarabine (purine analogue used in acute myeloid leukemia), which has been used with variable success. Persistent high-risk MS LCH despite various salvage treatments may be cured with bone marrow hematopoietic stem cell transplantation. Lung or liver transplantation in some instances has been used in refractory disease. For the first time, BRAF -V600E has also been suggested as a prognostic marker, correlating with high-risk LCH and increased resistance to first-line therapy in a French LCH cohort. Targeted therapy with MAPK/BRAF inhibitors has been under active investigation, with case reports highlighting improved outcomes in refractory disease; however, larger series with longer follow-up are needed to assess efficacy and toxicity.
Note that transition from bona fide LCH to later Langerhans cell sarcoma (dendritic sarcoma, Langerhans' phenotype) is exceptionally rare, and there is reason to believe that the LCH in these instances was cytogenetically different to begin with (see Langerhans Cell Sarcoma ).
Like the juvenile xanthogranuloma family, adult xanthogranuloma can vary from single lesions to multiple lesions, xanthoma disseminatum, and “systemic-form” Erdheim-Chester disease (ECD). Localized lesions are common around the orbit, where a number of clinical variants are recognized: adult-onset xanthogranuloma, necrobiotic xanthogranuloma, adult-onset asthma, periocular xanthogranuloma, and the ocular involvement of Erdheim-Chester disease. Historically, ECD has been grouped under the category of the systemic xanthogranuloma (XG) family given its shared immunophenotype, but it has been long recognized that it has a distinct clinical and radiographic presentation. Accumulation of data over the past few years has shifted our understanding of ECD toward a clonal inflammatory myeloid neoplasm, similar to LCH. ECD is now recognized as a distinct entity in the updated 2016/2017 WHO tumors of hematopoeitic and lymphoid tissues. An increasing number of reports link LCH and ECD, either within the same lesion/site or at different sites within the same patient during his or her lifetime, while also sharing clonal mutations in the mitogen-activated protein kinase (MAPK) pathway. These are the drivers that led to the recent proposal in classifying LCH and ECD together within the “L” (Langerhans) group of the revised classification of histiocytoses and neoplasms of the macrophage-dendritic cell lineages (see Box 19.1 ). For the purposes of this chapter we opt to leave ECD as its own category, acknowledging the shared molecular phenotype with LCH and shared immunophenotype with the XG family of lesions.
ECD is characterized by symmetric bony sclerosis on x-ray and 99T bone scintigraphy and organ involvement of the CNS, orbit, lung, retroperitoneum, kidney, heart, and skin. The biology is chronic and often aggressive, depending on involvement site. “High-risk” disease is seen with severe involvement of any one organ. ECD has a predilection for the hypothalamic-pituitary axis with the development of diabetes insipidus, much like Langerhans cell histiocytosis, and in a number of instances it has been associated with LCH at various sites.
Solitary and localized forms have no distinctive imaging features, but ECD is, in part, a radiologic diagnosis characterized by polyostotic sclerosis. This disease involves the bones around the knee, less frequently the elbow. The diaphyses and metaphyses have a coarsely increased trabecular pattern of medullary sclerosis and cortical thickening on radiography (Paget's disease–like pattern). Bone scintigraphy is said to show pathognomonic bilateral and symmetric increased uptake affecting both diaphysis and metaphysis of the femur and tibias. Computed tomography demonstrates increased density, and MRI highlights low-signal marrow replacement on fat-suppressed T1-weighted images, mixed signals on T2, and some enhancement after gadolinium contrast. FDG-PET has been used to identify new lesions and to monitor response to therapy.
Clinicoradiographic findings in the context of a histiocytic proliferation of epithelioid to xanthomatous cells with an XG phenotype; now considered as an inflammatory myeloid neoplasm
True incidence unknown, 500 to 550 case reports
Bone, retroperitoneum, including perinephric and periaortic; lung, including pleural, pericardial, skin, orbit; and CNS, including frequent posterior pituitary involvement
Chronic disease with remissions and progression dependent on sites of involvement
High morbidity and mortality depending on extent of organ involvement
Adults 40 to 80 years of age, with male predominance; few pediatric cases reported
Constitutional: fevers, fatigue, night sweats; CNS manifestations, including ataxia, diabetes insipidus, endocrinopathy, double vision, exophthalmos, and psychiatric features; dyspnea; bone pain; xanthelasma and rash
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here