Chapter 28 Mechanisms of Skin Damage
Understanding the mechanisms of skin damage in cutaneous lupus erythematosus (CLE) requires knowledge of the current categorization of cutaneous LE types, the histopathology and the immunopathology of the specific cutaneous manifestations of LE that occur in patients with this autoimmune disease, insight into the molecular genetics that predisposes individuals to the disease, elucidation of the cellular and molecular biologic abnormalities that underlie the increased ultraviolet light (UV) photosensitivity, and insight into the immune pathomechanisms that orchestrate an abnormal immune response against the skin.
Cutaneous histopathology can provide a very useful clue for the diagnosis of CLE. However, it should be noted that correct classification of LE into the four major categories—systemic LE (SLE), subacute CLE (SLCE), chronic CLE—largely relies on the clinical picture and key laboratory findings. Variants of the aforementioned specific CLE manifestations include discoid LE (DLE), LE profundus (lupus panniculitis), bullous LE, hypertrophic LE and chilblain LE, all of which, like SCLE and CCLE, may present as isolated skin disorders or which may represent cutaneous manifestations of SLE. It must be emphasized that the histopathology of all of these forms depends on the age of the lesion and previous therapy. Accordingly, skin biopsy specimens from early lesions may yield nonspecific signs of inflammation and topical or systemic corticosteroids can easily attenuate the true histopathologic picture. In this chapter we will outline the dermatopathologic key findings of the specific cutaneous LE lesions followed by a summary of the major pathogenetic concepts that have been developed thus far to explain skin damage in patients with LE.
Epidermal changes such as hydropic (vacuolar) degeneration of the basal layer, scattered “Civatte bodies” (dead keratinocytes), epidermal atrophy, and compact orthohyperkeratosis are key features in SLE (i.e., in ACLE, also known as “malar rash”), SCLE, and DLE lesions1–3 (Fig. 28.1). These changes are identified by routine hematoxylin and eosin (H & E) staining in formalin-fixed biopsy specimens. Neonatal LE displays similar epidermal findings as compared to SCLE, and is sometimes more pronounced with cleft formation between the dermis and epidermis. Prominent vacuolar epidermal degeneration together with massive accumulation of inflammatory cells in the basal membrane zone (“interface dermatitis”) can progress to subepidermal blister formations known as bullous LE.1 The epidermal blister roof of patients with bullous LE is mostly intact in contrast to toxic epidermal necrolysis-like ACLE, an extremely rare and only recently described subtype of LE characterized by a pan-necrotic epidermis.1 Vacuolar degeneration and keratinocyte injury of the basal layer may be regarded as early pathogenetic steps in the development of CLE. Using in situ nick translation and TUNEL it was demonstrated that apoptotic nuclei accumulate in skin of patients with CLE after ultraviolet (UV) light exposure supporting the pathogenetic concept of increased apoptosis in skin of patients with CLE.1 As a consequence, epidermal atrophy and possibly reactive epidermal orthohyperkeratosis may develop. Epidermal hyperkeratosis is very prominent and typically includes the adnexal structures (“follicular plugging”) in DLE1,2 (Fig. 28.2). In hypertrophic LE, epidermal hyperkeratosis is even more increased with parakeratosis and acanthosis. Another pathologic finding of the dermoepidermal junction, especially in long-standing CLE lesions, is thickening of the basement membrane.1,2 It is most apparent in DLE, less in SCLE, and often absent in ACLE lesions, and can be visualized by the periodic acid-Schiff (PAS) stain (Fig. 28.3). Hydropic degeneration of the basal membrane zone, apoptotic keratinocytes, and atrophy are also found in chilblain LE.1 In rare cases of verrucous chilblain LE, epidermal hyperkeratosis, patchy parakeratosis, acanthosis, and hypergranulosis are the dominant features.1 Hydropic degeneration of the basement membrane zone, epidermal atrophy, follicular plugging, and basement membrane thickening occur in the majority of patients with LE panniculitis (LE profundus),9,10 while such changes are only occasionally detected in patients with LE tumidus.11,12
Fig. 28.1 Histopathology of acute cutaneous lupus erythematosus (classical malar rash) in a patient with systemic lupus erythematosus. Note hydropic degeneration of the basal layer, apoptotic keratinocytes (Civatte bodies), epidermal atrophy, and compact orthohyperkeratosis. In addition, inflammatory cells, mainly lymphocytes and few neutrophils are present in a patchy distribution in the papillary dermis (hematoxylin and eosin stain).
Fig. 28.2 Classical discoid lupus erythematosus. Note the prominent follicular plugging. The infiltrate consisting mainly of lymphocytes is accentuated around the hair follicles and blood vessels. There is also dense mucin deposition in the upper dermis (hematoxylin and eosin stain).
Another consistent finding in all CLE forms is the presence of an inflammatory infiltrate consisting mostly of lymphocytes. The cutaneous location and pattern of the inflammatory infiltrate detected by H & E staining differs in its location depending on the category of LE. The lymphocyte infiltrate in cutaneous lesions of patients with SLE (malar rash) can be sparse especially in early lesions and is typically located in the upper dermis around the blood vessels1 (Fig. 28-1). In more advanced lesions it becomes more prominent, involving the dermoepidermal junction (interface dermatitis), sometimes with extravasation of erythrocytes, and deposition of fibrinoid material around blood vessels and between collagen fibers. In some biopsy specimens of ACLE from patients with SLE, there are also signs of leucocytoclastic vasculitis, that is, nuclear dust, fibrinoid necrosis of the vessel wall, neutrophil infiltration, and extravasation of erythrocytes. In bullous LE of patients with SLE, there is in addition a prominent mixed mononuclear/neutrophilic infiltrate along with dermal microabscesses.1 The blister fluid contains fibrin and neutrophils. SCLE lesions and neonatal LE share similar patterns of the inflammatory cell infiltration. The lymphocytes are mostly confined to the upper dermis leading to a band-like infiltrate with interface dermatitis1–3 (Fig. 28.4). Erythrocyte extravasation and dermal fibrin deposition can occur. A striking feature of DLE distinguishing all other CLE forms is the prominent periadnexal inflammatory infiltrate1,2 (Fig. 28-2). The epidermal changes of DLE and LE hypertrophicus as outlined above (follicular plugging due to hyperkeratosis) may represent a follicular response to proinflammatory and proliferative signals released by infiltrating lymphocytes. Besides its striking periadnexal location, the inflammatory infiltrate in classical CLE lesions displays a patchy, sometimes a band-like (lichenoid), pattern. While the inflammatory infiltrate of chilblain LE is likewise situated in the upper dermis, around the blood vesicles, and occasionally around the hair follicles (especially in the verrucous subtype),1 the infiltrates in LE tumidus and LE panniculitis are mainly present in deeper layers of the skin. In LE tumidus, perivascularly situated lymphocytes are found in the superficial and deep dermis and only infrequently around the skin adnexal structures.11,12 In LE panniculitis, lymphocytic infiltration is present, sometimes together with eosinophils, of the subcutaneous fat leading to panniculitis, fat necrosis, and hyalinization of adipose lobules.9,10 The pattern of the panniculitis is lobular, and sometimes paraseptal. Lymphoid follicles and germinal centers are often detected. Periadnexal infiltrates are less frequently seen.
Fig. 28.4 Histopathological changes in subacute cutaneous lupus erythematosus. Note less dense and prominent inflammatory infiltrates as compared to Fig. 28.2. Inflammatory cells, mainly lymphocytes, are primarily found in the upper dermis close to the dermo-epidermal junction (interface dermatitis) as well as perivascularly (insert). There is hydropic degeneration of the basal layer (hematoxylin and eosin stain).
Another consistent feature of virtually all specific CLE lesions is dermal mucin deposition.1 It can be visualized by colloidal iron stain or Alcian blue stain (Fig. 28.5). Mucin deposits are most prominent in LE tumidus and may give rise, when excessively prominent, to so-called papular mucinosis. The biochemical nature of the deposited mucopolysaccharides in CLE (as well as in other inflammatory skin disorders) is undefined. Proinflammatory cytokines such as interleukin-1 (IL-1) released by inflammatory cells may be involved in inducing increased mucopolysaccharid synthesis by dermal fibroblasts, but the exact pathogenesis remains unknown. However, mucin deposition alone is not a specific dermatopathologic finding and can frequently be detected in many inflammatory and noninflammatory conditions of the skin.
Early immunodermatologic work on cutaneous lesions of patients with SLE strongly suggested a pathogenetic role of precipitated immunoglobulins at the dermoepidermal junction in this autoimmune disorder. Due to its characteristic band-like staining pattern this phenomenon in the skin of patients with LE has been coined “lupus band.” The technique now routinely performed to detect immunoglobulins, fibrin, and complement components in lesional and nonlesional skin specimens of patients with LE is called direct immunofluorescence (DIF). It is most reliably performed on snap-frozen skin specimens. The intensity of fluorescence in skin biopsy specimens (lupus band test [LBT]) depends on the biopsy site, the acuity of a lesion, and previous treatment. Facial lesions may give false positive results whereas very early ones and those pretreated with topical corticosteroids and immunomodulators or systemic medication may yield false negative results. Moreover, immune complexes and complement along the dermoepidermal junction can be detected in a number of other inflammatory skin disorders. Therefore, a positive lupus band test must be interpreted in the context of the clinical picture and laboratory data of the patient. Although the overall diagnostic relevance of DIF analysis has declined during the last several years and proper clinical characterization of CLE lesions, serologic tests, and routine histopathology may be sufficient for establishing the correct diagnosis, the LBT in nonlesional skin has still a high predictive value for the diagnosis of SLE. Moreover, DIF studies may be helpful in discriminating inflammatory skin disorders with similar histopathologic pictures as LE. Finally, DIF studies can provide some important information on the pathogenesis of CLE. In general, most DIF studies have been undertaken in patients with SLE, DLE, and SCLE while comparatively less information is available regarding the LBT in lesional skin of the other CLE subsets. At least three patterns of DIF in skin of patients with LE can be distinguished.1
The most striking immunopathologic feature in CLE (and SLE) is the presence of deposited immunoglobulins (IgG, IgM, and IgA), complement (especially C3), and other serum proteins (e.g., fibrin, albumin, factor B, and properdin) at the dermoepidermal junction. Ultrastructural studies using immune electron microscopy have shown that the immune deposition takes place in the sub–lamina densa region. Several morphologic variants of the immune deposits at the dermoepidermal junction have been described including linear, granular, or shaggy. In addition globular deposits consisting of immunoglobulins, complement, and fibrin can frequently be detected in lesional skin (and nonlesional skin of patients with SLE). These ovoid (cytoid) bodies are scattered along the dermoepidermal junction but can also be found in the superficial dermis. Numerous studies on the immune deposition in patients with the three major LE forms with cutaneous involvement have resulted in a typical distribution of the LBT positivity in lesional and nonlesional skin (Table 28.1).
|LE Subtype||Lesional Skin||Nonlesional Skin|
DLE, discoid lupus erythematosus; SCLE, systemic cutaneous lupus erythematosus; SLE, systemic lupus erythematosus.
In cutaneous lesions of patients with SLE, the lupus band test is positive in 90 to 100%. IgG, IgM, IgA, C3, and fibrin are most often detected. Most importantly, the LBT is positive in 50 to 90% in nonlesional, sun-protected skin of patients with SLE. In lesional skin of DLE, immune deposits are present in about 60 to 95%. IgG3 and C3 are most frequently found and typically display a linear band-like pattern, and sometimes also a granular fluorescence, along the dermoepidermal junction (Fig. 28.6). The LBT is usually negative in nonlesional skin of patients with DLE, although in some cases deposits of C3 and IgM have been described. Lesional skin from patients with SCLE displays a similar pattern to that of DLE, and the composition of immunoglobulins and complement at the basement membrane zone is similar to that of DLE. The LBT is positive in 60 to 100% of skin biopsy specimens taken from lesional skin of patients with SCLE, and is consistently negative in nonlesional skin. In chilblain LE and LE panniculitis, the majority of patients have immune deposits (mostly IgM and/or IgG and/or C3) at the dermoepidermal junction in lesional skin,7,9,10 while in LE tumidus LBT positivity is rather heterogeneous.11,12
Fig. 28.6 Positive lupus band test in lesional skin from a patient with discoid lupus erythematosus. IgG deposits are visualized by an anti-human IgG antibody coupled to the fluorochrome FITC. A fluorescent bright green band is seen at the dermoepidermal junction. Note “nonspecific” immunostaining on collagen fibers in the dermis.
The precise pathomechanism of immune deposition at the dermoepidermal junction in the skin of patients with LE remains only partially understood. A pathogenetic role for deposited immunoglobulins has recently been emphasized in a fraction of patients with bullous SLE. These patients have circulating anti–basement membrane zone antibodies (mostly IgG, less frequently IgA) directed against type VII collagen.14,15 Using the salt-split skin technique, moreover, autoantibodies directed against several undefined proteins of 230, 200, 180, 130, and 97 kD from epidermal extracts, and 75 kD from dermal extracts, were identified.1 It has long been known that the fluorescence intensity of the LBT in SLE correlates with disease activity,1 as well as with the serum titer of antinuclear antibodies, suggesting a causal relationship.1 Both native and single-stranded DNA antibodies have an affinity for collagen present in the basement membrane, possibly leading to in vivo fixation of anti-DNA antibodies.1 As will be outlined below, autoantigens such as SSA/Ro are exposed on the surface of epidermal keratinocytes upon ultraviolet (UV) radiation. Although this phenomenon would not fully explain the band-like pattern of deposited immunoglobulins at the dermoepidermal junction in LE, it may suggest a contribution to its pathogenesis via in vivo fixation of antinuclear antibodies (ANAs) to exposed epidermal epitopes.
In addition to the lupus band, epidermal immunofluorescence occurring as a cytoplasmic and/or nuclear fluorescence has been described.1 A nuclear speckled immunofluorescence pattern can be detected in epidermal cells of nonlesional skin in patients with SLE. The presence of this DIF pattern in nonlesional skin correlates with the titer of anti-RNP antibody, suggesting a causal relationship and in vivo fixation. These deposits contain mostly IgG, less frequently IgM and IgA. However, similar DIF patterns have been seen in patients with other autoimmune disorders such as Sjögren’s syndrome or mixed connective tissue diseases. In lesional skin of SCLE there is another epidermal DIF pattern.1 Accordingly, a fine granular, dust-like deposition of IgG in the cytoplasm, nuclei, and intercellular space of the basal epidermis is detectable. This DIF pattern has been associated with the presence of circulating anti-Ro/SSA antibodies. It is also present in skin of patients with Sjögren’s syndrome and neonatal LE. In vivo fixation of anti-Ro/SSA antibodies in the pathogenesis of this phenomenon is supported by the fact that injection of anti-Ro/SSA antibodies into nude mice leads to analogous epidermal immunostaining in grafted human skin.1