* Interface dermatitis can be classified based upon the cell type
that dominates the infiltrate (ie, neutrophilic, lymphocytic, or
lymphohistiocytic) or by the intensity of the interface inflammation.
Regarding lymphocytic interface dermatitis, there are 2 broad
categories: cell-poor interface dermatitis, when only a sparse
infiltrate of inflammatory cells is present along the dermoepidermal
junction, or cell rich, which typically occurs as a heavy bandlike
infiltrate that obscures the basal layers of the epidermis. In the case
of lymphocytic interface dermatitis, the latter is often termed a
lichenoid interface dermatitis. This review focuses upon the mononuclear
cell-predominant forms of interface dermatitis.
(Arch Pathol Lab Med. 2008;132:652-666)
(Development and progression)
Systemic lupus erythematosus (Development and progression)
Skin (Development and progression)
Crowson, A. Neil
Magro, Cynthia M.
Mihm, Martin C., Jr.
|Publication:||Name: Archives of Pathology & Laboratory Medicine Publisher: College of American Pathologists Audience: Academic; Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2008 College of American Pathologists ISSN: 1543-2165|
|Issue:||Date: April, 2008 Source Volume: 132 Source Issue: 4|
The term interface dermatitis refers to the finding in a skin
biopsy of an inflammatory infiltrate that abuts or obscures the
dermoepidermal junction (DEJ). Historically, interface dermatitis has
been classified based in part upon the cell type that dominates the
infiltrate (ie, neutrophilic, lymphocytic, or lymphohistiocytic).
Interface dermatitis can also be classified by the intensity of the
interface inflammation; regarding lymphocytic interface dermatitis,
there are 2 broad categories in our classification scheme. These include
cell-poor interface dermatitis (Table 1), when only a sparse infiltrate
of inflammatory cells is present along the DEJ, or cell rich (Table 2).
The infiltrate in lesions of cell-rich lymphocytic interface dermatitis
typically occurs as a heavy bandlike process that obscures the basal
layers of the epidermis; this is often termed a lichenoid interface
CELL-POOR INTERFACE DERMATITIS: THE DERMATOPATHOLOGY OF CONNECTIVE TISSUE DISEASE AND ITS MIMICS
Cell-poor vacuolopathic interface dermatitis is defined by basilar keratinocyte and subepithelial vacuolopathy un accompanied by a significant inflammatory infiltrate. Usually, there is some lymphocyte tagging along the DEJ pointing to the immunopathogenetic basis, namely, cellular cytotoxicity, either in the context of a type 4 immune reaction or of antibody-dependent cellular cytotoxicity, or a type 2 immune reaction, when autoantibodies target components of the basement membrane zone to trigger an antibody-dependent cellular cytotoxicity reaction.
The prototypic differential diagnoses of a cell-poor interface dermatitis include erythema multiforme; autoimmune connective tissue disease, particularly systemic lupus erythematosus, dermatomyositis, and mixed connective tissue disease; graft-versus-host disease (GVHD); morbiliform viral exanthem; and morbiliform drug reaction (Table 1).
There are a variety of more recently described entities that also manifest a lymphocytic interface inflammatory infiltrate, such as the human immunodeficiency virus (HIV)-related dermatoses, the cutaneous eruption of lymphocyte recovery, and the superantigen id reaction. Cutaneous T-cell dyscrasias and cutaneous T-cell lymphoma can be associated with a combination of interface inflammation and basal layer epidermal colonization, and thus they enter into the differential diagnosis of interface dermatitis. (1,2) Specifically, so-called interface parapsoriasis (or cutaneous lymphoid dyscrasia of the hypopigmented interface type), pityriasis lichenoides, (3) and patch or early plaque-stage lesions of mycosis fungoides may manifest this histomorphology. (2,3) Furthermore, there are additional forms of autoimmune connective tissue disease (CTD) that have been associated with an interface dermatitis, including Sjogren syndrome, vitiligo, (4,5) and autoimmune thryoiditis. (6)
Erythema multiforme is a distinctive clinical pathologic entity with a wide variety of underlying causes. The classic lesion has a targetoid morphology with a peripheral rim of erythema and a central zone of pallor. Some lesions manifest a dusky or violaceous appearance with no true central clearing. Blisters may be observed. As the pathogenetic basis of erythema multiforme is one of cellular cytotoxicity, the sites of predilection are those where antigenic processing is maximal, which includes the palms and soles, but lesions may occur elsewhere and may become widespread.
Common to cases of drug- or infectious-based etiology are focal areas of basilar vacuolopathy accompanied by lymphocyte tagging along the DEJ; suprabasilar lymphocytosis around degenerating keratinocytes also may be seen (Figure 1). In those cases mediated by infection, one typically observes a fairly brisk angiocentric superficial and deep lymphocytic infiltrate, along with a cell-poor interface dermatitis with minimal epidermal injury. In contrast, cases of drug-based etiology may show a less intense dermal-based inflammatory cell infiltrate, often with tissue eosinophilia but with more pronounced degenerative epithelial changes, including discrete zones of confluent epidermal necrosis (Figure 2). Focal areas of basilar vacuolopathy accompanied by lymphocyte tagging along the DEJ are the hallmarks; suprabasilar lymphocytosis around degenerating keratinocytes also may be seen. In drug-based erythema multiforme, acrosyringeal accentuation of these interface inflammatory and degenerative epithelial changes is typical. There may be streak dyskeratosis, whereby basilar keratinocytes acquire an elongated cigar-shaped morphology with hypereosinophilic condensed cytoplasms and pyknotic elongated nuclei. As most clinicians biopsy acute lesions, the epidermis is typically of normal thickness and is surmounted by a basket weave pattern of orthokeratosis. Lesions of active herpes not infrequently show areas of interface dermatitis resembling erythema multiforme; characteristic viral cytopathic changes of herpes, however, may be observed.
The main consideration is that of acute autoimmune CTD, typically as seen in the setting of systemic CTD syndromes. In our experience, in lesions of systemic lupus erythematosus (SLE) (Figure 3) or dermatomyositis (Figure 4), the presence of striking dermal mucin deposition is a helpful distinguishing feature. As well, the degree of epithelial injury in such cases is usually less than that observed in erythema multiforme. Acrosyringeal accentuation, as seen in drug-associated erythema multiforme, is not present in lesions of acute systemic collagen vascular disease. Although tissue eosinophilia may be observed in drug-associated erythema multiforme, eosinophilis in lesions of CTD are uncommon outside of a few specific settings, such as drug-induced SLE and bullous SLE. In erythema multiforme there usually is no alteration of the stratum corneum, pointing to the transient and acute nature of the eruption. In contrast, in the interface dermatitis of CTD, hyperkeratosis and parakeratosis are frequent. Also, a function of the transient and acute nature of erythema multiforme is the unaltered architecture of the epidermis with preservation of the retia, in contrast to CTD, where one typically observes an atrophying interface dermatitis with retiform effacement. Furthermore, pauci-inflammatory mucin deposition is highly characteristic for lesions of collagen vascular disease.
The distinction of erythema multiforme from acute GVHD (Figure 5) may be difficult. There are, however, 3 useful morphologic clues: (1) in acute GVHD, the interface change is characteristically very focal, with accentuation at tips of rete ridges; (2) there is a dearth of tissue eosinophilia; and (3) follicular involvement is very common and at times can be the anatomic structure manifesting the brunt of interface change. A recent study demonstrated that intraepidermal bile pigment deposition can be seen in biopsies of GVHD specimens and correlates with hyperbilirubinemia and/or liver involvement. While this finding is only observed in a minority of cases of GVHD, it is a highly specific finding; it is not observed in erythema multiforme unless, of course, there is unrelated liver disease. We have never found this a useful morphologic clue.
Autoimmune progesterone dermatitis is temporally associated with the menstrual period. The manifestations are polymorphous and include urticaria, erythema multiforme, and eczema, although we have even encountered erythema nodosum-like lesions. Histologically, the cases show a variable
histomorphology dependent on the clinical presentation. Cases resembling erythema multiforme show a vacuolar interface dermatitis with varying degrees of keratinocyte necrosis. During each cycle, the eruptions may appear at previously affected sites, hence mimicking the clinical features of a fixed drug eruption. This rare phenomenon is attributed to an autoimmune reaction to female sex hormones. The condition can improve with tamoxifen, which suppresses ovulation and the postovulation rise in endogenous progesterone levels. In extreme cases, an oophorectomy is performed.
Erythema multiforme is a type 4 cellular cytotoxic reaction. Provocation of lesions is typically related to ingestion of drugs or to infection with viruses or microbial pathogens, such as Mycoplasma sp7 or herpes, including in the context of recurrent herpes labialis. Classical targetoid lesions are more common in patients whose erythema multiforme is triggered by herpes simplex. The disease is typically self-limited, but it may recur. Drug-induced forms of erythema multiforme can be quite severe. When the eruption is extensive, there may be severe oral mucosal involvement, which defines Stevens-Johnson syndrome (SJS). In SJS, more than 10% of the body surface is affected by epidermal necrosis and detachment, in a fashion reminiscent of, but less marked than, toxic epidermal necrolysis (TEN), in which more than 30% of the skin surface is so affected. (8) Case fatality rates are 1% to 5% in the former instance and 25% to 35% in the latter, where management in a dedicated burn unit is necessary. (9) Genetic factors appear to predispose patients to SJS, at least in certain populations. (10) A viral etiology is rarely implicated for these severe forms of erythema multiforme. There are rare reports of erythema multiforme following allergic contact dermatitis with nickel and poison ivy (ie, Rhus dermatitis). Eruptions mimicking TEN also have been described in the setting of drug-induced linear immunoglobulin (Ig) A disease, paraneoplastic pemphigus, and GVHD. In all of these aforesaid conditions where the clinical features are truly indistinguishable from true drug-induced TEN, the baseline morphology would capture features unique to each of the disorders, such as areas of neutrophilic interface in IgA disease, acantholysis in the setting of pemphigus, and striking follicular involvement in the setting of GVHD. The epidermal necrosis in erythema multiforme is mediated by cytotoxic T cells of the [T.sub.C]1/[T.sub.C]2 subsets11 through provocation of apoptosis and by acting through interaction between circulating soluble Fas and its receptor, FasL. (12) This feature is the logic behind the therapeutic application of intravenous immunoglobulin, which binds soluble Fas in the peripheral blood stream and has thus been thought to have a potential role in the treatment of TEN. (8) Synergistic action with helper T cells, including those of the [T.sub.H]2 subset, is implied by demonstration of the thymus- and activation-regulated chemokine. (13)
As mentioned above, herpes-associated erythema multiforme occurs in association with recurrent herpes simplex infection. Herpes DNA polymerase gene (Pol)and pol gene expression have been shown in recurrent erythema multiforme lesions in this setting, the latter through immunohistochemistry using antibody directed against pol in lesional skin of patients with herpes-associated erythema multiforme. In addition, the T-cell receptor variable chain repertoire in such patients is composed primarily of [V.sub.[beta]]2 chain-positive cells, suggesting a selective homing of lymphocytes to sites of viral antigen expression. (14)
Antibodies directed against desmosomal plaque proteins desmoplakins 1 and 2 have been described in a subset of patients with erythema multiforme major. Such studies suggest a humoral-based etiology in the propagation of lesions of erythema multiforme in these patients. The epitope is localized at the carboxy-terminal domain of desmoplakin and is responsible for the assembly of keratin filaments with desmosomes. Purified human antibody directed against the carboxy terminus of desmoplakins 1 and 2 when injected into newborn mice produces a constellation of changes that resembles erythema multiforme, suggesting a role for these antibodies in a subset of patients with erythema multiforme. (15,16)
THE CONNECTIVE TISSUE DISEASES
The manifestations of connective tissue disease in the skin encompass vasculopathy and vasculitis of leukocytoclastic, granulomatous, and lymphocytic subtypes; panniculitis; and dermal and epidermal infiltrates. The vasculitides and panniculitides are not considered in this review. With respect to the epidermal and dermal findings, the characteristic morphology seen in most skin lesions of lupus erythematosus, dermatomyositis, relapsing polychondritis (C.M.M. and A.N.C., personal observation, 1998), Sjogren syndrome, perniosis, and mixed connective tissue disease comprises a variable superficial or superficial and deep lymphocytic infiltrate in concert with a lymphocytic interface dermatitis ranging from a subtle cell-poor vacuolopathic injury pattern to a lichenoid infiltrate. (17-24) The hallmark of the cell-poor interface dermatitis is a sparse number of lymphocytes scattered along the DEJ with concomitant degenerative epithelial changes manifested by basilar vacuolopathy and dyskeratosis.
Relatively specific to the connective tissue diseases are hyperkeratosis with follicular and acrosyringeal plugging, epidermal atrophy, basement membrane zone thickening, and prominent dermal mucinosis. In those cases of cell-poor interface dermatitis owing to hypersensitivity reactions, where the insult is acute, no alteration of the stratum corneum or of the basement membrane zone is seen, nor is dermal mucinosis conspicuous. Chronic vasculopathic changes defined by reduplicated capillary and venular basement membranes imparting a hyaline appearance to the structure in concert with vascular ectasia also are useful morphologic clues.
With respect to the subclassification of connective tissue disease, a skin biopsy processed for both routine microscopy and for immunofluorescence analysis can provide information of essential value in support of the clinical diagnosis. However, knowledge of the clinical findings (Table 3), including in the context of extracutaneous disease and serology (Table 4), must be carefully integrated to arrive at a correct diagnosis. The pathologist is in a powerful position to positively influence this process.
CELL-POOR INTERFACE DERMATITIS OF AUTOIMMUNE CONNECTIVE TISSUE DISEASE
The 2 main connective tissue disease syndromes characterized by a cell-poor vacuolar interface dermatitis are systemic lupus erythematosus (17) and dermatomyositis. (19) Dermatologists apply the designation "acute" lupus erythematosus to the macular erythematous lesions of sudden onset in patients with SLE. In such acute lesions, the changes are subtle: the epidermis and the stratum corneum may not show any appreciable alterations, save for a focal and often subtle basilar vacuolopathy with a few lymphocytes tagging along the DEJ. (17) Patchy parakeratosis often is seen, and when the lesions have been present for a few weeks or more, there may be atrophy with attentuation of the retiform pattern. What is more striking are alterations of the dermis comprising splaying of the collagen fibers by mucin, along with superficial vascular plexus ectasia. Critical to the diagnosis is the presence of clinical signs and symptoms that fulfill the American College of Rheumatology criteria for SLE. (25) The differential diagnosis of this particular reaction pattern includes dermatomyositis, which will be discussed on the following page. We also have observed exactly this morphology in patients with constitutional symptoms simulating SLE in the setting of parvovirus B19 infection and Lyme disease. (26,27) In any patient who presents with (1) an acute symptom complex characterized by a skin rash with arthralgia and constitutional symptoms and (2) a skin biopsy histomorphologically resembling acute lupus erythematosus as defined by the criteria above, we recommend that consideration be given to the possibility of parvovirus B19 infection and Lyme disease, the latter particularly if the patient is from an endemic area.
Lupus erythematosus (LE) is an autoimmune disorder affecting skin, hematopoietic, and lymphoreticular organs, joints, kidney, lung, serosa, and cardiovascular structures in concert or in isolation. Lupus erythematosus is subdivided clinically into systemic (SLE), subacute cutaneous (SCLE), and discoid (DLE) forms, each with its own characteristic skin findings (Table 5). (17,18,25,28,29) A diagnosis of SLE is based on the presence of 4 or more of the criteria of the American College of Rheumatology. (25) Patients with SCLE manifest photodistributed, annular, papulosquamous eruptions accompanied by extracutaneous manifestations that, if present, are mild in nature, such as micro hematuria or arthralgia. (17,18) In DLE, disease is restricted to a cutaneous expression as one or more scaling plaques in photodistributed areas, typically involving the head and neck region.
Some authors aver that light microscopy has limitations in the subclassification of LE (29) and that the differences between subtypes reflect lesional age, a position to which we emphatically do not subscribe. Although it is true that a sparse lymphocytic interface infiltrate in the incipient lesion of DLE may mimic the more ominous systemic form and that discoid lesions may be seen in SLE patients, the pattern of disease progression, seen in some 5% of patients, is from DLE to SLE and not vice versa. (17,18) Certain features prove particularly helpful in subclassification. (17,18) Lesions of SLE, for example, show a pauci-inflammatory interface dermatitis with subtle basal layer vacuolopathy and no basement membrane zone thickening, keratotic follicular plugging, or acanthosis (Figure 3). Skin biopsies from patients with SCLE demonstrate suprabasilar exocytosis of lymphocytes with satellitosis to necrotic keratinocytes, slight or absent thickening of the basement membrane zone or follicular plugging, and no significant deep perivascular or periadnexal infiltrate. Atrophy is variable but usually is present in lesions of SLE and SCLE. Lesions of DLE generally manifest a heavier superficial and deep perivascular and periappendageal lymphocytic infiltrate, basement membrane zone thickening, keratotic follicular plugging, and variable acanthosis and atrophy. (17,18)
The differential diagnosis is that of lymphocytic interface dermatitis and embraces dermatomyositis, Sjogren syndrome, drug-related lupus erythematosus-like eruptions, polymorphous light eruption, perniosis, Jessner lymphocytic infiltrate, vitiligo, erythema multiforme, and certain delayed-type hypersensitivity and id reactions and viral exanthemata. (19,20,22,30-32) Kikuchi necrotizing lymphadenitis can produce skin lesions with a lymphocytic vasculopathy and cell-poor interface dermatitis identical to that seen in SLE, (33) but these may in turn reflect a forme fruste of SLE. The idiopathic lichenoid eruptions lichen planus, lichen nitidus, and lichen striatus can be problematic if parakeratosis and atrophy are seen. (32) Another pitfall is the distinction of lichenoid actinic keratoses from lesions of DLE in sun-damaged skin. Correlation with clinical, serologic, and direct immunofluorescence findings is therefore imperative. A clinician may be advised to submit a fresh biopsy of lesional or nonlesional skin, depending upon the clinical situation, in a physiologic medium such as normal saline or Michel transport medium for a lupus band test (LBT). (17,31)
Direct Immunofluorescence: The LBT and Its Implications for Pathogenesis of Connective Tissue Diseases.--The examination by fluorescence microscopy of frozen sections cut from lesional or nonlesional sun-exposed or non-sun-exposed skin received fresh, in saline, or in transport medium and incubated with fluorescein-conjugated antibodies monospecific for IgG, IgA, IgM, and complement constitutes the LBT. (17,31) The definition of a positive LBT is controversial, with differences in applicable criteria partly explaining the disparate results from various series. To be considered a positive LBT, deposition of IgM in sun-exposed skin should assume a continuous band over at least 50% of the width of the biopsy specimen and be at least moderate in intensity. (17,31) Weak IgM decoration along the DEJ is common in sun-exposed skin of patients who do not have LE, including in the context of normal skin specimens from healthy young adults; one study showed weak interrupted linear and granular IgM and Clq deposition along the DEJ in 25% of such samples, (34,35) whereas deposition of IgG, IgA, and C3 was seen in less than 5%. (17,31) In sun-protected skin, an interrupted band of IgM of at least moderate intensity is sufficient for designation as a positive LBT. The deposition of IgG is usually less intense than that of IgM, but false-positive results are rare. Thus, even an interrupted weak band of IgG deposition in sun-exposed skin. (17,18,31) can be considered a positive LBT. Concomitant IgA deposition enhances the specificity of the assay.
Sun-exposed lesional skin should be used to substantiate an initial diagnosis of LE so as to avoid a false-negative result due to reduced sensitivity in sun-protected skin. After the diagnosis of LE is made, a biopsy of non-lesional skin may be performed to further assist in sub-classification and prognostication; a positive LBT in sun-protected skin is predictive of systemic disease in the context of renal and other organ involvement. Although a skin lesion less than 2 months old may give negative direct immunofluorescence results, a positive LBT is seen in 90% of lesional skin biopsies from DLE and SLE patients. A negative LBT is almost invariably seen in nonlesional skin of DLE patients, whereas more than 80% of nonlesional skin biopsies show a positive LBT in SLE patients. (36) The LBT is positive in more than 90% of sun-exposed nonlesional skin biopsies from SLE patients with active disease versus only one third of patients with inactive disease. (36) Patients with SCLE manifest a positive LBT in only 30% to 50% of cases. (18)
Indirect Immunofluorescence: The Membrane Attack Complex of Complement and Its Putative Pathogenetic Role.--An indirect immunofluorescence methodology to detect the presence of the membrane attack complex of complement ([C.sub.5b-9]) in frozen sections may be a helpful adjunct to the LBT. (18) Keratinocyte fluorescence for [C.sub.5b-9] correlates with seropositivity for antibodies to extractable nuclear antigens Smith (Sm), Ro/SSA, La/SSB, or RNP. Patients with such antibodies often manifest distinctive clinical features and may be seronegative for antinuclear antibody. (18) Antibodies to La correlate with SICCA syndrome, and in the setting of SCLE may be predictive of a higher incidence of pulmonary disease, whereas patients with SLE who have antibodies to Ro are at greater risk for photosensitive skin eruptions, interstitial pneumonitis, myositis, myocarditis, and complete heart block. (37,38) Patients with antibodies to RNP frequently have Raynaud disease, sclerodactyly, and myopathy. (39) The Ro, La, Sm, and RNP antigens are small ribonucleoprotein macromolecules resident in nuclei and, in the case of Ro, the cytoplasms of all eukaryotic cells. (39) Binding of the Ro auto-antibody may be dependent on relocation of nuclear and cytoplasmic Ro antigens to the cell surface, an event that follows UV light exposure, viral infection, or estrogen therapy. (40-42) The presence of the membrane attack complex of complement in the basement membrane zone of skin lesions of patients with SLE and DLE and of renal tubules in patients with lupus nephritis, coupled with its absence in uninvolved skin, suggests its pathogenic role. (18,43) The surface binding of anti-Ro antibodies may allow adherence of [C.sub.5b-9], the latter forming plasmalemmal pores that, in concert with antibody-dependent cellular cytotoxicity, may be an important mechanism of keratinocyte injury. (18,43,44)
Vascular deposition of [C.sub.5b-9] is seen in dermatomyositis, in mixed connective tissue disease (MCTD), and in SLE patients who possess the lupus anticoagulant, antibodies to Ro, or biopsy-proven lymphocytic vasculitis. (18,19,38,45,46) Vasculitis has been shown in skeletal muscle biopsy specimens of LE patients with anti-Ro and anti-La antibodies, perhaps reflecting localization of the Ro antigen to endothelia, as has been shown in vitro. (38,40,47) In addition to myositis, SLE patients with anti-Ro antibodies manifest cutaneous lesions virtually identical to those seen in dermatomyositis. (38)
Dermatomyositis combines an inflammatory myopathy with characteristic skin lesions: the often-subtle heliotrope rash, the Gottron papule, a violaceous or hypopigmented papule over the joints of the fingers, erythema of the upper back (the "shawl sign"), extensive erythema of the extensor surfaces of the arms, scaly alopecia, and cuticular overgrowth with periungual telangiectasias. (19,48-50) Although myositis usually eventuates, skin involvement may be unaccompanied initially by objective evidence of muscle disease; such cases are referred to as amyopathic dermatomyositis or dermatomyositis sine myositis. (51) In adult populations, roughly one third of patients have an underlying malignancy, with statistical correlation demonstrated to lymphoma and to carcinomas of ovary, lung, pancreas, colon, and stomach. (48,52) Other manifestations include arthritis, myocarditis, and esophageal and pulmonary disease. Dermatomyositis appears to represent an aberrant immune response directed principally at endothelia in an immunogenetically predisposed individual following antigenic stimuli, such as neoplasms, drugs, or infections. With respect to the latter, it has been demonstrated that symptomatology in some cases of dermatomyositis reflects the presence of endotheliotropic viruses, such as parvovirus B19, that may alter endothelial antigenicity. (26,53)
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
Although some observers doubt that histologic criteria exist to distinguish dermatomyositis from other connective tissue diseases, such as SLE, SCLE, and MCTD, a constellation of light microscopic and immunofluorescent features, in concert with modern serology, is virtually pathognomonic for each of these entities. (17-21) Skin lesions of dermatomyositis manifest an atrophying cell-poor lymphocytic interface dermatitis accompanied by dermal mucinosis and vascular alterations that vary according to the age of the lesion biopsied and the presence or absence of myopathy: in patients with myopathic dermatomyositis, a characteristic injury pattern comprising a variably cell-poor, often thrombogenic lymphocytic vasculopathy mainly affecting the dermal papillae capillaries is seen.
The immunofluorescent profile of dermatomyositis comprises a negative LBT in conjunction with membrane attack complex ([C.sub.5b-9]) deposition along the DEJ and within blood vessels. (19) Similar vascular deposition is seen in the setting of SLE with antibodies to Ro and MCTD in the setting of antibodies to RNP. (21,38) All 3 conditions manifest myopathy, interstitial lung disease, and vasculitis, perhaps reflecting the common localization of antibodies to endothelial RNA antigens, the expression of which is up-regulated owing to endogenous or exogenous triggers.
The differentiating points of dermatomyositis from most cases of LE include active vascular injury characterized by endothelial cell necrosis and intraluminal fibrin deposition along with the end sequelae of vasculopathy, namely, reduction of vascular density and vascular ectasia (Figure 4). Hypovascularity is more conspicuous in cases of myopathic versus amyopathic dermatomyositis, implying that cutaneous vascular changes mirror those in muscle, whereby a critical reduction in vascular density is necessary to generate objective evidence of myopathy. (19,54,55) In MCTD, a lymphocytic interface injury pattern in concert with a lymphocytic vasculopathy can closely mimic dermatomyositis. Sclerodermoid tissue alterations are fairly common in the former but rare in the latter. Eosinophils are observed in biopsies of dermatomyositis in approximately 10% to 20% of cases (A.N.C. and C.M.M., personal observation, 2000) but are rare in idiopathic LE and MCTD, possibly reflecting a pathogenic role for aberrant delayed-type hypersensitivity in a minority of cases. The other causes of cell-poor vacuolopathic interface dermatitis, such as certain viral exanthems and hypersensitivity reactions, do not typically produce the same vascular injury patterns.
MIXED CONNECTIVE TISSUE DISEASE
Mixed connective tissue disease was first defined by Sharp et al (56) as a distinct rheumatic disease syndrome associated with high titers of antibody to an extractable nuclear antigen (nRNP). The typical features include Raynaud phenomenon, polyarthritis, polyserositis, myositis, sclerodactyly, restrictive lung disease, lymphadenopathy, and esophageal dysfunction. The cutaneous manifestations include SLE-like malar erythema, discoid plaques, an SCLE-like photodistributed eruption, swollen hands, sclerodactyly, and vasculitis.
Biopsies of the photodistributed eruptions show a cell-poor or lichenoid interface dermatitis with suprabasilar exocytosis around necrotic keratinocytes in the absence of deep periadnexal or perivascular extension or conspicuous follicular plugging. (21)
Nuclear keratinocyte immunoreactivity with IgG and [C.sub.5b-9] is demonstrated in all cases studied, accompanied by a positive LBT in roughly one half of cases. The in vivo speckled nuclear staining for IgG observed within keratinocytes of lesional and nonlesional skin correlates with antibodies to nRNP, the serologic hallmark of MCTD. Granular vascular deposition of immunoreactants, including [C.sub.5b-9], also is seen. (21)
The histopathology of MCTD mimics SCLE by virtue of the lymphocytic interface dermatitis, but it differs by showing vasculopathic alterations comprising ectasia, hypovascularity, and luminal thrombosis confined to the superficial vascular plexus, and sometimes by a concomitant sclerodermoid tissue reaction. (21) As mentioned, distinction from dermatomyositis can be problematic. Perniosis can demonstrate a similar pattern of interface injury with associated lymphocytic vasculopathy, but the exocytosis of lymphocytes in perniosis tends to be directed to retia and acrosyringia. (22) The interface dermatitis of erythema multiforme often is accompanied by pronounced papillary dermal edema, tissue eosinophilia, and epidermal colloid body formation in the absence of atrophy, vascular density reduction, or alterations of the stratum corneum.
Acute GVHD.--Clinical Features.--Acute GVHD manifests as erythematous patches occurring classically within 30 days following bone marrow transplantation. The skin is the most common site of involvement of GVHD, although other organ systems, such as the liver and gastrointestinal tract, also are involved frequently. The erythematous patches are typically acral in location and may form blisters in severe examples. Rarely, acute GVHD will develop in patients who receive autologous blood transfusions of nonirradiated blood or blood products that have not been subjected to leukopheresis.
Histopathology.--Acute GVHD produces a cell-poor lymphocytic interface inflammatory process that is often accentuated in the acrosyringia or in the terminations of hair follicles within the epidermis (Figure 5). (4) A grading scheme has been produced (Table 5). Damage to the epidermis results in loss of keratinocyte nuclear polarity, often attributed to the injection or administration of antineoplastic agents. These changes, in concert with individual cell necrosis of keratinocytes, occurring within 3 weeks of transplantation are more likely to be a sequel of the administration of cytoreductive agents than to acute GVHD. Eosinophils are sometimes present in lesions of acute GVHD. The most severe form of GVHD is indistinguishable from toxic epidermal necrolysis.
Differential Diagnosis.--The differential diagnosis of acute GVHD is that of the cell-poor interface dermatitides discussed at the onset of this review.
Interface Dermatitis of HIV Infection.--Clinical Features.--HIV infection may be associated with a variety of inflammatory dermatoses ranging from eosinophilic folliculitis to seborrheic dermatitis. Patients also may develop a peculiar multifocal plaque and patchlike erythema with scaling that becomes violaceous over time and is associated with pronounced hyperpigmentation. This eruption is often itchy and often shows a photodistribution with accentuation in the face and neck.
Histopathologic Features.--Common to the biopsies from many of these different pathophysiologic entities is a mononuclear cell interface dermatitis lateral to the main area of tissue pathergy. The infiltrate is typically lymphocytic with deep extension and accompanied by admixed eosinophils, histiocytes, and plasma cells.
Differential Diagnosis.--The differential diagnosis includes both cell-poor and cell-rich interface dermatitis of diverse ideology. This entity is included in the section on cell-poor interface dermatitis, as that is the classic and most common histomorphology of the interface inflammatory infiltrate in such lesions. (4)
Vitiligo is an acquired disorder in which an autoinflammatory lymphocytic infiltrate targets melanocytes in various locations, typically resulting in loss of pigment in skin, mucous membranes, and/or hair. (5) It appears that there is a complex interplay of genetic and environmental factors. Although there are case reports of vitiligo and hair pigment loss concurrent with the presence of underlying central nervous system tumors and with exacerbation of multiple sclerosis, case control studies fail to confirm a statistically significant increase of vitiligo in the latter setting. (57) Other autoimmune disorders of a systemic nature, including perncious anemia and type 1 diabetes mellitus, have been shown to have a statistical link to vitiligo, as does an increased incidence of alopecia areata and hypothyroidism. (58) On occasion, the areas of skin depigmentation in cases of vitiligo manifest a blaschkolinear distribution, implying a role for genetic mosaicism in the expression of the disease. (59) Triggering of the autoimmune diathesis by hepatitis C virus is implied by one case report in which the vitiligo lesions improved with antiviral therapy in a hepatitis C patient. (60)
At the advancing edge of a vitiligo lesion, skin biopsy shows a lymphocytic interface dermatitis that is variably cell poor. In the older center of such a lesion, postinflammatory melanophage accumulation may be seen in concert with a decrease or absence of melanocytes.
Pityriasis lichenoides is a self-limited dermatosis that usually presents in the first to third decades of life and shows a predilection for males. The onset ranges from an acute eruption termed pityriasis lichenoides et varioliformis acuta (PLEVA) to a chronic eruption, pityriasis lichenoides chronica (PLC), which tends to heal with postinflammatory hyperpigmentation. Lesions number from dozens to hundreds and frequently affect the anterior trunk and flexor aspects of the proximal extremities preferentially. The etiologic basis has long been postulated to be viral in nature, based in part on the clinical behavior and in part on phenotypic studies that show a characteristic infiltrate comprising CD8+ lymphocytes admixed with a minor populace of Langerhans cells or indeterminate cells. Although there are rare case reports of the detection of cytomegalovirus in endothelia of cases of PLEVA, (61) larger series have shown clonal restriction in lesions of PLEVA (1,3,62-64) and PLC. (1,3,62-64) The T-cell populations in both forms of pityriasis lichenoides show deletion of postthymic T-lineage markers, indicating that the process represents a true lymphoid dyscrasia. (63) Recently, we have shown relative defects in activity of the [CD4.sup.+]/[CD25.sup.+] T-regulatory cell population. (3)
In PLEVA, the DEJ often is obscured by a pure population of lymphocytes associated with basal layer vacuolopathy, colloid body formation, and variable but sometimes confluent epidermal necrosis. Intraepidermal hemorrhage, endothelial swelling, papillary dermal edema, bleeding, and wedge-shaped superficial and deep dermal lymphoid infiltrates are characteristic. (1) The parakeratotic scale often contains neutrophils. A rare extreme expression of the disease is the ulceronecrotic variant, in which necrotizing lymphocytic vasculitis produces confluent epidermal necrosis. (65) Parakeratosis often is confluent and epidermal injury less striking in PLC (Figure 6). (66) The epidermis may show superficial pallor, and melanophages often are present in the papillary dermis, reflecting basal layer injury. As PLEVA and PLC represent poles of a spectrum of injury, not all cases can be comfortably classified as one or the other; some authorities sign out most cases simply as "pityriasis lichenoides" without further qualification. (16)
The differential diagnosis includes pityriasis rosea, in which discrete mounds of parakeratin overlie foci of epidermal spongiosis, and small-plaque parapsoriasis, where the degree of epithelial injury generally is less and the infiltrate more sparse. (66) The common viral exanthem pattern may mimic PLEVA by virtue of a perivascular and interface lymphocytic infiltrate accompanied by basilar vacuolopathy with scattered cytoid bodies and streak dyskeratosis, but the epidermis usually is surmounted by a basket-weave pattern of orthokeratinization, and the lymphocytic infiltrate is usually only superficial. Similar features also typify evolving lesions of erythema multiforme, which may show eosinophils in the dermal inflammatory populace, a finding that helps to enable distinction. Secondary syphilis merits consideration, but it usually has a significant plasma cell component, which is at variance with the pure lymphoid populace of pityriasis lichenoides. Connective tissue diseases with pronounced keratinocyte degeneration, such as SCLE, can be problematic if a superfical shave biopsy is submitted and the deep infiltrates of pityriasis lichenoides, which would enable distinction, cannot be assessed. Large plaque parapsoriasis may be considered, as it manifests an interface pattern of colonization with a wiry pattern of horizontal collagen fiber thickening typically absent in lesions of PLEVA but common in lesions of PLC. (63)
HYPOPIGMENTED INTERFACE VARIANT OF PRELYMPHOMATOUS CUTANEOUS LYMPHOID DYSCRASIA
This entity has fallen under the appellation hypopigmented mycosis fungoides and manifests a characteristic race, sex, and age predilection. (1) Patients are usually younger and of African American extraction; there is a female predilection. The lesions often involve truncal areas similar to typical mycosis fungoides. While hypopigmented mycosis fungoides has been described, a hypopigmented variant of large plaque parapsoriasis as a precursor state has not been specifically alluded to in the literature. There is, however, a form of interface parapsoriasis. Many of the cases of hypopigmented mycosis fungoides described in the literature do not manifest a clinical course or a histology specimen diagnostic of fully evolved mycosis fungoides, and are thus better characterized as a form of cutaneous lymphoid dyscrasia falling under the designation of hypopigmented large plaque parapsoriasis and/or hypopigmented epitheliotropic T-cell dyscrasia.
The histology comprises a low-density epitheliotropic lymphocytic infiltrate with characteristic colonization of the basal layer unaccompanied by significant destructive epithelial changes. The lack of epithelial injury distinguishes this process from a true immunologically mediated interface dermatitis. There is often a haphazard single-cell pattern of lymphocyte migration into the upper layers of the epidermis. Pautrier microabscesses are not seen; the presence of such discrete aggregates of neoplastic lymphocytes in the epidermis would warrant categorization as mycosis fungoides. Many of the cells in the epidermis have a distinctly cerebriform appearance; a small cell-dominant infiltrate predominates. There may be some accentuation of migration to involve hair follicles and acrosyringia.
There is a significant reduction in CD7 and CD62L expression; often CD62L shows a greater loss than CD7. From a phenotypic perspective, there is a predominance of CD8 lymphocytes over those of the CD4 subset. (1) If the CD8 lymphocytes are reactive and induce lesional resolution and/or contribute to immunologic surveillance to prevent disease progression, one would expect to see greater epithelial injury of bystander keratinocytes as the basis for the leucoderma.
We have studied similar cases and have observed the same restricted oligoclonal T-cell repertoire between biopsies, adding credence to the designation of these cases as forms of cutaneous lymphoid dyscrasia.
CELL-RICH INTERFACE DERMATITIS: THE LICHENOID
Cell-rich interface dermatitis is defined by us as an inflammatory infiltrate along the DEJ that is of sufficient intensity that it obscures, at least focally, the basilar keratinocytes. When this process is bandlike in character it may be termed lichenoid. Interface dermatitis can comprise a mainly lymphocytic population, a mainly lymphohistiocytic infiltrate, or a granulocyte-predominant interface injury pattern. The latter may, in turn, be classified as a neutrophil-predominant pattern, as seen, for example, in linear IgA disease or bullous lupus erythematosus, or an eosinophil-predominant pattern, as seen in bullous pemphigoid. The granulocyte-predominant injury pattern is beyond the scope of this review, which confines itself to the lymphocyte-predominant interface injury pattern. Lichenoid interface dermatitis is seen in numerous conditions, including lichen planus, lichenoid hypersensitivity reactions of drug- or contact-based etiology, lichenoid reactions in the setting of hepatobiliary disease, (67) secondary syphilis, and autoimmune CTD. Lichenoid pigmentary purpura of Gougerot and Blum is a form of pigmented purpuric dermatosis which often is associated with a drug-based etiology and is discussed in more detail below. Most of the foregoing will be considered with a specific emphasis on distinctive light microscopic features. When a lichenoid interface dermatitis is present, the onus is on the clinician to consider, to investigate, and to exclude, where appropriate, these various disorders. The idiopathic lichenoid disorders include lichen planus, lichen nitidus, and lichen striatus. The prototype is lichen planus. (32)
Considered an idiopathic dermatosis, lichen planus manifests as violaceous, itchy, flat-topped, polygonal papules covered by a reticulated surface scale termed Wickham striae. Lesions typically manifest on the volar aspect of the forearms and other flexural surfaces of acral parts; genitalia are often involved. Lesions may be widespread. Oral changes accompany the cutaneous eruption in roughly one half of cases, and these manifest as linear or reticulate whitish plaques and as lacy white patches along the occlusal lines of the buccal mucosa and elsewhere. Nail changes are frequently seen and manifest as dystrophies with ridging and splitting of the distal aspect of the nail plate. Variants of lichen planus include the atrophic form (lichen planus actinicus), hypertrophic lichen planus, bullous lichen planus, and linear lichen planus. Hypertrophic lichen planus is particularly problematic diagnostically, as significant pseudoepiltheliomatous hyperplasia occurs in response to chronic rubbing and irritation, classically on the shins, and closely mimics well-differentiated squamous cell carcinoma histologically. Scarring alopecia may be seen in association with lichen planus; this is termed lichen planopilarias. A lichen planus-like eruption can be seen in patients infected with hepatitis C virus or hepatitis B virus (68); case control studies show conflicting data regarding the significance of the association, but most observers accept that one exists. (69,70)
Compact orthohypokeratosis overlying an epidermis that shows wedge-shaped thickening of the granular cell layer and a "sawtooth" pattern of acanthosis is prototypic for lichen planus (Figure 7). A dense, bandlike lymphocytic infiltrate obscures the DEJ; in our lexicon, this is the defining feature of a lichenoid dermatosis. There is no deep extension of the infiltrate in idiopathic lichen planus, and eosinophils should number no more than 1 to 3 per any given tissue section of a 4-mm punch biopsy specimen. As they undergo destruction, the basal keratinocytes release melanin into the dermis with compensatory accumulation of dermal melanophages to cause postinflammatory hyperpigmentation clinically. Repair of the basal layers of the epidermis is associated with the "squamotization" of the basal layer, with replacement of the normal cuboidal progenitor cells in that anatomical location. Cleft-like spaces may form between the undersurface of the epidermis and dermis; these are termed Max-Joseph spaces. This process may eventuate in clinically evident bullous lesions (ie, bullous lichen planus).
As lesions of lichen planus are itchy, they may be rubbed and chronically irritated; parakeratosis may supervene. With persistent rubbing and irritation, pseudoepitheliomatous change may be superimposed to generate hypertrophic lichen planus. End-stage lesions may show compact orthohyperkeratosis with slight thickening of the epidermis but lack the significant inflammatory component, therefore having a nondiagnostic histomorphology. Biopsies from mucosal sites similarly show a bandlike lymphocytic and histocytic infiltrate obscuring the epithelial-submucosal interface, but in mucosal sites, tissue eosinophilia is often profound. Atrophic lichen planus (sometimes called lichen planus actinicus) has atrophy with attenutation of the retiform pattern.
From a pathophysiologic perspective, the lymphoid cells are T cells expressing CD3 and CD4. Increased numbers of [gamma][delta] T cells are often present along the DEJ; this is a form of T lymphocyte that is only a minor fraction of circulating T cells in humans and responds to a broad array of microbial pathogens, including in the context of those with superantigen properties. This is one clue to the possible infection-triggered etiology of lichen planus, including in the context of the hepatitis C-triggered lesions. (68)
If significant tissue eosinophilia is present in the lichenoid inflammatory process, a lichenoid drug eruption or fixed drug eruption should be suspected. (71) Solitary lesions of benign lichenoid keratosis or lichenoid actinic keratosis show a dome-shaped contour with a peripheral actinic lentigo in the former instance and basal layer crowding and atypia in the latter. Furthermore, benign lichenoid keratosis occurs typically on the trunk in the clinical setting of a suspect basal cell carcinoma, whereas lichenoid actinic keratosis is seen in the setting of dermatoheliosis and is frequently a multifocal process. A clinician should always specify when a lichenified dermatosis, as opposed to a solitary lesion, is biopsied.
Lichen nitidus classically presents clinically as miliary papules involving the arms, trunk, and penis, (72-75) and histologically as a superficial lymphocytic and granulomatous infiltrate that imparts a "ball in claw" appearance to the epidermis and subjacent dermal papillae (Figure 8). (32)
The presentation of lichen striatus is as a linear eruption involving an extremity, biopsies of which prototypically show a hybrid lichenoid and ezcematous dermatitis with suprabasilar dyskeratosis. The lesions initially manifest a spongiotic dermatitis histologically, with the lichenoid mononuclear cell infiltrate supervening and then being followed by fibrosis with horizontal papillary dermal fibrosis. In our hands, lichen striatus has proven to be one of the disorders manifesting a lichenoid and granulomatous morphology. An additional and characteristic finding in our series was lymphocytic eccrine hidradenitis and erythrocyte extravasation. (32)
LICHENOID AND GRANULOMATOUS DERMATITIS
AS AN ENTITY SUI GENERIS
Introduction and Clinical Features
In 2000, we described 40 patients in whom biopsies showed a lichenoid and granulomatous dermatitis associated clinically with a variety of disorders. The individual lesions presented most frequently as lichenoid papules which, in roughly 20% of the cases, ultimately were held to be examples of the idiopathic lichenoid eruptions lichen nitidus, lichen striatus, and even lichen planus. Other cases were held to represent a reflection of underlying systemic autoimmune disrders, microbial id reactions, and drug eruptions. Implicated drug classes included lipid-lowering agents, p-blockers, ACE inhibitors, H2 antagonists, plaquenil, and antibiotics, several of which are known to provoke lichenoid eruptions. (32) Some of these patients had underlying medical illnesses with proven associations with cutaneous granulomatous infiltrates, such as thyroiditis, diabetes mellitus, hepatitis C, rheumatoid arthritis, and Crohn disease, (32,76) suggesting a form of granulomatous Koebnerization of a lichenoid drug reaction. With respect to the infectious associations, these included patients with active infection or an id response to nonviable microbial antigen. The organisms were among those with superantigen properties, namely, viruses, mycobacterial and treponemal species, and streptococci. Superantigens are microbial proteins that interact with the variable region of the T-cell receptor and the conserved residues of class II antigen-presenting cells, thereby stimulating a much larger percentage of the T-cell repertoire than do traditional bacterial oligopeptides. (30,77-83) Follicular inflammation, lymphocytic eccrine hidradenitis, and perineural infiltrates are also characteristic, particularly in postherpetic lesions.
The defining histomorphology is a bandlike infiltrate of lymphocytes and histiocytes closely applied to the undersurface of the epidermis and obscuring the DEJ. Additional light microscopic features in the drug-associated cases included parakeratosis, keratinocyte necrosis, acrosyringeal accentuation, a lichenoid pigmentary purpura-like reaction pattern, granulomatous vasculitis, tissue eosinophilia, plasmacellular infiltrates, and sparing of the deep dermis. Some cases manifested lymphoid atypia in the setting of ingestion of drugs with immune-dysregulating properties.
[FIGURE 9 OMITTED]
[FIGURE 10 OMITTED]
LICHENOID CTD SYNDROMES
The prototypic lichenoid connective tissue disease syndromes include SCLE, anti-Ro-associated SLE, and MCTD.
The lichenoid connective tissue disease syndromes have in common a polycyclic annular and/or papulosquamous photodistributed eruption involving the head, neck, arms, and the "v" of the chest and upper back. In SCLE, there are no significant extracutaneous stigmata of collagen vascular disease. Serologic testing reveals anti-Ro antibodies in the majority of cases. Some cases of SCLE are of drug-based etiology, where the main implicated drugs include calcium channel blockers, (23) Griseofulvin, thiazides, and antihistamines. (24) In anti-Ro-associated SLE, while the cutaneous eruption may be morphologically indistinguishable from SCLE, there are other extracutaneous stigmata indicative of a systemic CTD syndrome, including renal dysfunction, musculoskeletal complaints, pulmonary disease, and central nervous system manifestations. (38) Also, the patients may have cutaneous and extracutaneous stigmata indicative of vascular compromise, as manifested by ulceration, digital infarcts, palpable purpura, mononeuritis multiplex, gastrointestinal ulcerations, and myocardial infarction. In MCTD, the patients have a constellation of characteristic extracutaneous manifestations, which include Raynaud phenomenon, sclerodactyly, and pulmonary hypertension. (21) In MCTD, while patients may have a skin rash that resembles SCLE, other characteristic features allow the distinction, namely, anti-RNP antibodies, sclerodactyly, variable myositis, Raynaud phenomenon, and pulmonary hypertension.
The prototypic histomorphology of the lichenoid CTD syndromes is one of variable epidermal hyperplasia and atrophy accompanied by an interface dermatitis that ranges in quality from being lichenoid in nature to a cell-poor vacuolar dermatitis. There is variable middermal and deep dermal perivascular extension of the infiltrate. Dermal mucinosis is ubiquitous. Also characteristic is the presence of suprabasilar dyskeratosis, with lymphocyte satellitosis around degenerating keratinocytes. Other concomitant inflammatory cell elements, such as epithelioid histiocytes, plasma cells, and eosinophils, are uncommon, although it is emphasized that in a small minority of cases of drug-associated SCLE, eosinophils and granulomata may be observed. (23,24,71) It has been our experience that a significant vasculopathy is not present in SCLE. In contrast, in both anti-Ro-associated SLE and MCTD, a microangiopathy similar to that encountered in dermatomyositis is a frequent finding. In particular, one may observe zones of vascular density reduction, endothelial cell necrosis, and variable luminal fibrin deposition. (38)
LICHENOID DRUG ERUPTION
Lichenoid drug eruptions are caused by a variety of medications ranging from thiazide diuretics to antihistaminics and antihypertensive agents with [beta]-blocking and ACE inhibitor properties. (24,71) As mentioned above, the lichenoid drug eruption is characterized by deep extension of an infiltrate that includes eosinophils but has other features typical for lichen planus morphologically at both a clinical and histologic level.
FIXED DRUG ERUPTIONS
The fixed drug eruption is characterized by the recurrence of one or a few plaques of macular erythema that manifest a persistent or slowly expanding lesion at one or a few anatomical locations in association with the repeated ingestion of a drug of a particular class or a similar class. These lesions typically form a violaceous hue and leave significant postinflammatory hyperpigmentation. Certain classes of antibiotics and nonsteroidal anti-inflammatory agents taken on an over-the-counter basis by patients are frequently included. (71)
The morphology of the fixed drug eruption is characterized by a heavy bandlike infiltrate, which may be patchy and is associated with particular homing of the lymphoid infiltrate to the retia; clusters of colloid (Civatte) bodies around the tips of rete ridges are a characteristic finding. Full-thickness necrosis due to confluent colloid body formation may be seen. The infiltrates often include scattered tissue eosinophils, and postinflammatory hyper-pigmentation at a histologic level is often significant. On occasion, a neutrophilic fixed-drug eruption with abundant karryorhectic debris may be seen.
LICHENOID PURPURA OF GOUGEROT AND BLUM
Introduction and Clinical Features
The pigmentary purpuras are a heterogeneous group of dermatoses defined by specific clinicopathologic features but sharing dermal lymphocytic infiltrates and hemorrhage at the light microscopic level. The prototype, Schanmberg disease, manifests as a purpuric macular eruption of the shins in middle-aged patients. A variety of other forms are described, including the eczematoid purpura of Doucas and Kapetanikus, the annular purpura of Majochi, and 3 variants that produce a bandlike and often heavy infiltrate in the upper dermis. These latter variants include lichen aureus, the lichenoid pigmentary purpura of Gougerot and Blum, and atypical pigmentary purpura, (84) a term applied by us to cases of pigmentary purpura in which individual lesions, although clinically presenting as pigmentary purpura, show lymphoid atypia by light microscopy. Previous reports suggest that the pigmented purpuric dermatoses with heavy lymphoid infiltration may presage fully evolved mycosis fungoides. A drug-based etiology is clearly implicated in some one third of cases of atypical pigmentary purpuras, the culprit agents being calcium channel blockers, lipid-lowering agents, [beta]-blockers, ACE inhibitors, antihistamines, antidepressants, or analgesics. (84,85) Although these atypical and lichenoid pigmentary purpuras may express clonal restriction of T cells, showing loss of pan-T-cell markers in a fashion identical to mycosis fungoides, most are reversible with appropriate drug modulation, or remain biologically stable. We consider them a forme fruste of lymphoproliferative disease which, in the vast majority of cases, has an indolent course.
Although lesions of drug-related atypical pigmentary purpura may mimic mycosis fungoides or idiopathic atypical pigmentary purpura histologically, the cytomorphology of the intraepidermal populace, when more atypical than the dermal based populace, is predictive of mycosis fungoides versus drug-related or idiopathic cases.
In lesions of chronic GVHD, a lichenoid morphology supervenes. (4) Ultimately, chronic inflammation in this setting may result in a sclerodermoid tissue reaction. We have seen accelerated variants of lichenoid GVHD in the setting of peripheral blood stem cell transplantation.
LICHENOID MYCOSIS FUNGOIDES
This is an unusual variant that shows true lichenoid destructive interface change exactly recapitulating classic lichen planus. However, almost invariably there are areas of haphazard lymphocyte migration into the epidermis that are indistinguishable from those morphologic changes encountered in typical mycosis fungoides. Other discriminating features are enhanced cytologic atypia.
Interface dermatitis can be broadly broken down into cell-poor and cell-rich inflammatory processes and addressed as well by the character of the cellular infiltrate. A bandlike inflammatory infiltrate that obscures the DEJ, that is, a lichenoid inflammatory process, calls to mind a set of differential diagnostic considerations that is distinct from those that flow from a cell-poor interface injury pattern. A comprehension of the pathobiology of the different entities that fall into these broad categories of injury will point the astute pathologist toward a recognition of distinctive histomorphologic features which can, with clinicopathologic correlation, enable a precise etiopathologic diagnosis and so promote an enhancement of patient care.
(1.) Magro CM, Guitart J, Crowson AN. Precursor lesions of cutaneous T cell lymphoma. In: Magro CM, Crowson AN, Mihm MC, eds. The Cutaneous Lymphoid Proliferations. New York, NY: John Wiley and Sons; 2007:93-140.
(2.) Magro CM, Mihm MC, Crowson AN. Mycosis fungoides. In: Magro CM, Crowson AN, Mihm MC, eds. The Cutaneous Lymphoid Proliferations. New York, NY: John Wiley and Sons; 2007:267-299.
(3.) Magro CM, Crowson AN, Morrison C, Li J. Pityriasis lichenoides chronica: stratification by molecular and phenotypic profile. Hum Pathol. 2007;38:479-490.
(4.) Horn TD. Interface dermatitis. In: Barnhill R, Crowson AN, eds. Textbook of Dermatopathology. 2nd ed. New York, NY: McGraw-Hill Co; 2004:35-60.
(5.) Spritz RA. The genetics of generalized vitiligo and associated autoimmune diseases. Pigment Cell Res. 2007;20:271-278.
(6.) Cheng W, Gilliam AC, Castrovinci A, Pazirandeh M. Anti-thyroid autoantibody-associated interface dermatitis in individuals with undifferentiated connective tissue disease--an unrecognized subset of autoimmune disease? J Rheumatol. 2007;34:81-88.
(7.) Schalock PC, Dinulos JG, Pace N, Schwarzenberger K, Wenger JK. Erythema multiforme due to Mycoplasma pneumoniae infection in two children. Pediatr Dermatol. 2006;23:546-555.
(8.) French LE. Toxic epidermal necrolysis and Stevens Johnson syndrome: our current understanding. Allergol Int. 2006;55:9-16.
(9.) Gravante G, Delogu D, Marianetti M, Trombetta M, Esposito G, Montone A. Toxic epidermal necrolysis and Steven Johnson syndrome: 11-years-experience and outcome. Eur Rev Med Pharmacol Sci. 2007;11:119-127.
(10.) Ueta M, Sotozono C, Tokunaga K, Yabe T, Kinoshita S. Strong association with HLA-A*0206 and Stevens-Johnson syndrome in Japanese. Am J Ophthalmol. 2007;143:367-368.
(11.) Nishio D, Izu K, Kabashima K, Tokura Y. T cell populations propogating in the peripheral blood of patients with drug eruptions. J Dermatol Sci. 2007;48: 25-33.
(12.) Stur K, Karlhofer EM, Stingle G. Soluble Fas ligand: a discriminating feature between drug-induced skin eruptions and viral exanthems. J Invest Dermatol. 2007;127:802-807.
(13.) Quaglino P, Caproni M, Antiga E, et al. Serum levels of the Th1 promoter IL-12 and the Th2 chemokine TARC are elvated in erythema multiforme and Stevens-Johnson syndrome/toxic epidermal necrolysis and correlate with soluble Fas ligand expression: an immunoenzymatic study form the Italian Group of Immunopathology. Dermatology. 2007;214:296-304.
(14.) Kokuba H, Imafuku S, Huang S, Aurelian L, Burnett JW. Erythema multi-forme lesions are associated with expression of herpes simplex virus (HSV) gene and qualitative alterations in the HSV-specific T-cell response. Br J Dermatol. 1198;138:952-964.
(15.) Fukiwake N, Moroi Y, Urabe K, Ishii N, Hashimoto T, Furue M. Detection of autoantibodies to desmoplakin in a patient with oral erythema multiforme. Eur J Dermatol. 2007;17:238-241.
(16.) Crowson AN. Superficial and deep perivascular dermatitis. In: Barnhill R, Crowson AN, eds. Textbook of Dermatopathology. New York, NY: McGraw-Hill Co; 2004:79-96.
(17.) Crowson AN, Magro CM. The cutaneous pathology of lupus erythematosus: a review. J Cutan Pathol. 2001;28:1-23.
(18.) Magro CM, Crowson AN, Harrist TJ. The use of antibody to [C.sub.5b-9] in the subclassification of lupus erythematosus. Br J Dermatol. 1996;134:855-862.
(19.) Crowson AN, Magro CM. The role of microvascular injury in the pathogenesis of cutaneous lesion of dermatomyositis. Hum Pathol. 1996;1926:14-19.
(20.) Teramoto N, Katayama I, Arai H, et al. Annular erythema: a possible association with Sjogren's syndrome. J Am Acad Dermatol. 1989;20:596-601.
(21.) Magro CM, Crowson AN, Regauer S. The dermatopathology of mixed connective tissue disease. Am J Dermatopathol. 1997;19:205-212.
(22.) Crowson AN, Magro CM. Idiopathic perniosis and its mimics: a clinical and histological study of 38 cases. Hum Pathol. 1997;28:478-484.
(23.) Crowson AN, Magro CM. Subacute cutaneous lupus erythematosus arising in the setting of calcium channel blocker therapy. Hum Pathol. 1997;28:67-73.
(24.) Crowson AN, Magro CM. Lichenoid and subacute lupus erythematosus-like dermatitis associated with antihistamine therapy. J Cutan Pathol. 1999;26:95-99.
(25.) Tan EM, Cohen AS, Fries JF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 1 982;2 5:1271-1277.
(26.) Magro CM, Dawood MR, Crowson AN. The cutaneous manifestations of human parvovirus B19 infection. Hum Pathol. 2000;31:488-497.
(27.) Crowson AN, Magro CM, Dawood MR. A causal role for parvovirus B19 infection in adult dermatomyositis and other autoimmune syndromes. J Cutan Pathol. 2000;27:505-515.
(28.) David-Bajar KM. Subacute cutaneous lupus erythematosus. J Invest Dermatol. 1993;100:2S-8S.
(29.) Jerdan MS, Hood AF, Moore GW, Callen JP. Histopathologic comparison of the subsets of lupus erythematosus. Arch Dermatol. 1990;126:52-55.
(30.) Magro CM, Crowson AN. Interface and granulomatous dermatitis as a manifestation of antecedent microbial infection: the superantigen id reaction. J Cutan Pathol. 1998;25:538-544.
(31.) Magro CM, Crowson AN. The application of immunofluorescence testing in diagnostic dermatopathology. Adv Dermatol. 1999;15:441-486.
(32.) Magro CM, Crowson AN. Lichenoid and granulomatous dermatitis: a novel cutaneous reaction pattern. Int J Dermatol. 2000;39:126-133.
(33.) Crowson AN, Allen P, Cornelison R, et al. Kikuchi's histiocytic necrotizing lymphadenitis with skin involvement: a clinicopathologic study. Lab Invest. 2002; 82:97A.
(34.) Leibold AM, Bennion S, David-Bajar K, Schleve MJ. Occurrence of positive immunofluorescence in the dermoepidermal junction of sun-exposed skin of normal adults. J Cutan Pathol. 1994;21:200-206.
(35.) Weigand DA. Lupus band test: anatomic regional variations in discoid lupus erythematosus. J Am Acad Dermatol. 1986;14:426-428.
(36.) Jaworsky C. Connective tissue diseases. In: Elder DE, Johnson BE, Elenitsas R, Johnson BE, Murphy GF, eds. Lever's Histopathology of the Skin. 9th ed. Philadelphia, Pa: JB Lippincott Co; 2005:293-322.
(37.) Ben-Chetrit E. The molecular basis of the SSA/Ro antigens and the clinical significance of their autoantibodies. Br J Rheumatol. 1993;32:396-402.
(38.) Magro CM, Crowson AN. The cutaneous pathology associated with seropositivity for antibodies to Ro/SSA: a clinicopathological study of 23 adult patients without subacute cutaneous lupus erythematosus. Am J Dermatopathol. 1999;21: 129-137.
(39.) McCauliffe DP, Sontheimer RD. Molecular characterization of the Ro/SS-A autoantigens. J Invest Dermatol. 1 993;100:73S-79S.
(40.) Furukawa F, Kahihara-Sawami M, Lyons MB, Norris DA. Binding of antibodies to ENA SS-A/Ro and SS-B/La is induced on the surface of human keratinocytes by UVL: implications for the pathogenesis of photosensitive cutaneous lupus. J Invest Dermatol. 1990;94:77-85.
(41.) Tesar JT, Armstrong J. Expression of Ro/SSA and La/SSB antigens on epithelial cell surface following in vitro adenovirus (AV) infection [abstract]. Arthritis Rheum. 1986;S74:C20.
(42.) Norris DA. Pathomechanisms of photosensitive lupus erythematosus. J Invest Dermatol. 1993;100:58S-68S.
(43.) Helm KF, Peters MS. Deposition of membrane attack complexin cutaneous lesions of lupus erythematosus. J Am Acad Dermatol. 1993;28:687-691.
(44.) Crowson AN, Magro CM. [C.sub.5b-9] deposition at the dermoepidermal junction in lupus erythematosus [letter]. J Am Acad Dermatol. 1994;31:515-516.
(45.) Magro CM, Crowson AN. The immunofluorescence findings in cutaneous lesions of dermatomyositis: a comparative study versus lupus erythematosus. J Cutan Pathol. 1997;24:543-552.
(46.) Lim KL, Abdul-Wahab R, Lowe J, Powell RJ. Muscle biopsy abnormalities in systemic lupus erythematosus: correlation with clinical and laboratory parameters. Ann Rheum Dis. 1994;53:178-182.
(47.) Lanto B, Bohm F, Maffert W, Sonnichsen N. The cytotoxic effect of anti-Ro- (SS-A) antibodies and UVA light on human endothelial cells in vitro. Dermatol Monatsschr. 1990;176:305-311.
(48.) Kovacs SO, Kovacs C. Dermatomyositis. J Am Acad Dermatol. 1998;39: 899-920.
(49.) Callen JP. Dermatomyositis. Lancet. 2000;355:53-57.
(50.) Sontheimer RD, Costner MI. Dermatomyositis. Chap 157. In: Wolff K, Goldsmith L, Katz S, Gilchrest B, Paller A, Leffell D, eds. Dermatology in General Medicine. 7th ed. New York, NY: McGraw-Hill. In press.
(51.) Euwer RL, Sontheimer RD. Amyopathic dermatomyositis: a review. J Invest Dermatol. 1993;100:124S-127S.
(52.) Hill CL, Zhang Y, Sigurgeirsson B, et al. Frequency of specific cancer typesin dermatomyositis and polymyositis: a population-based study. Lancet. 2001; 357:96-100.
(53.) Magro CM, Crowson AN, Dawood M, Nuovo GJ. Parvoviral infection of endothelial cells in cutaneous lesions--possible role in vasculitis and autoimmune diseases. J Rheumatol. 2002;29:1227-1235.
(54.) Kissel JT, Halterman RK, Rammohan KW, Mendell JR. The relationship of complement-mediated microvasculopathy to the histological features and clinical duration of disease in dermatomyositis. Arch Neurol. 1991;48:2 6-30.
(55.) Emslie-Smith AM, Engel AG. Microvascular changes in early and advanced dermatomyositis: a quantitative study. Ann Neurol. 1990;27:343-356.
(56.) Sharp GC, Irwin WS, Tan EM, Gould RG, Holman JR. Mixed connective tissue disease: an apparently distinct rheumatic disease syndrome associated with a specific antibody to an extractable nuclear antigen. Am J Med. 1972;52:148-159.
(57.) Ramagopalan SV, Dyment DA, Valdar W, et al. Autoimmune disease in families with multiple sclerosis: a population-based study. Lancet Neurol. 2007; 6:604-610.
(58.) Gopal KV, Rama Rao GR, Kumar YH, et al. Vitiligo: a part of a systemic automimmune process. Indian J Dermatol Venereol Leprol. 2007;73:162-165.
(59.) Schallreuter KU, Kruger C, Rokos H, Hasse S, Zothner C, Panske A. Basic research confirms co-existence of acquired Blaschkolinear vitiligo and acrofacial vitiligo. Arch Dermatol Res. 2007;299:225-230.
(60.) Taffaro M, Pyrsopoulos N, Cedron H, etal. Vitiligo improvementin a hepatitis C patient after treatment with PEG-interferon alpha 2A and ribavirin: a case report [published online ahead of print April 12, 2007]. DigDis Sci.
(61.) Tsai KS, Hsieh HJ, Chow KC, Lin TY, Chiang S, Huang HH. Detection of cytomegalovirus infection in a patient with febrile ulceronecrotic Mucha-Habermann's disease. Int J Dermatol. 2001;40:694-698.
(62.) Dereure O, Levi E, Kadin ME. T-cell clonality in pityriasis lichenoides et varioliformis acuta: a heteroduplex anaylsis of 20 cases. Arch Dermatol. 2000; 136:483-486.
(63.) Magro CM, Morrison C, Kovatich A, Burns F, Crowson AN. Pityriasis lichenoides is a cutaneous T-cell dyscrasia: a clinical, genotypic, and phenotypic study. Hum Pathol. 2002;33:788-795.
(64.) Shieh S, Mikkola DL, Wood GS. Differentiation and clonality of lesional lymphocytes in pityriasis lichenoides chronica. Arch Dermatol. 2001;137:305-308.
(65.) Lopez-Estebaranz JL, Vanaclocha F, Gil R, etal. Febrile ulceronecrotic Mucha-Habermann disease. J Am Acad Dermatol. 1993;29:903-906.
(66.) Benmaman O, Sanchez JL. Comparative clinicopathological study on pityriasis lichenoides chronica and small plaque parapsoriasis. Am J Dermatopathol. 1988;10:189-196.
(67.) Magro CM, Crowson AN, Mihm MC. Cutaneous manifestations of gastrointestinal disease: In: Elder DE, Johnson BE, Elenitsas R, Johnson BE, Murphy GF, eds. Lever's Histopathology of the Skin. 9th ed. Philadelphia, Pa: JB Lippincott Co; 2005:419-434.
(68.) Crowson AN, Nuovo G, Ferri C, Magro CM. The dermatopathological manifestations of hepatitis C infection: a clinical, histologic and molecular assessment of 35 cases. Hum Pathol. 2003;6:573-579.
(69.) de Mattos Camargo Grossman S, de Aguiar MC, Teixeira R, do Carmo MA. Oral lichen planus and chronic hepatitis C: a controversial association. Am J Clin Pathol. 2007;127:800-804.
(70.) Noam Y, Dagon N, Shinar E, Meir G. Association between hepatitis C virus infection and oral lichen planus in Israeli patients. Isr Med Assoc J. 2007;9:370-372.
(71.) Crowson AN, Brown TJ, Magro CM. Progress in the understanding of the pathology and pathogenesis of cutaneous drug eruptions: implications for management. Am J Clin Dermatol. 2003;4:407-428.
(72.) Ellis FA, Hill WF. Is lichen nitidus a variety of lichen planus? Arch Dermatol Syph. 1938;38:569-573.
(73.) Senea FE, Caro MA. Lichen striatus. Arch Dermatol Syph. 1941;43:116-133.
(74.) Lee H. Lichen striatus. Lancet. 1951;1:615-617.
(75.) Staricco RG. Lichen striatus: a study of 15 new cases with special emphasis on the histopathological changes and review of the literature. Arch Dermatol. 1959;79:311-324.
(76.) Magro CM, Crowson AN. Drug-induced immune- dysregulation as a cause of atypical cutaneous lymphoid infiltrates: a hypothesis. Hum Pathol. 1996;26: 125-132.
(77.) Dellabona P, Peccoud J, Kappler J, Marrack P, Benoist C, Mathis D. Superantigens interact with MHC class II molecules outside of the antigen groove. Cell. 1990;62:1115-1121.
(78.) Mollick JA, Miller GG, Musser JM, Cook RG, Grossman D, Rich RR. A novel superantigen isolated from pathogenic strains of Streptococcus pyogenes with aminoterminal homology to staphylococcal enterotoxins B and C. J Clin Invest. 1993;92:710-719.
(79.) Reda KB, Kapur V, Mollick JA, Lamphear JG, Musser JM, Rich RR. Molecular characterization and phylogenetic distribution of the streptococcal superantigen gene (ssa) from Streptococcus pyogenes. Infect Immun. 1994;62:1867-1874.
(80.) Drake CG, Kotzin BL. Superantigens: biology, immunology, and potential role in disease. J Clin Immunol. 1992;12:149-162.
(81.) Ohmen JD, Barnes PF, Grisso CL, et al. The T cell receptors of human gamma delta T cells reactive to Mycobacterium tuberculosis are encoded by specific V genes but diverse V-J junctions: M tuberculosis contains a superantigen [abstract]. J Immunol. 1993;150:94A.
(82.) Abe J, Takeda T, Watanabe Y, et al. Evidence for superantigen production by Yersinia pseudotuberculosis. J Immunol. 1993;151:4183-4188.
(83.) Cole BC, Atkin CL. The mycoplasma arthritidis T-cell mitogen, MAM: a model superantigen. Immunol Today. 1991;12:271-276.
(84.) Crowson AN, Magro CM. Atypical pigmentary purpura: a clinical, pathological and genotypic study of 40 cases. Hum Pathol. 1999;30:1104-1112.
(85.) Magro CM, Crowson AN. Drug induced immune-dysregulation as a cause of atypical lymphoid infiltrates. Hum Pathol. 1996;27:125-132.
A. Neil Crowson, MD; Cynthia M. Magro, MD; Martin C. Mihm, Jr, MD
Accepted for publication October 9, 2007.
From the Departments of Dermatology, Pathology and Surgery, University of Oklahoma and Regional Medical Laboratories, St John Medical Center, Tulsa, Okla (Dr Crowson); the Department of Pathology and Laboratory Medicine, Weill Medical College, Cornell University, New York, NY (Dr Magro); and the Departments of Dermatology and Pathology, Harvard Medical School, Massachusetts General Hospital, Boston (Dr Mihm).
The authors have no relevant financial interest in the products or companies described in this article.
Presented in part at the 28th Annual Course, Current Concepts in Surgical Pathology, Massachusetts General Hospital and Harvard Medical School, Boston, November 2006.
Reprints: A. Neil Crowson, MD, Regional Medical Laboratories, St John Medical Center, 1 923 S Utica Ave, Tulsa, OK 74104 (e-mail: email@example.com).
Table 1. Prototypes of Cell-Poor Interface Dermatitis Erythema multiforme Autoimmune connective tissue disease, particularly Systemic lupus erythematosus Dermatomyositis Mixed connective tissue disease Graft-versus-host disease Morbiliform viral exanthem Morbiliform drug reaction Table 2. Prototypes of Cell-Rich Interface Dermatitis Idiopathic lichenoid disorders Lichen planus Lichen nitidus Lichen striatus Lichenoid autoimmune connective tissue disease, particularly Discoid lupus erythematosus Anti-RO-positive systemic lupus erythematosus Mixed connective tissue disease Lichenoid and granulomatous dermatitis Lichenoid purpura Lichenoid and fixed drug reaction Table 3. Lupus Erythematosus: Comparison of Clinical and Direct Immunofluorescence Manifestations by Subtype * Feature SLE SCLE DLE Lesion character Scale Fine, easily Fine, easily Thick, detached detached adherent Follicular atrophy Absent Absent Present Photodistribution Present Marked Present Scarring Absent Absent Present Atrophy Usually absent Usually absent Present, often marked Pigmentary Slight Slight Often marked alteration Telangiectasia Present Present Present [greater than or 100% 40% 10% equal to] 4 ACR criteria for SLE (incidence) Positive lupus band >90% 60% 90% test Lesional skin Active 30% 0-10% disease 90% Nonlesional Inactive sun-protected skin disease 30% * SLE indicates systemic lupus erythematosus; SCLE, subacute cutaneous lupus erythematosus; DLE, discoid lupus erythematosus; and ACR, American College of Rheumatology. After Crowson and Magro. (17) Table 4. Seropositivity by Connective Tissue Disease Classification * ANA C ssDNA dsDNA Sm SLE 95-100 0-10 70 70 <30 ([dagger]) SCLE 70 0 10-20 0 0 DLE 20-30 10-20 10-20 0 CRST 90 90 0 0 0 PSS 70 10 10-20 0 0 MCTD 95-100 0 30 0 0-5 SSA/Ro SSB/La nRNP Histone Scl 70 SLE 40 10-20 40-50 30 0 SCLE 60-80 10 0 0 0 DLE 10-20 10 0 0 0 CRST 0 0 0 0 0 PSS 10 10 20 0 30 MCTD 0 0 100 0 0 * Seropositivity is expressed as percentage of cases. ANA indicates antinuclear antibody; C, complement; ssDNA, single/stranded DNA; dsDNA, double stranded DNA; Sm, Smith; nRNP, nuclear ribonucleoprotein; SLE, systemic lupus erythematosus; SCLE, subacute cutaneous lupus erythe/matosus; DLE, discoid lupus erythematosus; CRST, CREST syndrome; PSS, progressive systemic sclerosis/ scleroderma; and MCTD, mixed connective tissue disease. After Crowson and Magro. (17) ([dagger]) Dependent upon ethnicity. Table 5. Grades of Graft-Versus-Host Disease Grade 0 No pathologic change Grade 1 Basal vacuolization Grade 2 Basal vacuolization, keratinocyte necroses, and dermal inflammation Grade 3 Confluence of basal vacuoles Grade 4 Separation of epidermis from dermis
|Gale Copyright:||Copyright 2008 Gale, Cengage Learning. All rights reserved.|