Deposition of membrane attack complex in cutaneous lesions of lupus erythematosus

Clinical and laboratorv studies Deposition of membrane attack complex in cutaneous lesions of lupus erythematosus Klaus F. Helm, MD,* and Margot S. Peters, MD Rochester, Minnesota

Background: The membrane attack complex (MAC; C5b-9) localizes in the basement membrane zone (BMZ) of cutaneous lesions from patients with lupus erythematosus (LE) and has been implicated in the pathogenesis of this disease. Objective: Our purpose was to compare the frequency of MAC deposition with that of immunoglobulin and C3 deposition (lupus band). Methods: We studied 56 skin biopsy specimens from 42 patients with LE by direct immunofluorescence. Results: MAC was deposited in a granular pattern at the BMZ in 29 of 38 biopsy specimens (76%) from lesional skin; 5 of 38 specimens (13%) had focal or weak segmental deposition and 4 of 38 (11 %) were negative. In contrast, IgG, IgM, IgA, and C3 were detected in 16 of38 specimens (42%),30 of38 (79%), 40f38 (11%), and 22 of38 (58%), respectively. None of the uninvolved skin biopsy specimens was MAC positive, although 4 of 18 (22%), 5 of 18 (28%), 1 of 18 (6%), and 2 of 18 (11 %) were positive for IgG, IgM, IgA, and C3, respectively. Conclusion: MAC deposition in lesional skin appears to be a relatively sensitive and specific marker for cutaneous LE and may be useful as an adjunct to the "lupus band" test. (J AM ACAD DERMATOL 1993;28:687-91.)

The membrane attack complex (MAC) is formed by the activation of either the alternative or classical complement cascade through noncovalent binding of the complement proteins C5b through C9. The resultant complex may lyse cells by producing holes in the cell membrane, although the MAC has been detected by immunofluorescent techniques in lesional skin of several immunologically mediated skin disorders such as lupus erythematosus (LE), 1 bullous pemphigoid.i pemphigus.' dermatitis herpetiformis," and Henoch-Schonlein purpura.' Biesecker et al.' showed that deposition of MAC occurred only in lesional skin from patients with discoid or systemic LE. MAC deposition was not seen in any of the biopsy specimens from skin that appeared to be normal from the patients with LE. We examined the sensitivity and specificity of MAC deposition in skin biopsy specimens from both From the Immunodermatology Laboratory, Mayo Clinic and Mayo Foundation. Accepted for publication Nov. 21, 1992. Reprint requests: M. S. Peters, MD, Department of Dermatology, Mayo Clinic, 200 First St. SW, Rochester. MN 55905. "'Present address: Division of Dermatology, Department of Medicine, Hershey Medical Center, Hershey, Pa. Copyright @ 1993 by the American Academy of Dermatology, Inc.

0190-9622/93 $1.00 +.10 16/1/44462

lesional and uninvolved (normal-appearing) skin in patients with LE and control subjects.

METHODS We examined 56skin biopsy specimens from the tissue bank (stored at -70 0 C) in our Immunodermatology Laboratory. The specimens were originally collected from 42 patients with LE: 12 patients with systemic LE; 22 with discoidLE; 6with subacutecutaneous LE; 1with LE and rheumatoid arthritis; and 1 with a collagen vascular disorder suggestive of LE. Fourteen patients had two biopsy specimens (II patients had specimens taken from both involved and uninvolved sites, 2 patients had two biopsy specimens from involved sites, and 1 patient had two biopsy specimens from uninvolved skin). There were 11 control specimens: 3 from patients with pemphigoid, 4 with urticaria or urticarial vasculitis, 2 with eczema, 1 with relapsing polychondritis, and 1 with a skin ulcer. All biopsy specimens had been snap-frozen in liquid nitrogen immediately after collection, and. none was placed in transport medium. MAC deposition was detected by an indirect immunofluorescent technique with a fluorescein-conjugated monospecific rabbit anti-human MAC antibody at a dilution of 1:40 (Calbiochem, La Jolla, Calif.), and goatanti-rabbit secondary antibody (Sanofi Diagnostics, Chaska, Minn.) at a dilution of 1:40. The MAC was prepared by Calbiochem according to the methods of Bhakdi et aI.6 and consisted of a polyclonal antibody against the

687

Journal of the American Academy of Dermatology May 1993

688 Helm and Peters

Fig. 1. A and B, Thick, granular deposition of membrane attack complex along the dermoepidermal junction of lesional skin in a patient with discoid lupus erythematosus.

(x400.)

Table I. Detection of immunoreactants in lesional skin from patients with lupus" Type of deposition Granular Immunoreactant

No.

MAC IgG IgM IgA C3

29 16

30 5

22

J

Focal or weak

%

76 42

79 13

58

No.

I

5

7

5 5 8

Negative

%

No.

13 18 13 13 21

4 15 3

29 8

I

%

11

39 8

76 21

"Thirty-eight biopsy specimens were obtained from 34 patients: 19 with discoid, 7 with systemic, and 6 with subacute cutaneous lupus erythematosus, and 2 with an overlap syndrome.

C5b-C9 complex that had been preabsorbed to each of the individual components. The sensitivity of MAC deposition in lesional skin was calculated by dividing the number of immunofluorescence-positive lesional skin biopsy specimens by the total number oflesional skin biopsy specimens from patients with LE. Specificity was calculated by dividing the number of uninvolved skin biopsy specimens negative for MAC by the number of uninvolved skin biopsy specimens from patients with LE. Specimens were examined blindly with respect to disease and skin involvement and were graded as negative if no staining was present along the basement membrane zone (BMZ), as weak focal or segmental staining if the intensity of fluorescence was low and the deposits were discontinuous along the BMZ, and as positive if the staining was a bright granular continuous band.

RESULTS MAC deposition in lesional skin from patients with LE was present as a continuous granular pattern along the BMZ in 29 of 38 specimens (76%) (Fig. 1); deposition was weak focal or segmental in 5 of 38 (13%) and negative in 4 of 38 specimens (11 %). As seen in Table I, MAC deposition occurred more frequently than all of the other immunoreactants except for IgM. No correlation between the deposition of MAC and the presence or absence of other immunoglobulins was seen. Of the different subtypes, MAC was detected most frequently in lesional skin from patients with discoid LE (18 of 22, 82%). This was in contrast with MAC deposition in six of eight patients (75%) with systemic LE, three

Journal of the American Academy of Dermatology Volume 28, Number 5, Part I

Helm and Peters 689

Table II. Detection of imrnunoreactants in uninvolved skin from patients with lupus* Type of deposition Positive Immunoreactanl

No.

o

MAC IgG IgM IgA C3

4 5 1 2

I

Weak or focal

%

o

22

28 6 11

No.

I

4

Negative

%

No.

22

14 14 8 17 14

o 28 o

o 5 o 2

11

I

%

78 78

44

94 78

*Eighteen biopsy specimens were obtained from 17 patients: 4 with discoid, 7 with systemic, and 5 with subacute cutaneous lupus erythematosus, and 1 with an overlap syndrome.

Table III. Sensitivity and specificity of immunoreactants for lesional skin including weak or focally positive values

Table IV. Sensitivity and specificity of immunoreactants for lesional skin excluding weak or focally positive values

Immanoreactant

SellSitivity* (%)

Specfficlty] ( %)

Immunoreactant

Sensitivity* (%)

Speciftcityt (%)

MAC IgG IgM IgA C3

89 61 92

78 78

MAC IgG IgM IgA C3

76 42 79

100 78 72

58

88

26 79

44 94 78

13

94

*Sensitivity was calculated by dividing the number of immunofluorescence-positive lesional skin biopsy specimens by the total number of lesional skin biopsy specimens from patients with lupus erythematosus. [Speciflcity was calculated by dividing the number of uninvolved skin biopsy specimens negative for MAC by the total number of uninvolved skin biopsy specimens from patients with lupus erythematosus.

*Sensitivity was calculated by dividing the number of immunofluorescence-positive lesional skin biopsy specimens by the total number of lesional skin biopsy specimens from patients with lupus erythematosus. tSpecificity was calculated by dividing the number of uninvolved skin biopsy specimens negative for MAC by the total number of uninvolved skin biopsy specimens from patients with lupus erythematosus.

of six (50%) with subacute cutaneous LE, and one of two (50%) with overlap syndrome. Conversely, no specimen (0 of 18) of uninvolved skin from a patient with LE was positive for MAC, whereas these specimens were frequently positive for other immunereactants (Table II). Staining of blood vessels with MAC was seen in 7 of 38 biopsy specimens from Iesional skin, 1 of 18 from uninvolved skin, and 2 of 11 from controls. The values for sensitivity and specificity of MAC deposition depended on the inclusion or exclusion of the weak or focally positive specimens with either the true positiveor negative values. As Tables III and IV show, values ranged from 76% to 94% for sensitivity and 78% to 100% for specificity. The sensitivity of MAC deposition increased by including the weak values with the positive ones, but the specificity decreased from 100% to 78%. These data compare favorably with those derived with other immunoreac-

tants; for example, the specificity of IgA deposition at 94% is high but the sensitivity is low, and, conversely, IgM deposition is highly sensitive but not specific (Table III). DISCUSSION

Burnham et al.7 first described the deposition of immunoglobulin at the epidermal BMZ of involved skin in patients with LE. The presence of immunoreactants along the dermoepidermaljunction of uninvolved (normal-appearing) skin in patients with LE has since become known as a positive "lupus band." The mechanism of formation of the lupus band is not known. Various hypotheses include suggestions that the band is formed by antibodies against both nuclear and BMZ components''; that some antinuclear antibodies may have an affinity for the dermoepiderma1 junction's 10 and that this affinity may be charge-dependent!"; and finally, that

690

Helm and Peters

nuclear constituents released during epidermal differentiation may bind to the BMZ and subsequently lead to localization of antinuclear antibodies. 12 The specificity of the lupus band-defined as the presence of granular deposition ofimmunoglobulins along the derrnoepidermal junction-has been questioned. Immunoglobulins along the BMZ can be seen in a variety of dermatoses, including rosacea, and in sun-damaged and telangiectatic skin.13 Fabre et a1. 14 detected immunoreactants in normal-appearing skin of 20% of healthy adults. Gruschwitz et al. 15found immunoreactants in 88% of lesional skin of actinic keratoses and in 80% of nonlesional skin. Williams et a1. 16 found that direct immunofluorescence was most helpful in evaluating cases of possible systemic LE. However, testing was only supportive in the investigation of chronic discoid LE because test results in these patients only rarely contained IgG and the presence of IgM was considered nonspecific. Our study showed that the deposition of MAC can be used to eliminate false-positive lupus band tests because the deposition of MAC was highly specific for lesional skin in patients with LE. As in other studies, wefound that the presence of IgM was sensitive but not specific. 15, 17, 18 The specificity of the lupus band increased if more than one immunoreactant was present; however, the sensitivity decreased. For example, Smith et a1. 18 noted a specificity of 94% (if three or more immunoreactants were present) but a sensitivity of only 47%. The antibody used to detect the MAC is important because French et aI.,19 using a monoclonal antibody against the C9 antigen present in the MAC, found staining in both uninvolved and involved skin from patients with a positive lupus band. The use of the MAC as an adjunctive antibody in the immunofluorescence diagnosis of cutaneous lesions from patients with LE has the distinct advantage of being highly sensitive and specific. The presence of a lupus band in non-sun-exposed nonlesional skin has been associated with a worse prognosis in patients with lupus2Q-22; consequently, staining for MAC cannot be used as a prognostic indicator because it is not detected in nonlesional skin. In addition, although testing for MAC deposition has a diagnostic role, the presence of MAC in a skin biopsy specimen does not necessarily indicate that immune complexes or complement is involved in the pathogenesis of the lesions.

Journal of the American Academy of Dermatology May 1993 REFERENCES 1. Biesecker G, Lavin L, Ziskind M, et al. Cutaneous localization of the membrane attack complex in discoid and systemic lupus erythematosus. N Engl J Moo 1982;306:26470. 2. Dahl MY, Falk RJ, Carpenter R, et al. Deposition of the membrane attack complex of complement in bullous pemphigoid. J Invest Dermatol 1984;82: 132-5. 3. Kawana S, Geoghegan WD, Jordon RE, et al. Deposition of the membrane attack complex of complement in pemphigus vulgaris and pemphigus foliaceus skin. J Invest DermatoI1989;92:588-92. 4. Dahl MY, Falk RJ, Carpenter R, et al. Membrane attack complex of complement in dermatitis herpetiformis. Arch Dermato11985;121 :70-2. 5. Kawana S, Shen GH, Kobayashi Y, et a1. Membrane attack complex of complement in Henoch-Schonlein purpura skin and nephritis. Arch Dermatol Res 1990;282:183-7. 6. Bhakdi S, Muhly M, Roth M. Preparation and isolation of specific antibodies to complement components. Methods EnzymoI1983;93:409-20. 7. Burnham TK, Neblett TR, Fine G. The application of the fluorescent antibody technic to the investigation of lupus erythematosus and various dermatoses. J Invest Dermatol 1963;41:451-6. 8. Landry M, Sams WM Jr. Systemic lupus erythematosus: studies of the antibodies bound to skin. J Clin Invest 1973;52:1871-80. 9. Natali PG, Tan EM. Experimental skin lesions in mice resembling systemic lupus erythematosus. Arthritis Rheum 1973;16:579-89. 10. Izui S, Lambert P-H, Miescher P A. In vitro demonstration of a particular affinity of glomerular basement membrane and collagen for DNA: a possible basis for a local formation of DNA-anti-DNA complexes in systemic lupus erythematosus. J Exp Med 1976;144:428-43. 11. Joselow SA, Mannik M. Localization of preformed, circulating immune complexes in murine skin. J Invest Dermatol 1984;82:335-40. 12. Gilliam IN. DNA binding to normal skin connective tissue as a localizing factor for DNA-anti-DNA complexes [Abstract). Arthritis Rheum 1977;20:117-8. 13. Jablonska S, Chorzelski T, Maciejowska E. The scope and limitations of the immunofluorescence method in the diagnosisoflupus erythematosus. Br J DermatoI1970;83:242-7. 14. Fabre YC, Lear S, Reichlin M, et al. Twenty percent ofbiopsy specimens from sun-exposed skin of normal young adults demonstrate positive immunofluorescence. Arch Dermatol 1991;127:1006-11. 15. Gruschwitz M, Keller J, Hornstein CP. Deposits ofimmunoglobulins at the derma-epidermal junction in chronic light-exposed skin: What is the value of the lupus band test? Clin Exp Dermatol 1988;13:303-8. 16. Williams REA, MacKie RM, O'Keefe R, et al. The contribution of direct immunofluorescence to the diagnosis of lupus erythematosus. J Cutan PathoI1989;16:122-5. 17. Wojnarowska F, Bhogal H, Black MM. The significance of an IgM band at the derma-epidermal junction. J Cutan Pathol J986;13:359-62. 18. Smith CD, Marino C, Rothfield NF. The clinical utility of the lupus band test. Arthritis Rheum 1984;27:382-7. 19. French LE, Polla LL, Tschopp J, et a1. Membrane attack complex (MAC) deposits in skin are not always accompanied by S protein and clusterin. J Invest Dermatol 1992; 98:758-63.

Journal of the American Academyof Dermatology Volume 28, Number 5, Part I 20. Davis BM, Gilliam IN. Prognostic significance of subepidermalimmunedepositsin uninvolved skinof patientswith systemiclupuserythematosus:a 1O-year longitudinal study. J InvestDermatol 1984;83:242-7. 21. GilliamIN, Cheatum DE, Hurd ER, et al. Immunoglobulin in clinically uninvolved skin in systemiclupus erythematosus: association withrenal disease. J Clin Invest 1974; 53:1434-40.

Helm and Peters 22. ProvostIT, Andres G, Maddison PJ, et al. Lupus band test in untreated SLE patients: correlation of immunoglobulin deposition in the skin of the extensor forearm with clinical renal disease and serological abnormalities. J Invest Dermatol 1980;74:407-12.

Clinical and immunologic studies in reticular erythematous mucinosis and Jessner's lymphocytic infiltrate of skin Suzanne W. Braddock, MD,a H. David Kay, PhD,b Diane Maennle, MD,c Thomas L. McDonald, PhD,d Samuel J. Pirruccello, MD,c Aneal Masih, MD,c Lynell W. Klassen, MD,b and Alisa R. Sawka" Omaha, Nebraska

Background: Little is understood about reticular erythematous mucinosis and Jessner's lymphocytic infiltrate of skin. Objective: Our purpose was to define reticular erythematous mucinosis and Jessner's lymphocytic infiltrate of skin further with focus on immunologic studies. Methods: In patients with reticular erythematous mucinosis and Jessner's lymphocytic infiltrate of skin, we measured circulating immune complexes before, during, and after therapy. We examined natural killer cells in a functional assay; we performed direct immunofluorescence and T- and B-cell marker studies in skin biopsy specimens. Results: The infiltrate in reticular erythematous mucinosis is composed of helper T cells. Circulating immune complexes are increased in both reticular erythematous mucinosis and Jessner's lymphocytic infiltrate of skin and decrease with hydroxychloroquine therapy and clinical clearing. Natural killer cell function is decreased in reticular erythematous mucinosis and Jessner's lymphocytic infiltrate of skin. Conclusion: Changes in circulating immune complex titers accompanying therapy with hydroxychloroquine and clinical clearing, with recurrence of the condition and increase in circulating immune complexes on discontinuation of treatment, point to a possible relation between these events. (J AM ACAD DERMATOL 1993;28:691-5.)

Reticular erythematous mucinosis has not been well characterized immunologically. 1, 2 The current study was undertaken to further characterize immune function in this disorder. Five patients with reticular erythematous mucinosis who fulfilled ac-

cepted criteria for the diagnosis of the disorder were studied and compared with three patients with lessncr's lymphocytic infiltrate of skin, a disorder that has certain similarities to reticular erythematous mucinosis.

From theDepartment of Internal Medicine, Section ofDermatology," Section ofRheumatology and Immunology," Department ofPathology"/Microbiology,d University of Nebraska Medical Center. Accepted for publication Sept. 21, 1992. Reprint requests: Suzanne W. Braddock, U.N.M.C., 600 S. 42nd St., Omaha, NE 68198-4360. Copyright@ 1993 bythe American Academy of Dermatology, Inc. 0190-9622/93 $1.00+.10 16/1/42828

CASE REPORTS Case 1. A 65-year-old white woman had an asymptomatic, erythematous plaque measuring 4 em on the posterior mid thorax of 4 years' duration. The patient was in excellent general health. Histologic examination of a biopsy specimen revealed a normal epidermis and a dense perivascular lymphocytic infiltrate in the superficial der-

691