Immunopathology of pityriasis lichenoides acuta

I

I

II

II II II

'

I

9

I

I

I

Immunopathology of pityriasis lichenoides acuta Jan E. Muhlbauer, M.D., Atul K. Bhan, M.D., Terence J. Harrist, M.D., Richard A. Moscicki, M.D., Rhonda Rand, M.D., Wright Caughman, M.D., Bob Loss, M.D., and Martin C. Mihm, Jr., M.D.

Boston, MA Eleven biopsy specimens (five papules and six dusky or crested lesions) from four patients with pityriasis lichenoides et varioliformis acuta (PLEVA) were studied by direct immunofluoreseence and immunoperoxidase technics. Slight vascular deposits of IgM and C3 were present in most lesions. Slight perivascular deposits of fibrin were observed in early lesions; more extensive perivascular and interstitial deposits of fibrin were detected in advanced lesions. Most of the infiltrating cells were T lymphocytes; cells with cytotoxic/suppressor phenotype (T8-positive) were generally more numerous than ceils with helper/inducer phenotype (Leu-3a-positive, T4-positive). A marked increase in epidermal T8-positive ceils over epidermal Leu-3a/T4-positive cells was found in late lesions. Moreover, a reduction of the ratio of circulating T4-positive to TS-positive cells was observed in most cases. The number of epidermal T6-positive (Langerhans/indeterminate) ceils was decreased in the lower as compared with the upper stratum spinosum. About 5% of perivascular infiltrating cells were T6-positive. These results suggest that cell-mediated immune mechanisms are probably important in the pathogenesis of PLEVA. (J AM ACADDErMA'teL 10:783-795, 1984.)

Pityriasis lichenoides acuta (pityriasis lichenoides et varioliformis acuta, PLEVA, Mucha-Habermann disease) is an idiopathic eruption characterized by papules that may become hemorrhagic or crusted and resolve as varioliform scars.t,2 Lesions appear in crops that subside in several weeks, but often recur over a period of months or years. A hypersensitivity reaction to an infectious agent has been suggested by the occurrence of small outbreaks ~ and by an association with preFrom the Departments of Pathology and Dermatology, Massachusetts General Hospital, Harvard Medical School. Supported in part by Grants CA 29601 and HL 18646 from the National Institutes of Health. Accepted for publication Oct. 18, 1983. Reprint requests to: Dr. Atul K. Bhan, Immunopatbology Unit, Cox 5, Massachusetts General Hospital, Boston, MA 02114.

vious upper respiratory tract infections, z toxoplasmosis, 4 and viral infections. 5 Because circulating immune complexes as well as vascular and epidermal basement membrane zone deposits of IgM or C3 have been observed in some patients, PLEVA is assumed by some to be immune complex-mediatedG'7; others, however, could not confirm the direct immunofluorescence (DIF) findings. s-~1 The routine histologic features include a superficial perivascular mononuclear cell infiltrate, endothelial cell swelling with invasion of vessel walls by mononuclear cells and erythrocytes, and the presence of erythrocytes and lymphoid cells within the epidermis. 1't2'13 Using an immunoperoxidase technic, lesions in different stages of evolution were studied with monoclonal antibodies that react with different lymphocyte 783

Journal of the American Academy of Dermatology

Muhlbauer et al

784

Table I. Clinical data on patients with PLEVA No.

Age

Sex

Duration

1

19

M

5 yr

2

23

M

9 mo

3

29

F

6 mo

4

64

M

10 mo

Type of lesion and site

Age of lesion

Study performed

Papule, leg Papule, leg Papule, elbow Papule, back Dusky plaque, flank Crusted papule, flank Papule, ann Crusted papule, arm Papule, axilla Dusky papule, forearm Crusted papule, neck Papule, arm Crusted papule, arm Papule, right wrist Papule, forearm Dusky papule, finger

5-7 days 5 days 2-3 days 5 days 7 days 14 days Early Late 3-4 days 7-8 days Late 7 days 14 days 7 days 3 days 14 days

Routine Routine DIF/IP DIF/IP DIP/IP DIP/IP Routine Routine DIF/IP DIF/IP DIF/IP Routine/DIF/IP Routine/DIF/IP Routine DIF/1P DIF/IP

DIF: Directimmunofluorescence;IP: immunoperoxidase. subsets, monocytes, and Langerhans cells to characterize the infiltrating mononuclear ceils in PLEVA. DIF was used to detect immunoglobulin, complement, and fibrin deposits. In addition, correlative studies were performed examining the distribution of T cell subsets among peripheral blood lymphocytes of these patients, utilizing fluorescence-activated flow cytometry. MATERIALS A N D M E T H O D S Patients

Four patients with recurrent eruption of pink, scaly, hemorrhagic and crusted papules, which healed with varioliform scarring and with biopsy specimens interpreted as typical for PLEVA, t'~z'ta were prospectively studied" (Table I). Permission was obtained from each patient to perform additional punch biopsy procedures. One individual (Patient 2), had upper respiratory tract symptoms and fever at the onset of eruption. In Patient 1, necrotic lesions were large (up to 1.5 crn diameter) and resolved as deep varioliform scars. Patient 4 had adult-onset diabetes mellitus for many years before his rash developed. A VDRL test result was nonreactive in Patients 1, 2, and 4; a serologic test for syphilis was not performed in Patient 3. An enzyme-linked immunosorbent assay for Toxoplasma gondii IgG (M. A. Bioproducts, Walkersville, MD) was negative in the serum of Patients 2 and 4 and was not performed in Patient 3. Mid-range positive test results (0.325 and

0.284) were obtained on separate occasions in Patient 1, consistent with previous toxoplasmosis infection at an indeterminate time. Skin specimens. No therapy was used during 4 weeks prior to biopsy, except Patient 3 had applied emollients and topical steroids to lesions on the legs (biopsies were performed only on lesions on the arms), and Patient 4 had been taking digoxin, propranolol hydrochloride, indomethacin, and chlordiazepoxide. One percent lidocaine (Xylocaine), with or without epinephrine, was injected around the 2- to 14-day-old lesions from which a 4-ram punch biopsy was obtained (Table I). The specimens were fixed in formalin ~br routine study, or frozen in OCT Compound (Ames Co., Division of Miles Laboratory, Inc., Elkhart, IN) for DIF and immunoperoxidase studies (Table I). DIF procedure. The methods for the DIF procedure have been described previously.~4 In brief, frozen sections, 4/~m thick, were air-dried and then incubated alone in a moist chamber at room temperature with fluoresceinated antihuman IgG, IgM, IgA, albumin, fibrin, and C3 (N. L. Cappel Laboratories, Cochranville, PA). The sections were washed in phosphatebuffered saline solution (PBS) mounted in Elvanol (E. I. DuPont de Nemours & Co., Inc., Wilmington, DE), and viewed through a fluorescence microscope with epi-illumination. Intensity and quantity of fluorescence were graded by two observers, utilizing a scale of 1+ (slight but definitely present), 2+ (moderate), and 3+ (marked).

Volume I0 Number 5, Part 1 May, 1984

lmmunopathology of pityriasb lichenoides acuta 785

T a b l e I I . DIF in P L E V A Lesion Location

Epidermis or stratum corneum (diffuse deposition)

Epidermal basement membrane zone (linear) Epidermal basement membrane zone (colloids) Vascular/perivascular

Dermis (interstitial deposition)

Papule (~

Dusky/crusted (63

IgG IgM IgA C3 Fibrin Albumin Fibrin

1 1 1 1 2 2 2

5 4 4 3 3 3 4

IgM

3

3

!gM* C3 Fbrint IgG IgA Fibrin Albumin

4 4 5 0 0 1 2

4 6 6 3 3 6 3

Immunoreactant

*lgM deposition was confinedto one vessel in each of three papules and three dusky/crusted lesions. tFibrin deposition was slight (1 +) in papules, but marked (3+) in dusky/crusted lesions. Monoclonai antibodies. A series of monoclonal antibodies reactive with lymphocyte subsets, monocytes, and Langerhans cells were used. 15-2~ AntiLeu-1 antibody (Becton, Dickinson & Co., Sunnyvale, CA) reacts with approximately 100% of peripheral E-rosette-positive T cells. Anti-T3 antibody (OKT3, Orthoclone antibodies; Ortho Diagnostic Systems Inc., Raritan, NJ) reacts with all mature peripheral blood T cells. 2~ Anti-T6 antibody, which reacts with a majority of cortical thymocytes but not with peripheral T cells, ~5 also reacts with Langerhans/indeterrninate cells, l~-la Anti-Leu-3a (Becton, Dickinson) and anti-T4 antibodies 2~ react with T cell antigens associated with helper/inducer function. Anti-T8 antibody reacts with an antigenic determinant on T cells associated with cytotoxic/suppressor function.21 Anti-I 1 antibody reacts with the nonpolymorphic regions of human h-like (HLA-DR) antigens and labels the vast majority of B cells and monocytes, 20% of null cells, 30% - 40% of activated T cells, and virtually no resting T cells. 22AntiM 1 antibody reacts with monocytes/macrophages, neuIrophils, and natural killer cells. 2~ A monoclonal antiIgM antibody 2~ reactive with B cells and an anti-B1 antibody that reacts with all peripheral B cells except plasma cells24 were also used. Immunoperoxidase procedure. The frozen sec-

tions, 4 / , m thick, were air-dried, fixed in acetone for 5 minutes, and stained by a 3-step avidin-biotin technic (Vector Laboratories, Burlingame, CA), 25 Sections were first incubated with a 1 : 100 to 1 : 500 dilution of mouse ascites, containing the monoclonal antibody, for 60 minutes at room temperature, succeeded by 30minute incubations with biotin-labeled horse antimouse Ig and avidin-biotin-peroxidase complex. Between each incubation, sections were washed in three changes of PB S. Staining was achieved by incubation of sections in an acetate buffer solution (pH 5.0) that contained 3amino-9-ethylcarbazole (Aldrich Chemical Co., Milwaukee, WI), dimethylformamide, and hydrogen peroxide. The sections were washed in acetate buffer, mounted in Elvanol, and counterstained with Gill's triple-strength hematoxylin. Controls included sections incubated in the first step with PBS or nonspecific ascitic fluid instead of monoclonal antibodies. The potency and reactivity of the monoclonal antibodies were monitored by frequent testing on lymphoid tissue, including thymus and tonsils, ls,ls,~4 Counting methods. Epidermal dendritic cells were counted on sections stained with anti-T6 antibody as described previousiy, zG Epidermal cell counts were simiIarly performed on sections stained with anti-T

786

Journal of the American Academy of Dermatology

Muhlbauer et al

Fig. 1. Frozen tissue section of a crested lesion stained with fluoresceinated antihuman fibrin. Deposits are observed in perivascular/vascular (short arrows) and interstitial (long arrows) locations. (Bar, 25/zm; original magnification, x400.) cell antibodies. Counts of epidermal T6-positive cells on each section were subdivided by position of cells in the center (that portion overlying the most dense area of papillary dermal mononuclear cell infiltrate) or the periphery, and by position of cells in the upper or lower portion of the epidermis. The number of T4-positive, Leu-3a-positive, and T8-positive cells were counted within perivascular areas of at least five corresponding high-power fields on serial sections (