Henoch-Schönlein vasculitis: Direct immunofluorescence study of uninvolved skin

IIIII

I

I

II

IIIII III

II

II

II

Henoch-Sch6nlein vasculitis: Direct immunofluorescence study of uninvolved skin Harriet M. Van Hale, M.D., Lawrence E. Gibson, M.D., and Arnold L. Schroeter, M.D. Rochester, MN Henoch-Sch6nlein purpura is a multisystem disease believed to be a consequence of entrapment of circulating IgA-containing immune complexes in blood vessel walls throughout the skin, kidney, and gastrointestinal tract. In this direct immunofluorescence study, twenty-five skin biopsy specimens from twenty patients with Henoch-Sch6nlein purpura were examined (9 from uninvolved, normal-appearing skin). A distinct stippled pattern of vascular fluorescence was found in 87% of lesion biopsies; 75% of these contained deposits of IgA. In uninvolved skin, seven (78%) showed immunoglobulin in vessel walls and six (67%) contained IgA, suggesting that immune complexes are deposited with equal frequency in normal-appearing and lesional skin of patients with Henoch-Sch6nlein purpura. Biopsy of uninvolved, rather than of purpuric, skin for direct immunofluorescence studies may be more helpful in confirming the diagnosis of Henoch-Sch6nlein purpura because tissue morphology is usually of better quality. (J AM ACAD DERMATOL15:665-670, 1986.)

One of the most characteristic features of Henoch-Sch6nlein purpura, an acute systemic necrotizing vasculitis primarily involving skin, gastrointestinal tract, and joints, has been IgA deposits in glomerular mesangium and vessel walls of skin lesions, detected by direct immunofluorescence microscopy. Direct immunofluorescence results on skin from areas o f active vasculitis are often difficult to interpret because destruction of vessels and fibrin deposition tend to be widespread and immunoreactants are often rapidly dispersed and removed. For this reason, our laboratory requests biopsy specimens from perilesional or normal skin in cases of suspected vasculitis. The objective of the following study was comparison of the incidences of vascular IgA deposits in affected and uninFrom the Department of Dermatology, Mayo Clinic and Mayo Foundation. Accepted for publication May 7, 1986. Reprint requests to: Dr. Lawrence E. Gibson, Mayo Clinic, 200 First St. SW, Rochester, MN 55905.

volved skin of patients with Henoch-Schrnlein purpura. MATERIALS AND METHODS

Of the twenty patients with Henoch-Schrnlein purpura in the study, thirteen were seen at the Mayo Clinic and seven were not but their skin biopsy specimens had been sent directly to our laboratory. Complete clinical and laboratory data were obtained by review of records or from the primary physicians. In all cases, either 3-mm or 4-mm punch biopsy specimens of skin were used for direct immunofluorescence studies (total, 25 specimens: 16 from active skin lesions and 9 from uninvolved skin). Specimens obtained at the Mayo Clinic were immediately frozen in liquid nitrogen; others were mailed in Michel's medium. Frozen sections (4 ~m thick) were stained with fluorescein-conjugated goat antihuman IgG (43.75 Ixg/ml), IgM (36 p,g/ml), IgA (36.3 p~g/ml), C3 (34.2 p~g/ml), and fibrinogen (16 unitage) and examined for fluorescence with a Leitz Ortholax II microscope with epi-illumination. Hematoxylin-eosin-stained sections of biopsy specimens from lesional or perilesional skin of fifteen of 665

666

Van Hale et al

Journal of the American Academy of Dermatology

Fig. 1. Palpable purpuric skin lesions in patient with Henoch-Sch6nlein vasculitis. Fig. 2. Necrotizing vasculitis of dermal vessels. (Hematoxylin-eosin stain; x 60.) Inset, High power view of leukocytoclastic vasculitis and surrounding hemorrhage. (Hematoxylin-eosin stain; x 110.) Fig. 3. Finely stippled fluorescence of IgA in blood vessel walls. ( x 200.) Fig. 4. Broad, continuous, vascular fluorescence of fibrin. (× 60.)

the twenty Henoch-Sch6nlein purpura patients were also reviewed. RESULTS The patients, eleven male and nine female, ranged from 4 to 77 years of age; approximately half were children. All had cutaneous petechiae or purpura, mainly on the lower extremities (Fig. 1). Eleven had symptoms of joint involvement consisting of joint tenderness, pain, or swelling;

this primarily involved the hips, knees, or ankles (Table I). Fifteen patients complained of abdominal pain o r experienced nausea, vomiting, or bloody diarrhea. Symptomatic renal involvement was absent; however, thirteen had gross or microscopic hematuria, proteinuria, or renal casts on urinalysis. With the exception of the urinalysis results, most routine laboratory values were normal. In fifteen cases skin biopsy specimens were

Volume 15 Number 4, Part 1 October, 1986

Henoch-Sch6nlein vasculitis

667

T a b l e I. Clinical and pathologic data on twenty patients with Henoch-Sch6nlein purpura

•

Petechiae/ purpura

1~ tien t ~

Arthritis/ I Gastrointestinal arthralg~ symptoms

1

18/M

+

-

+

2 3 4 5 6 7 8

8/M 17/F 18/F 11/M 40/M 4/M 26/M

+ + + + + + +

+ + + + + +

+ + +

+ + + + + +

9

59/F

+

-

+

+

10 11 12

17/M 77/F 47/F

+ + +

+

+ + +

+ +

13

53/M

+

-

+

14

36/M 27/M

+

-

-

+

-

+

15 16

14/F

+

+

+

+

17 18 19 20

12/F 29/M 9/F 18/F

+ + + +

+ + +

+ + + +

+ + +

Histopathology

Uninvolved

NA NA NA NV NV NV NA Perivascular lymphocytic infiltrates Dermal hemorrhage NV NV Perivascular lymphocytic infiltrates Nonspecific inflammation and hemorrhage NV Perivascular lymphocytic infiltrates NV NV NV NA NV

+A +A +A

+A +A +A

+A +A + + +A

+A +A

+A +A

+

+A +A +A +A +A

DIF: Direct immunofluorescence;NA: not available; NV: necrotizingvasculitis; UA: urinalysis. *A, IgA in dermal vessels.

taken for light microscopy as well as for direct immunofluorescence studies. Ten of these showed necrotizing vasculitis (Fig. 2), and perivascular lymphocytic or mixed inflammatory cell infiltrates were found in the five others. Four of the five patients with nonspecific biopsies had vascular IgA deposits in skin and all had clinical syndromes o f cutaneous purpura and gastrointestinal symptoms consistent with Henoch-Sch6nlein purpura (Table I). Results of direct immunofluorescence are shown in Table II. B a s e m e n t membrane fluorescence was absent in all specimens studied. Deposits of IgG also were absent in vessels of all patients. The most frequent positive findings were vascular deposits o f IgA and fibrinogen. IgA deposits were finely stippled (Fig. 3), whereas fibrinogen was formed in broad, continuous, vascular and perivascular bands (Fig. 4). The incidences of vascular fluorescence with individual immunoreac-

tants did not differ between involved (87%) and uninvolved (78%) skin. In blood vessels, IgA was found in 75% of skin lesions and in 67% of clinically normal-appearing skin (Table I/i). O f nine biopsies from uninvolved skin, two showed negative findings on direct i m m u n o f l u o r e s c e n c e examination; six o f the remaining seven that showed positive findings had IgA deposits in blood vessels. Lesional skin taken f r o m three of these six patients also demonstrated vascular IgA deposits. DISCUSSION

Clinical descriptions of the s y m p t o m c o m p l e x now referred to as H e n o c h - S c h 6 n l e i n purpura first appeared in the mid and late 1800s. ~'2 In 1965, Miescher et aP performed the first direct immunofluorescence studies on skin b i o p s y specimens from two patients with H e n o c h - S c h 6 n l e i n purpura. The presence o f fibrinogen without immunoglobulins and complement led them to speculate that

668

Journal of the American Academy of Dermatology

Van Hale et al

Table II. Direct immunofluorescence results in twenty-five biopsies Positive in blood vessels

Involved skin Uninvolved skin

Number

IgG

IgM

IgA

C3

Fibrin

16 9

0 0

4 3

12 6

8 5

14 5

Table III. Summary of blood vessel fluorescence by direct immunofiuorescence Lesional skin

Nonlesionai skin

14/16 (87%)--positive vessels 12/16 (75%)--positive IgA in vessels

7/9 (78%)--positive vessels 6/9 (67%)--positive IgA in vessels

the mechanism of vascular damage in HenochSch6nlein purpura did not involve immune complex deposition. Eight years later, Baart de la Faille-Kuyper et aP demonstrated vascular deposits of IgA and perivascular fibrin in involved and uninvolved skin of four patients with HenochSch6nlein purpura. IgG, IgM, and complement were also found, but less frequently. This finding was confirmed in other laboratories and soon became one of the criteria for diagnosis of HenochSch6nlein purpura. The finding of vascular immune reactants also supports the concept that the vascuIitis was indeed an immune complex-mediated phenomenon. Since then, little attention has been directed to the clinically normal-appearing skin of patients with this syndrome. Immunoglobulins and complement are found in the skin in various clinical entities, but IgA deposits are not seen frequently. They are characteristic of dermatitis herpetiformis 5 and linear IgA dermatosis, 6 are common in skin of patients with alcoholic liver disease 7 and IgA nephropathy, 4'8 and are occasionally seen in bullous pemphigoid and p e m p h i g u s : Several reports indicate that vascular IgA can be found in skin in chronic glomerular diseases other than IgA nephropathy. 9:° Purpuric and normal-appearing skin from fourteen children with Henoch-Sch6nlein purpura was studied prospectively by direct immunofluorescence in 1977 by Giangiacomo and Tsait~; thirteen biopsy specimens showed IgA deposits in vessels

from areas of purpura, whereas only six showed uninvolved skin to be positive and three of these stained weakly. Baart de la Faille-Kuyper et al ~2 found granular, vascular IgA in nonlesional biopsy specimens of uninvolved skin of twelve patients with Henoch-Sch6nlein purpura. These twelve, however, were taken from a direct immunofluorescence study of 262 patients with various nephropathies and therefore may show a bias toward patients with renal involvement. In our study, vascular IgA was found in six of nine specimens (67%) of uninvolved skin. Biopsy specimens of skin lesions of three of these six also showed IgA in vessels. Six of these nine patients also had findings on urinalysis (erythrocytes, protein, and urinary casts) indicative of renal involvement; however, there was no correlation between positive urinalysis and IgA deposits in normal skin. The presence of IgA and fibrinogen with the relative absence of other immunoreactants makes Henoch-Sch6nlein purpura different from other forms of cutaneous necrotizing vasculitis that typically have vascular deposits of IgG, IgM, and early complement components. C3, properdin, and factor B have also been demonstrated in skin and kidney without evidence of earlier complement components, thus suggesting that complement is activated by the alternate pathway in HenochSch6nlein vasculitis. 4'1a:4 These facts may be related to the pathophysiology of the syndrome. One half to two thirds of patients with Henoch-Sch6nlein purpura have had upper respiratory infection or streptococcal throat infections days to weeks prior to the onset of purpura. 15Exposure to certain foods and drugs has also been related to its onset. ~5 In our series of cases, five patients had sore throats (with cultures positive for group A ~-hemolytic streptococcus in three), one had an upper respiratory infection, and one had a flulike illness. Two

Volume 15 Number 4, Part 1 October, 1986

patients had urine cultures positive for Escherichia coli, and two reported exposure to possible toxins immediately before skin lesions developed. The presence of IgA deposits in various tissues, in addition to increased serum IgA 16and circulating IgA immune complexes, 17 may indicate that this is an exaggerated response to antigens entering the body via mucous membranes. In disorders in which IgA has been found in skin in association with liver disease, the pathophysiology has also been related to alcohol or immunologic injury to intestinal mucosa, 7 as is possible in dermatitis herpetiformis. Regarding the relationship of Henoch-Sch6nlein purpura to IgA nephropathy, it has been suggested that IgA nephropathy simply may be a monosymptomatic form of Henoch-Sch6nlein purpura. ~2 Like cutaneous vessels in Henoch-Sch6nlein purpura, the mesangial lesions in both syndromes contain predominantly IgA, C3, and fibrinogen without early complement components of the "classic" pathway. 4'13'14Also, stippled IgA deposits in cutaneous vessels of apparently normal skin have been found often in IgA nephropathy, s't2 This indicates that widespread tissue deposition of IgA or IgA immune complexes or both occurs in both syndromes. The factors involved in induction of actual clinical lesions, however, remain unknown. In our study, immunglobulins and complement components were found with approximately equal frequency in both purpuric (87%) and uninvolved (78 %) skin in patients with Henoch-Sch6nlein purpura. There also appears to be little difference in IgA deposits at these sites (75% in affected skin and 67% in uninvolved skin). Five patients had direct immunofluorescence studies on both lesional and nonlesional skin, and three of these showed IgA deposits in both. These findings indicate that biopsy of clinically normal skin in patients with Henoch-Sch6nlein purpura is equally diagnostic for this syndrome by direct immunofluorescence. Direct immunofluorescence studies of cutaneous vasculitic lesions older than 48 hours are frequently negative. This is because of destruction and removal of antibodies in the affected necrotic blood vessels and has been demonstrated experimentally by Gower et al. t8 For this reason we believe that the use of uninvolved skin or skin near lesions in Henoch-Sch6nlein purpura and

Henoch-Sch6nlein vasculitis

669

other forms of vasculitis may prove more helpful because the blood vessels, and hence the immunoreactants, are more likely to be intact. A prospective study should clarify this point. The overall sensitivity of direct immunofluorescence for vasculitis in Heuoch-Schfnlein purpura was 90% in our study, slightly higher than has been found for other vasculitides (81%). ~9This study emphasizes the widespread nature of IgA immune deposits in skin in Henoch-Sch6nlein purpura, which parallels similar involvement in other organ systems. REFERENCES

1. Schoenlein JL. Allgemeine und specielle Pathologic und Therapie. Nach dessen Voflesungen niedergeschrieben und hrsg. yon einigen seiner Zuh0rer. Herisau, 1837, Lit.-Compt., vol. 2, p. 48. 2. Henoch E: Ueber eine eigenthtimliche Form yon Purpura. Bed klin Wchnschr 11:641-643, 1874. 3. Miescher PA, Paronetto F, Koffler D: Immunofluorescent studies in human vasculitis, in Grabar P, Miescher PA, editors: Immunopathology: IVth International Symposium. New York, 1965, Gmne & Stratton Inc., pp. 446456. 4. Baart de la Faille-Kuyper EH, Kater L, Kooiker C J, Dorhout Mees El: IgA-deposits in cutaneous blood-vessel walls and mesangium in Henoch-SchSnlein syndrome. Lancet 1:892-893, 1973. (Letter to Editor.) 5. Lever WF, Schaumburg-Lever G: Histopathology of the skin, ed. 6. Philadelphia, 1983, J. B, Lippincott Co. 6. McGuire J, Nordlund J: Bullotls disease of childhood. Arch Dermatol 108:284, 1973. 7. Swerdlow MA, Chowdhury LN, Mishra V, Kavin H: IgA deposits in skin in alcoholic liver disease. J AM ACaD DERMATOL9:232-236, 1983. 8. Lee S, Nevins TE, Michael AF: Diagnostic significance of skin biopsies in IgA nephropathy. Kidney Int 12:515, 1977. (Abst.) 9. Hirbec G, Belghiti D, Wechsler J, et al: Immunofluorescence study of skin biopsy specimens from patients with chronic glomerular diseases. Kidney Int 12:374, 1977. (Abst.) 10. Thompson AJ, Chan Y-L, Woodroffe AJ, et al: Vascular IgA deposits in clinically normal skin of patients with renal disease. Pathology 12:407-413, 1980. 11. Giangiacomo J, Tsai CC: Dermal and glomerular deposition oflgA in anaphylactoid purpura. Am J Dis Child 131:981-983, 1977. 12. Baart de la Faille-Kuyper EH, Kater L, Kuijten RH, et al: Occurrence of vascular IgA deposits in clinically normal skin of patients with renal disease. Kidney Int 9:424-429, 1976. 13. Baart de la Faille-Kuyper EH, van der Meer JB, Kater L, Mul N: Alternate pathway complement activation by IgA in SehSnlein-Henoch syndrome. Neth J Med 17:511, 1974.

670

Journal of the American Academy of Dermatology

Van Hale et al

14. Tsai CC, Giangiacomo J, ZucknerJ: Dermal IgA deposits in Henoch-Sch6nlein purpura and Berger's nephritis. Lancet 1:342-343, 1975. 15. Allen DM, Diamond LK, Howell DA: Anaphylactoid purpura in children (Sch6nlein-Henoch syndrome): Review with a follow-up of the renal complications. Am J Dis Child 99:833-854, 1960, 16. Trygstad CW, Stiehm ER: Elevated serum IgA globulin in anaphylactoid purpura. Pediatrics 47:1023-1028, 1971.

17. Levinsky RJ, Barratt TM: IgA immune complexes in Henoch-Sch6nlein purpura. Lancet 2:1 t 00-1103, 1979. 18. Gower RG, Sams WM Jr, Thome EG, et al: Leukocytoclastie vasculitis: Sequential appearance of immunoreactants and cellular changes in serial biopsies. J Invest Dermatol 69:477-484, 1977. 19. Schroeter AL, Powell FC: Cutaneous vascular immunofluorescence: A sensitivity and specificity study, in MacDonald DM, editor: Immunodermatology. London, 1984, Butterworth Publishing Co., pp. 219-223.

ABSTRACTS

Icotidine, an antagonist of histamine at both H1 and H2 receptors

Food allergy or intolerance in severe recurrent aphthous ulceration of the mouth

Ganellin CR, Balkemore RC, Brown TH, et al: Allergy Proc 7:126-133, 1986 (New England and Regional Allergy Proceedings)

Wright A, Ryan FP, Willingham SE, et al: Br Med J 292:1237-1238, 1986

lcotidine is a single chemical with the ability to antagonize histamine at both H1 and I-I2 receptors. J. Graham Smith, Jr., M.D.

Hepatitis B serum markers in porphyria cutanea tarda Valls V, Enriquez de Salamanca R, Lapena L, et al: J Dermatol 24:24-29, 1986 Hepatitis B virus (HBV) infection markers were studied in 100 Patients from Spain with porphyria cutanea tarda. The overall prevalence of HBV serologic markers was 57%, with 40 patients having antiHBs, 9 antiHBc, 6 HBsAg, and 2 antiHBc and antiHBe. Seropositivity was significantly related to age and to previous venesection of these patients, with 68% of the 47 who had previously had phlebotomy and 47% of the 53 nonphlebotomized patients having HBV markers (p < 0.05). J. Graham Smith, Jr., M.D.

Standards and guidelines for cardiopulmonary resuscitation (CPR) and emergency cardiac care (ECC) Montgomery WH, McIntyre KM, Atkins JM, et al: J A M A 255:2905-2909, 1986 This is an update of standards and guidelines for eardiopulmonary resuscitation and emergency cardiac care for proper training and performance of these procedures. J. Graham Smith, Jr., M.D.

Seborrheic pemphigoid (German text) Schneider I, Husz S: Hautarzt 37:149-151, 1986 Schnyder described seborrhelc pemphigus in 1969. It is rare and was seen only in older female patients. Authors observed this dermatosis in a 67-year-old man. Treatment with azathioprine and low doses of prednisolone caused prompt remission. Alfred Hollandert tDeeeased.

Fifteen patients with aphthous stomatitis of up to 20 years' duration were studied. Three cleared with a gluten-free diet, 2 with an azo-free diet (excluding tartrazine, sunset yellow, new coccine, and benzoic acid), and one with a milk-free diet. One patient had some improvement with a milk- and azo-free diet, another with a milkfree diet, and another with an azo-free diet. One patient with I-year duration of disease cleared spontaneously; two had no response, and 3 did not complete the study. J. Graham Smith, Jr., M.D.

Prolonged viability of human skin xenografts in rats by cyclosporine Biren CA, Barr RJ, McCullough JL, et al: J Invest Dermatol 86:611-614, 1986 Cyclospodne maintains split-thickness or full-thickness human skin grafts in rats significantly longer (up to 255 days) than in controls. This may serve as an in vivo model for testing percutaneous drug penetration and pharmacokinetics as well as for study of various cutaneous neoplasms in dermatoses. J. Graham Smith, Jr., M.D.

Treatment of chronic urticaria with PUVA or UVA plus placebo: A double-blind study Oiafsson JH, Larko O, Roupe G, et al: Arch Dermatol Rcs 278:228-231, 1986 In a randomized double-blind study 11 patients with chronic urticaria received psoralens plus ultraviolet A (PUVA) and 8 received UVA plus a placebo, In the PUVA group 7 patients showed improvement, 3 noticed no change, and i became worse. In the group receiving UVA plus placebo, 5 patients experienced an improvement whereas the other 3 showed no change. There was no statistically significant difference between the two groups. However, the probability of achieving this degree of improvement in both groups by chance was less than 1%. PUVA may be no better than UVA in the treatment of chronic urticaria. J. Graham Smith, Jr., M,D,