- Email: [email protected]
Skin lesions, angioedema, eosinophilia, and hypocomplementemia
724
November 1976 The Journal o f P E D i A T R I C S
Skin lesions, angioedema, eosinophilia, and hypoc omp lementemia A six-year-old girl with recurrent urticaria and angioedema, vasculitis, andprobable renal disease exhibited marked blood eosinophilia, increased levels o f serum IgE, circulating Clq precipitins, and hypocomplementemia with evidence of activation of complement by the classic pathway. Biopsies' of skin and muscle revealed heavy infiltrations o f the vessel walls with eosinophils. Immunofluorescence studies" revealed deposition o f lgM, IgE, and C3 in the vessel walls. Exacerbations of the disease were associated with an increase in the eosinophil count and a decrease in the serum levels of C4 and C3. Remission was achieved with cortieosteroid therapy. This patient has many Jbatures in common with the syndrome of skin lesions, angioedema, and hypocomplementemia recently described in adults.
R a i f S. G e h a , M . D . , * and K a m a l F. A k l , M . D . , B e i r u t , L e b a n o n
SKIN LESIONS and h y p o c o m p l e m e n t e m i a are manifestations of i m m u n e complex diseases? A m o n g these are cryoglobulinemia, systemic lupus erythematosus, cutaneous vasculitis associated with rheumatoid arthritis, and occasional drug reactions. The purpose of this communication is to report a sixyear-old girl with urticaria-like skin lesions, angioedema, eosinophilia, and a pathologic picture of vasculitis. I m m u nologic investigations revealed high serum levels of IgE, circulating Clq precipitins, hypocomplementemia, and deposition of immunoglobulins and complement in the blood vessel walls. There was no evidence of SLE, cryoglobulinemia, or rheumatoid arthritis. A syndrome similar to the one experienced by the patient has been recently described in adults? -~ CASE R E P O R T Patient H. K. is a female child who first presented in December, 1974, at the age of six years because of recurrent attacks of itching, rash, and swelling of the face and extremities of two From the Department of Pediatrics, American University Medical Center, and the National Tissue Typing Center. Supported by research grants from the American University of Beirut and the Lebanese National Council for Scientific Research. *Reprint address: Immunology Division, Children's Hospital Medical Center, 300 Longwood Ave,, Boston, Mass. 02115.
Vol. 89, No. 5, pp. 724-727
months' duration. The rash consisted of red itching papules 1 to 2 cm in diameter. The attacks were not precipitated by changes in weather or exposure to sunlight and were not affected by therapy with antihistamines. There was no personal or family history of allergy. From December, 1974, to April, 1975, the child was followed in the outpatient department of the American University Medical Center. Investigations revealed leukocytosis (18,300 to 28,500 white blood cells/ram ~) with persistent eosinophilia (43% to 57% of the total white blood cell count). Ova of Ascaris lumbricoides and Trichuris trichiura were present in stools. Treatment with anfihelmintics eliminated these ova from the stools but did not affect the attacks of urticaria and angioedema. The attacks occurred once every month, lasted for about a week, and were associated with an increase in the eosinophil count. In April, 1975, the child was admitted to the hospital for further investigation. Abbreviations used Ig: immunoglobulin C: complement SLE: serum lupus erythematosus Physical examination was negative except for the above described skin lesions. Laboratory data included: hemoglobin concentration 12.4 gm/dl; white blood cell count 13,400/mm 3 with 46% eosinophils; platelets 330,000/mm 3, reticulocyte count 0.8%; blood group B, Rh positive; anti-A titer 1:32; antistreptolysin O titer 12 Todd Units; concentrations of serum IgG 1,840 mg/dl, IgA 118 mg/dl, IgM 131 mg/dl, IgD 6 mg/dl, IgE 50,000 IU/ml. Bone marrow aspirate contained 33% eosinophils. The
Volume 89 Number 5
Skin lesions, angioedema, eosinophilia, and hypocomplementemia
7 25
Table I. Complement components
Clq Normal ranges Patient
70-130% 32%(15%)
]
C4
I
30-40 mg/dl 26(18)
C3
C3PA
100-200 mg/dl 76(56)
65-140% 110%
I
Cl inhibitor
CH50
75-135% 95%(110%)
35-45 28 (20)
For Clq, C4, and C1 inhibitor, normal ranges are expressed as percentage of reference pooled normal human serum • 2 SD from the mean and patient values as percentage of reference pooled normal human serum. CH50 values are expressed as hemolytic units as in reference 4. Patient values obtained during the attacks are between parentheses.
following laboratory determinations were normal or negative; urine and stool examinations; serum concentrations of urea, electrolytes, creatinine, HBsAg, alkaline phosphatase, transaminases, and protein electrophoresis; lupus erythematosus clot tests (repeated several times); antinuclear antibodies; Wasserman reaction; rheumatoid factor; antibodies to mitochondria, smooth muscle and thyroid; serum DNA-binding; urinary porphyrins; antibodies to hydatid antigen; roentgenograms of chest and joints; intravenous pyelogram; electrocardiogram. Creatinine clearance was 60 ml/minute/l.73 m ~ (repeated twice). Serum C3 level was found to be low (76 mg/dl). Skin testing showed normal delayed hypersensitivity to candida and streptokinase-streptodornase antigens. There was a normal number of circulating thymus-derived and bone marrow-derived lymphocytes (43% and 21%, respectively). There was normal in vitro proliferation of blood lymphocytes in response to stimulation with phytohemagglutinin, pokeweed mitogen, and Concanavalin A. Skin testing with common allergens and with penicilloyl polylysine did not give positive responses. Other special immunologic studies are detailed below. During an observation period of three months (April to June, 1975) the attacks increased in number, severity, and duration and were associated with a decrease in the serum C3 level (56 mg/dl) and an increase in the eosinophil count. Therapy was started with prednisone 2 mg/kg/day. In two weeks the patient became symptom free. In one month creatinine clearance increased from 60 to 110 ml/min/1.73 m 2, the eosinophil count dropped from 15,500/m 3 to 2,400/mn ~ and the serum IgE level decreased to 18,000 IU/ml. Serum levels of complement components reverted to normal after six weeks of therapy. An attempt to discontinue corticosteroid therapy was followed within two weeks by recurrence of the symptoms and by a drop in serum C3 concentration and an increase in the eosinophil count.
METHODS Sera and E D T A plasma were stored at - 7 0 ~ with 0.1% sodium azide. Hemolytic complement (CH50) was measured by the method of M a y e r ? Concentrations of Clq, C4, C3, and C3 proactivator were determined by radial immunodiffusion. ~ C1 esterase inhibitor was measured by the method of L a c h m a n and associates. 7 Circulating C3 breakdown products were detected by crossed flnmunoelectrophoresis. 8 Clq precipitins were studied by the method of Agnello and associatesY Cryoprecipitates were assessed as described by Brouet and associates? ~
Serum DNATbinding activity (antibodies to doublestranded D N A ) was measured by the Farr technique using a commercially available kit (Radiochemical Center, Amersham, England). Serum IgE levels were determined by radioimmunoassay using IgE Phadebas Kit (Pharmacia, Uppsala, Sweden). T cell n u m b e r was assessed by rosette format!on with sheep red blood cells?' B cell number was assessed by surface staining with fluorescein conjugated antisera to human immunoglobulins? 2 In vitro lymphocyte studies were performed as described in reference. 1:~C3 Coombs tests were performed as described by Kourilsky and associates? ~ Urinary histamine was determined by the method of Huff and associates? 5 RESULTS
Complement studies. C o m p l e m e n t components are shown in Table I. Concentrations o f C3 and of C t and C4, the early reacting components of the classic pathway o f complement activation, were low. During exacerbations of the disease these abnormalities became more marked and circulating C3 breakdown products became detectable. Concentrations of C3 proactivator and of C1 esterase inhibitor were normal. C o m p l e m e n t studies were normal in the father and mother of the patient. Other studies. Clq precipitation test and C3 Coombs test were positive during attacks. Sucrose density gradient ultracentrifugation of fresh serum taken during the attacks revealed the Clq precipitins to be localized to the region of the 7S marker. Ouchterlony analysis of Clq precipitins did not reveal any immunoglobulin components. Cryoprecipitins were never detected in the serum. Urinary excretion o f histamine increased three- to fivefold during the attacks. Biopsy of affected skin and underlying calf muscle revealed diffuse infiltration with polymorphonuclear leukocytes, round cells, and eosinophils in the dermis and in the fat and interstitial spaces between the muscle fibers. Eosinophils ,were the preponderant cells in these infiltrates. Walls of small blood vessels were infiltrated by the same ceils and few vessels were occluded. There was no necrosis of muscle fibers. N o larvae were seen. I m m u n o -
726
Geha and A k l
fluorescent staining revealed deposition of C3, IgM, and IgE in the vessel walls. Renal biopsy contained eight glomeruli which were normal by light microscopy. No kidney tissue was obtained for immunofluorescent Studies. DISCUSSION The two best-known syndromes in which hypocomplementemia has been associated with skin lesions are SLE and cryoglobulinemia. These two diseases were not demonstrated in our patient. Cryoglobulinemia is exceedingly rare in the pediatric age group and cryoglobulins were absent from the patient's serum. The skin lesions of the patient differed from those seen in SLE in that they were urticarial and tended to appear and clear rapidly. Immunofluorescence studies of the skin lesions failed to demonstrate the basement membrane deposition of immunoglobulins typical of SLE lesions. TM SLE clot tests were repeatedly negative. The findings in our patient are very similar to those described in nine adult patients recently reported by McDuffie and associates, ~ Sissons and associates, 3 and Agnello and colleagues? All nine patients had urticaria, angioedema , circulating Clq precipitins, and hypocomplementemia. None of these cases had either leukocytosis or eosinophilia, however, and their serum IgE levels were not reported. Our patient had low molecular weight precipitins which contained no immunoglobulins and which are clearly distinct from Soluble immune complexes detected in certain SLE sera by Clq precipitation. It is possible that the Clq precipitins in our patient represent viral or bacterial antigens that have fixed Clq and activated the complement system. 17 The high levels of serum IgE, the blood eosinophilia which varied with the severity of the disease, the tissue infiltration with eosinophils, the deposition of IgE, IgM, and C3 in the vessel walls, all favor the presence of an antigen acting as an allergen and causing a hypersensitivity angiitis. The exact nature of the antigen is unknown but it is unlikely to have been an intestinal parasite, hydatid, toxocara, or trichinella. Hypocomplementemia (low C3 and low CH50) was more pronounced d u r i n g exacerbations of the disease. The depressed serum levels of C 1 and C4 suggest that the hypocomplementemia was secondary to activation of the classic pathway of the complement system. There was no evidence of activation of the alternate pathway of complement as C3PA levels remained normal. In vivo activation of C3 was confirmed by demonstration of C3 breakdown products in the serum and on the surface of the red blood cells. Complement could have been activated by circulating immune complexes but no immunoglobulins were
The Journal of Pediatrics November 1976
detected in the Clq precipitins. A possible explanation for co~aplement activation could be an intrinsic failure of the patient's complement system resulting in failure to eliminate a viral or bacterial antigen which then continues to circulate as a C l q precipitin and constantly activates complement b y the classic pathway. A similar sequence of events occurs in patients with deficiencies of Clr, C2, or C4 who have been reported to develop an SLE-tike illness. During the acute attacks creatinine clearance was low. This could be secondary to a focal nephritis that was missed on kidney biopsy. Pruritus and skin rash occurred during the attacks together with an increase in urinary excretion of histamine, depression of C3, and increase in eosinophil count. Excess release of histamine could be secondary to excess amounts of a C3 anaphylatoxin or to interaction of antigen and IgE antibody on the surface of basophils and mast cells. Treatment with prednisone was very effective ~in controlling the symptoms of the patient and in reversing the complement abnormalities. The mechanism of the corticosteroid responsiveness as well as the frequency and the long-term prognosis of the present s y n d r o m e remain unknown. The authors thank Ms. Virginia Salameh and Orietta Geha for excellent technical assistance and Drs. Samir Najjar and Fred S. Rosen for reviewing the manuscript. REFERENCES
1. McDuffie FC: Serum complement levels in cutaneous diseases, Br J Dermatol 82(Suppl 5):20, 1970. 2. McDuffie FC, Sams WM, Malonado JE, Andreini PH, Conn DL, and Samayoa EA: Cutaneous vasculitis and hypocomplementemia, Proc Staff Meetings Mayo Clinic, 48:340, 1973. 3. Sissons JP, Williams DG, Peters DK, Boulton-Jones JM, and Goldsmith H J: Skin lesions, angio-edema, and hypocomplementemia, Lancet 2:1350, 1974. 4. Agnello V, Ruddy S, Winchester R, Christian CL, and Kunkel HG: Hereditary C2 deficiency in systemic lupus erythematosus and acquired complement abnormalities in an unusual SLE-related syndrome. In Immunodeficiency in man and animals, Birth Defects Original Articles Series, 11:312, 1975. 5. Mayer MM: Complement and complement fixation, in Kabat EA, editor: Kabat and Mayer's experimental immunochemistry, ed 2, Springfield, Ill, 1961, Charles C Thomas, Publisher, pp 133-240. 6. Mancini G, Carbonara AO, and Hermans JF: Immunochemical quantitation of antigens by single radial immunodiffusion, Immunochemistry 2:235, 1965. 7. Lachmann PJ, Hobart MJ, and Aston WP: Complement technology, in Weir M, editor: Handbook of experimental immunology, Oxford, 1973, Blackwell Scientific publications, pp 5.11-5.12.
Volume 89Number 5
8. 9.
10.
11. 12.
13.
Skin lesions, angioedema, eosinophilia, and hypocomylementemia
Laurell CB: Antigens-antibody cross electrophoresis, Anal Biochem 10:358, 1965. Agnello V, Winchester R J, and Kunkel HG: Precipitin reactions of the Clq components of complement with aggregated gammaglobulin and immune complexes in gel diffusion, Immunology 19:909, 1970. Brouet JC, Clamet JP, Danon F, Klein M, and Seligmann M: Biologic and clinical significance of cryoblobulins, Am J Med 57:775, 1974. Brain P, Gordon J, and Willetts WA: Rosette lbrmation by peripheral lymphocytes, Clin Exp hnmunol 6:681, 1970. Froland S, Natvig JB, and Berdal J: Surface-bound immunoglobulin as a marker of B lymphocytes in man, Nature [New Biol] 234:251, 1971. Geha RS, Rosen FS, and Merler E: Identification and characterization of subpopulations of lymphocytes i n human peripheral blood after fractionation on discontin-
14.
15.
16.
17.
727
uous gradients of albumin; the cellular defect in X-linked agammaglobulinemia, J Clin Invest 52:1726, 1973. Kourilsky O, Lucas JP, Poryau-Lemaux C, Neuilly G, and Rich~et G: Significance of complement-positive Coombs test in patients with glomerular disease, Lancet 2:683, 1974. HuffJA, Davis VE, and Brown H: A simplified technique for the estimation of histamine in blood and urine, J Lab Clin Med 67:1044, 1966. Svec K, Blair J, and Kaplan M: lmmunopathologic studies of systemic lupus erythematosus: 1. Tissue-bound immunoglobutins in relation to serum antinuclear immunoglobulins in systemic lupus erythematosus with LE cell factor, J Clin Invest 46:558, 1967. Agnello V, Winchester RJ, and Kunkel HG: Precipitin reactions of Clq with various gamma globulins and anionic macromolecules, J Immunol 107:309, 1971.
November 1976 The Journal o f P E D i A T R I C S
Skin lesions, angioedema, eosinophilia, and hypoc omp lementemia A six-year-old girl with recurrent urticaria and angioedema, vasculitis, andprobable renal disease exhibited marked blood eosinophilia, increased levels o f serum IgE, circulating Clq precipitins, and hypocomplementemia with evidence of activation of complement by the classic pathway. Biopsies' of skin and muscle revealed heavy infiltrations o f the vessel walls with eosinophils. Immunofluorescence studies" revealed deposition o f lgM, IgE, and C3 in the vessel walls. Exacerbations of the disease were associated with an increase in the eosinophil count and a decrease in the serum levels of C4 and C3. Remission was achieved with cortieosteroid therapy. This patient has many Jbatures in common with the syndrome of skin lesions, angioedema, and hypocomplementemia recently described in adults.
R a i f S. G e h a , M . D . , * and K a m a l F. A k l , M . D . , B e i r u t , L e b a n o n
SKIN LESIONS and h y p o c o m p l e m e n t e m i a are manifestations of i m m u n e complex diseases? A m o n g these are cryoglobulinemia, systemic lupus erythematosus, cutaneous vasculitis associated with rheumatoid arthritis, and occasional drug reactions. The purpose of this communication is to report a sixyear-old girl with urticaria-like skin lesions, angioedema, eosinophilia, and a pathologic picture of vasculitis. I m m u nologic investigations revealed high serum levels of IgE, circulating Clq precipitins, hypocomplementemia, and deposition of immunoglobulins and complement in the blood vessel walls. There was no evidence of SLE, cryoglobulinemia, or rheumatoid arthritis. A syndrome similar to the one experienced by the patient has been recently described in adults? -~ CASE R E P O R T Patient H. K. is a female child who first presented in December, 1974, at the age of six years because of recurrent attacks of itching, rash, and swelling of the face and extremities of two From the Department of Pediatrics, American University Medical Center, and the National Tissue Typing Center. Supported by research grants from the American University of Beirut and the Lebanese National Council for Scientific Research. *Reprint address: Immunology Division, Children's Hospital Medical Center, 300 Longwood Ave,, Boston, Mass. 02115.
Vol. 89, No. 5, pp. 724-727
months' duration. The rash consisted of red itching papules 1 to 2 cm in diameter. The attacks were not precipitated by changes in weather or exposure to sunlight and were not affected by therapy with antihistamines. There was no personal or family history of allergy. From December, 1974, to April, 1975, the child was followed in the outpatient department of the American University Medical Center. Investigations revealed leukocytosis (18,300 to 28,500 white blood cells/ram ~) with persistent eosinophilia (43% to 57% of the total white blood cell count). Ova of Ascaris lumbricoides and Trichuris trichiura were present in stools. Treatment with anfihelmintics eliminated these ova from the stools but did not affect the attacks of urticaria and angioedema. The attacks occurred once every month, lasted for about a week, and were associated with an increase in the eosinophil count. In April, 1975, the child was admitted to the hospital for further investigation. Abbreviations used Ig: immunoglobulin C: complement SLE: serum lupus erythematosus Physical examination was negative except for the above described skin lesions. Laboratory data included: hemoglobin concentration 12.4 gm/dl; white blood cell count 13,400/mm 3 with 46% eosinophils; platelets 330,000/mm 3, reticulocyte count 0.8%; blood group B, Rh positive; anti-A titer 1:32; antistreptolysin O titer 12 Todd Units; concentrations of serum IgG 1,840 mg/dl, IgA 118 mg/dl, IgM 131 mg/dl, IgD 6 mg/dl, IgE 50,000 IU/ml. Bone marrow aspirate contained 33% eosinophils. The
Volume 89 Number 5
Skin lesions, angioedema, eosinophilia, and hypocomplementemia
7 25
Table I. Complement components
Clq Normal ranges Patient
70-130% 32%(15%)
]
C4
I
30-40 mg/dl 26(18)
C3
C3PA
100-200 mg/dl 76(56)
65-140% 110%
I
Cl inhibitor
CH50
75-135% 95%(110%)
35-45 28 (20)
For Clq, C4, and C1 inhibitor, normal ranges are expressed as percentage of reference pooled normal human serum • 2 SD from the mean and patient values as percentage of reference pooled normal human serum. CH50 values are expressed as hemolytic units as in reference 4. Patient values obtained during the attacks are between parentheses.
following laboratory determinations were normal or negative; urine and stool examinations; serum concentrations of urea, electrolytes, creatinine, HBsAg, alkaline phosphatase, transaminases, and protein electrophoresis; lupus erythematosus clot tests (repeated several times); antinuclear antibodies; Wasserman reaction; rheumatoid factor; antibodies to mitochondria, smooth muscle and thyroid; serum DNA-binding; urinary porphyrins; antibodies to hydatid antigen; roentgenograms of chest and joints; intravenous pyelogram; electrocardiogram. Creatinine clearance was 60 ml/minute/l.73 m ~ (repeated twice). Serum C3 level was found to be low (76 mg/dl). Skin testing showed normal delayed hypersensitivity to candida and streptokinase-streptodornase antigens. There was a normal number of circulating thymus-derived and bone marrow-derived lymphocytes (43% and 21%, respectively). There was normal in vitro proliferation of blood lymphocytes in response to stimulation with phytohemagglutinin, pokeweed mitogen, and Concanavalin A. Skin testing with common allergens and with penicilloyl polylysine did not give positive responses. Other special immunologic studies are detailed below. During an observation period of three months (April to June, 1975) the attacks increased in number, severity, and duration and were associated with a decrease in the serum C3 level (56 mg/dl) and an increase in the eosinophil count. Therapy was started with prednisone 2 mg/kg/day. In two weeks the patient became symptom free. In one month creatinine clearance increased from 60 to 110 ml/min/1.73 m 2, the eosinophil count dropped from 15,500/m 3 to 2,400/mn ~ and the serum IgE level decreased to 18,000 IU/ml. Serum levels of complement components reverted to normal after six weeks of therapy. An attempt to discontinue corticosteroid therapy was followed within two weeks by recurrence of the symptoms and by a drop in serum C3 concentration and an increase in the eosinophil count.
METHODS Sera and E D T A plasma were stored at - 7 0 ~ with 0.1% sodium azide. Hemolytic complement (CH50) was measured by the method of M a y e r ? Concentrations of Clq, C4, C3, and C3 proactivator were determined by radial immunodiffusion. ~ C1 esterase inhibitor was measured by the method of L a c h m a n and associates. 7 Circulating C3 breakdown products were detected by crossed flnmunoelectrophoresis. 8 Clq precipitins were studied by the method of Agnello and associatesY Cryoprecipitates were assessed as described by Brouet and associates? ~
Serum DNATbinding activity (antibodies to doublestranded D N A ) was measured by the Farr technique using a commercially available kit (Radiochemical Center, Amersham, England). Serum IgE levels were determined by radioimmunoassay using IgE Phadebas Kit (Pharmacia, Uppsala, Sweden). T cell n u m b e r was assessed by rosette format!on with sheep red blood cells?' B cell number was assessed by surface staining with fluorescein conjugated antisera to human immunoglobulins? 2 In vitro lymphocyte studies were performed as described in reference. 1:~C3 Coombs tests were performed as described by Kourilsky and associates? ~ Urinary histamine was determined by the method of Huff and associates? 5 RESULTS
Complement studies. C o m p l e m e n t components are shown in Table I. Concentrations o f C3 and of C t and C4, the early reacting components of the classic pathway o f complement activation, were low. During exacerbations of the disease these abnormalities became more marked and circulating C3 breakdown products became detectable. Concentrations of C3 proactivator and of C1 esterase inhibitor were normal. C o m p l e m e n t studies were normal in the father and mother of the patient. Other studies. Clq precipitation test and C3 Coombs test were positive during attacks. Sucrose density gradient ultracentrifugation of fresh serum taken during the attacks revealed the Clq precipitins to be localized to the region of the 7S marker. Ouchterlony analysis of Clq precipitins did not reveal any immunoglobulin components. Cryoprecipitins were never detected in the serum. Urinary excretion o f histamine increased three- to fivefold during the attacks. Biopsy of affected skin and underlying calf muscle revealed diffuse infiltration with polymorphonuclear leukocytes, round cells, and eosinophils in the dermis and in the fat and interstitial spaces between the muscle fibers. Eosinophils ,were the preponderant cells in these infiltrates. Walls of small blood vessels were infiltrated by the same ceils and few vessels were occluded. There was no necrosis of muscle fibers. N o larvae were seen. I m m u n o -
726
Geha and A k l
fluorescent staining revealed deposition of C3, IgM, and IgE in the vessel walls. Renal biopsy contained eight glomeruli which were normal by light microscopy. No kidney tissue was obtained for immunofluorescent Studies. DISCUSSION The two best-known syndromes in which hypocomplementemia has been associated with skin lesions are SLE and cryoglobulinemia. These two diseases were not demonstrated in our patient. Cryoglobulinemia is exceedingly rare in the pediatric age group and cryoglobulins were absent from the patient's serum. The skin lesions of the patient differed from those seen in SLE in that they were urticarial and tended to appear and clear rapidly. Immunofluorescence studies of the skin lesions failed to demonstrate the basement membrane deposition of immunoglobulins typical of SLE lesions. TM SLE clot tests were repeatedly negative. The findings in our patient are very similar to those described in nine adult patients recently reported by McDuffie and associates, ~ Sissons and associates, 3 and Agnello and colleagues? All nine patients had urticaria, angioedema , circulating Clq precipitins, and hypocomplementemia. None of these cases had either leukocytosis or eosinophilia, however, and their serum IgE levels were not reported. Our patient had low molecular weight precipitins which contained no immunoglobulins and which are clearly distinct from Soluble immune complexes detected in certain SLE sera by Clq precipitation. It is possible that the Clq precipitins in our patient represent viral or bacterial antigens that have fixed Clq and activated the complement system. 17 The high levels of serum IgE, the blood eosinophilia which varied with the severity of the disease, the tissue infiltration with eosinophils, the deposition of IgE, IgM, and C3 in the vessel walls, all favor the presence of an antigen acting as an allergen and causing a hypersensitivity angiitis. The exact nature of the antigen is unknown but it is unlikely to have been an intestinal parasite, hydatid, toxocara, or trichinella. Hypocomplementemia (low C3 and low CH50) was more pronounced d u r i n g exacerbations of the disease. The depressed serum levels of C 1 and C4 suggest that the hypocomplementemia was secondary to activation of the classic pathway of the complement system. There was no evidence of activation of the alternate pathway of complement as C3PA levels remained normal. In vivo activation of C3 was confirmed by demonstration of C3 breakdown products in the serum and on the surface of the red blood cells. Complement could have been activated by circulating immune complexes but no immunoglobulins were
The Journal of Pediatrics November 1976
detected in the Clq precipitins. A possible explanation for co~aplement activation could be an intrinsic failure of the patient's complement system resulting in failure to eliminate a viral or bacterial antigen which then continues to circulate as a C l q precipitin and constantly activates complement b y the classic pathway. A similar sequence of events occurs in patients with deficiencies of Clr, C2, or C4 who have been reported to develop an SLE-tike illness. During the acute attacks creatinine clearance was low. This could be secondary to a focal nephritis that was missed on kidney biopsy. Pruritus and skin rash occurred during the attacks together with an increase in urinary excretion of histamine, depression of C3, and increase in eosinophil count. Excess release of histamine could be secondary to excess amounts of a C3 anaphylatoxin or to interaction of antigen and IgE antibody on the surface of basophils and mast cells. Treatment with prednisone was very effective ~in controlling the symptoms of the patient and in reversing the complement abnormalities. The mechanism of the corticosteroid responsiveness as well as the frequency and the long-term prognosis of the present s y n d r o m e remain unknown. The authors thank Ms. Virginia Salameh and Orietta Geha for excellent technical assistance and Drs. Samir Najjar and Fred S. Rosen for reviewing the manuscript. REFERENCES
1. McDuffie FC: Serum complement levels in cutaneous diseases, Br J Dermatol 82(Suppl 5):20, 1970. 2. McDuffie FC, Sams WM, Malonado JE, Andreini PH, Conn DL, and Samayoa EA: Cutaneous vasculitis and hypocomplementemia, Proc Staff Meetings Mayo Clinic, 48:340, 1973. 3. Sissons JP, Williams DG, Peters DK, Boulton-Jones JM, and Goldsmith H J: Skin lesions, angio-edema, and hypocomplementemia, Lancet 2:1350, 1974. 4. Agnello V, Ruddy S, Winchester R, Christian CL, and Kunkel HG: Hereditary C2 deficiency in systemic lupus erythematosus and acquired complement abnormalities in an unusual SLE-related syndrome. In Immunodeficiency in man and animals, Birth Defects Original Articles Series, 11:312, 1975. 5. Mayer MM: Complement and complement fixation, in Kabat EA, editor: Kabat and Mayer's experimental immunochemistry, ed 2, Springfield, Ill, 1961, Charles C Thomas, Publisher, pp 133-240. 6. Mancini G, Carbonara AO, and Hermans JF: Immunochemical quantitation of antigens by single radial immunodiffusion, Immunochemistry 2:235, 1965. 7. Lachmann PJ, Hobart MJ, and Aston WP: Complement technology, in Weir M, editor: Handbook of experimental immunology, Oxford, 1973, Blackwell Scientific publications, pp 5.11-5.12.
Volume 89Number 5
8. 9.
10.
11. 12.
13.
Skin lesions, angioedema, eosinophilia, and hypocomylementemia
Laurell CB: Antigens-antibody cross electrophoresis, Anal Biochem 10:358, 1965. Agnello V, Winchester R J, and Kunkel HG: Precipitin reactions of the Clq components of complement with aggregated gammaglobulin and immune complexes in gel diffusion, Immunology 19:909, 1970. Brouet JC, Clamet JP, Danon F, Klein M, and Seligmann M: Biologic and clinical significance of cryoblobulins, Am J Med 57:775, 1974. Brain P, Gordon J, and Willetts WA: Rosette lbrmation by peripheral lymphocytes, Clin Exp hnmunol 6:681, 1970. Froland S, Natvig JB, and Berdal J: Surface-bound immunoglobulin as a marker of B lymphocytes in man, Nature [New Biol] 234:251, 1971. Geha RS, Rosen FS, and Merler E: Identification and characterization of subpopulations of lymphocytes i n human peripheral blood after fractionation on discontin-
14.
15.
16.
17.
727
uous gradients of albumin; the cellular defect in X-linked agammaglobulinemia, J Clin Invest 52:1726, 1973. Kourilsky O, Lucas JP, Poryau-Lemaux C, Neuilly G, and Rich~et G: Significance of complement-positive Coombs test in patients with glomerular disease, Lancet 2:683, 1974. HuffJA, Davis VE, and Brown H: A simplified technique for the estimation of histamine in blood and urine, J Lab Clin Med 67:1044, 1966. Svec K, Blair J, and Kaplan M: lmmunopathologic studies of systemic lupus erythematosus: 1. Tissue-bound immunoglobutins in relation to serum antinuclear immunoglobulins in systemic lupus erythematosus with LE cell factor, J Clin Invest 46:558, 1967. Agnello V, Winchester RJ, and Kunkel HG: Precipitin reactions of Clq with various gamma globulins and anionic macromolecules, J Immunol 107:309, 1971.