Choroidal vascular occlusion in a child with a connective tissue disease and complement C4 deficiency

Choroidal Vascular Occlusion in a Child with a Connective Tissue Disease and Complement C4 Deficiency Nicola G. Ghazi, MD, Stephen A. Gollance, MD, W. Richard Green, MD Objective: To report the histopathologic findings in the eyes of a patient with a connective tissue disease and complement deficiency. Design: Human postmortem ocular histopathologic study. Intervention: A 15-year-old female died from complications of a connective tissue disease of uncertain etiology, particularly acute respiratory distress syndrome. Abnormalities seen in the eyes at autopsy were consistent with complement activation, granulocyte aggregation, and leukocyte embolization. Main Outcome Measures: Both eyes were examined by light microscopy. Results: Some choroidal vessels were occluded by platelet–fibrin thrombi and occasionally by aggregates of granulocytes and fibrin. Serous retinal detachment involving the macula and peripheral retina was present in both eyes. Conclusions: This is a report of the ocular histopathologic findings in a patient with connective tissue disease and complement C4 deficiency. The light microscopy findings were consistent with complement activation with granulocyte aggregation and leukocyte embolization and may represent another mechanism to explain the clinical findings in patients with connective tissue disease, particularly systemic lupus erythematosus. Ophthalmology 2002;109:1272–1277 © 2002 by the American Academy of Ophthalmology. The phenomenon of granulocyte aggregation with leukocyte embolization may occur in diverse clinical situations, including pulmonary dysfunction in hemodialyzed patients,1–3 Purtscher and Purtscher-like retinopathies with trauma,1 acute pancreatitis,4 – 6 childbirth,7 connective tissue disease,8 myocardial infarction,1 acute respiratory distress syndrome (ARDS),9 and vascular injury and acute reversible hypoxemia in systemic lupus erythematosus (SLE).10,11 In vitro12 and in vivo studies5,10 have shown evidence that complement activation with subsequent release of the granulocyte-aggregating factor C5a is the main mechanism behind this phenomenon. We report the case of a patient with a connective tissue disease, probably SLE, with complement C4 deficiency who died from complications of her disease, particularly ARDS. Postmortem examination showed fibrinous occlusion of some choroidal vessels, serous detachment involving the macula, and multifocal retinal pigment epithelial (RPE) lesions in both eyes. We believe that this case, like Purtscher and Purtscher-like retinopathies, represents another example in which tissue damage is secondary to Originally received: May 18, 2001. Accepted: November 27, 2001. Manuscript no. 210330. From The Eye Pathology Laboratory, Wilmer Institute, and Department of Pathology, The Johns Hopkins Medical Institutions, Baltimore, Maryland. Supported in part by The Independent Order of Odd Fellows, WinstonSalem, North Carolina (Dr. Green). Reprints requests to W. Richard Green, MD, Eye Pathology Laboratory, Maumenee 427, The Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, MD 21287-9248.

1272

© 2002 by the American Academy of Ophthalmology Published by Elsevier Science Inc.

complement activation and granulocyte aggregation. Of interest is the fact that the patient had complement C4 deficiency and histologic findings different from those of Purtscher retinopathy.4

Case Report This 15-year-old female died because of complications of a connective tissue disease of unknown etiology. She was first admitted 2 years before death to a pediatric service because of left lower lobe pneumonia and pulmonary edema. She was found to have proteinuria, hematuria, and bilateral nonpalpable purpuric lesions on the dorsum of her feet. A connective tissue disease was suspected, and extensive laboratory assessment was performed. Some findings suggestive of SLE included positive antinuclear antibody; positive rheumatoid factor; decreased levels of complement C4 (relative C4 deficiency); moderate restrictive pulmonary function test pattern; and a renal biopsy with immunoglobulin G, immunoglobulin M, and C3 deposition on immunofluorescence consistent with diffuse proliferative glomerulonephritis. Other findings not consistent with SLE included normal C3 levels, negative antidouble-stranded DNA antibody, and negative anti-SM antibody. Furthermore, a review of her kidney biopsy at The Johns Hopkins Hospital showed absence of immunoglobulin deposition on transmission electron microscopy. The course of her disease was complicated by nephrotic-range proteinuria; further decrease in her complement C4 level to as low as 0.67 mg/dl; Raynaud’s phenomena; pyoderma gangrenosum; traumatic, nonhealing leg ulcers with vasculitis requiring skin grafting; recurrent pneumonia; hypertension treated with nifedipine and doxazosin; and glaucoma that occurred with chronic systemic steroid use. The systemic steroid dosage was tailored according to her disease manifestations ISSN 0161-6420/02/$–see front matter PII S0161-6420(02)01080-1

Ghazi et al 䡠 Choroidal Vascular Occlusion in Connective Tissue Disease and ranged from 10 mg oral prednisone daily to 1 g intravenous Solu-Medrol three times daily for pyoderma gangrenosum and kidney disease, respectively. Her hypertension was unresponsive to medical therapy, and she progressively developed decreased urine output, anasarca, shortness of breath, left-sided chest pain, recurrence of her skin rash, and vasculitic leg ulcers that required hospitalization 2 years after her initial assessment. She was found to have pericardial effusion and oliguric renal failure, which required continuous hemodialysis. A repeat renal biopsy at that time disclosed membranoproliferative glomerulonephritis with immunoglobulin G, immunoglobulin M, and C3 deposition. During her hospital stay, ophthalmologic follow-up evaluation disclosed an uncorrected visual acuity of 20/160 in both eyes pin-holing to 20/25 in both eyes, applanation tonometry of 18 mmHg in the right eye and 16 mmHg in the left eye, and a cup-to-disc ratio of 0.3 in both eyes. A dilated fundus examination was not performed. She underwent debridement and subsequent skin grafting of her necrotic right lower extremity ulcer. Cultures from the debridement disclosed Proteus mirabilis, Escherichia coli, and Klebsiella pneumoniae. Her hemodynamic status worsened progressively during her stay, with the development of pleural and recurrent pericardial effusions that culminated in pulmonary edema and respiratory distress on her 22nd day of admission, requiring intubation. ARDS was suspected, and a chest x-ray disclosed bilateral pleural effusions and infiltrates. An endotracheal tube culture and a bronchoalveolar lavage disclosed Enterococcus and Staphylococcus aureus with Aspergillus fumigatus, respectively. She was treated with antibiotics and antifungal agents. Her cardiopulmonary status continued to decline despite maximal medical therapy and hemodialysis, and she died from cardiorespiratory arrest 38 days after admission. Autopsy findings included diffuse alveolar damage consistent with ARDS; acute tubular necrosis, membranoproliferative glomerulonephritis, and hypertensive changes in the kidneys; fibrinous pericarditis and epicarditis; pyoderma gangrenosum lesions of the legs; and adrenal gland atrophy consistent with chronic steroid use. No evidence of disseminated intravascular coagulopathy was present. Immunofluorescence disclosed granular and diffuse staining for immunoglobulin G and C3 and diffuse staining for immunoglobulin M and immunoglobulin GHc in the capillary loops of the kidneys. External examination of both eyes was unremarkable. Internal examination showed a temporal area of retinal detachment in the right eye and a nasal area in the left eye, and a detachment of the macula in both eyes. Microscopic examination of both eyes disclosed similar features. The temporal retina was detached by a dense proteinaceous material from the midperiphery to the equator. A 3-mm detachment of the macula was present (Fig 1). Minute areas of RPE detachment were present in the macular region and involved up to four RPE cells (Figs 1–3). Small nodules of hyperplastic pigment epithelium and several single, round pigmented cells were present along the inner surface of the RPE (Figs 1, 3, and 4). The choriocapillaris was intact, but rare vessels contained an intraluminal eosinophilic material. Occasional veins had platelet–fibrin thrombi in their lumen (Figs 2–5), and rare vessels in the choroid subjacent to areas of retinal detachment were occluded by an aggregate of granulocytes and fibrin (Fig 6). Many choroidal vessels had prominent mononuclear inflammatory cells, and the remainder of the choroid contained scattered mononuclear cells. Occasional ciliary body vessels also contained fibrin–platelet material. The retina and longitudinal and cross-sections of the optic nerve were unremarkable.

Discussion The association of autoimmune syndromes, particularly SLE, with complement C4 deficiency has been well documented.13–16 In fact, complement C4 deficiency is thought to be a strong predisposing factor for the development of SLE and immune complex disease.13,14 Almost 50% of patients with inherited deficiencies of C3 complement component, or other components of the classic pathway that activate C3 (C1, C4, and C2), have initially been seen with SLE or disorders that are similar to SLE. “The most common inherited deficiency of the complement system that is associated with SLE is that of C4A, one of the two isotypes of the fourth component of complement.”17 These patients even have a clinical expression of SLE that differs from SLE cases not associated with complement C4 deficiency.14,17 Because of the atypical presentation of these patients, the diagnosis can be problematic. Thus, the association of complement C4 deficiency with a connective tissue disease in our patient is not surprising and might be the reason behind the enigmatic nature of her disease. We believe that the histopathologic findings seen in this case are likely secondary to complement activation and granulocyte aggregation, a mechanism that has recently been considered responsible for the development of Purtscher and Purtscher-like retinopathies in various clinical settings.1,4 – 8 The evidence for the occurrence of such a mechanism in our case is fivefold. First, Abramson et al10 presented evidence that patients with active vasculitides, especially SLE, also have activation of the complement system. These patients were found to have elevated serum levels of C3a and C5a complement components by radioimmunoassay. They were also found to have activated circulating granulocytes with increased expression of the surface antigen CR3, a heterodimer that promotes chemotaxis, adhesion, and aggregation. These data led to the conclusion that in active SLE neutrophils are stimulated by activated components of the complement system to recruit increased numbers of CR3 molecules to their surface, thereby potentiating membrane–membrane interactions and aggregation. Our case had a connective tissue disease, strongly suspected to be SLE, with evidence of active vasculitis leading to tissue necrosis of her lower extremities (pyoderma gangrenosum). Second, the patient had documented infections with gram-positive, gram-negative, and Aspergillus species. It is now well documented18 that infectious processes, particularly infections with gram-negative organisms, lead to activation of the alternate pathway of the complement system through the interaction of bacterial lipopolysaccharide with the serum, with liberation of the granulocyte-aggregating factor C5a.12 Third, the patient died because of complications of her connective tissue disease, particularly ARDS, which has recently been shown to be, at least in part, secondary to complement activation and granulocyte aggregation with leukocyte embolization.1,9 One of these studies9 indicated that there was a highly significant correlation between the presence of polymorphonuclear-aggregating activity in the plasma (C5a) and the development of ARDS and that corticosteroid therapy might be beneficial in patients with ARDS because of its polymorphonuclear-aggre-

1273

Ophthalmology Volume 109, Number 7, July 2002

1274

Ghazi et al 䡠 Choroidal Vascular Occlusion in Connective Tissue Disease gation inhibitory action. Fourth, the patient had evidence of immunoglobulin G and immunoglobulin M and complement C3 deposition in one of her kidney biopsy specimens. This is consistent with a mechanism of nephritis, whereby the complement system is activated because of immunecomplex deposition in the basal lamina of the glomerular capillary bed.16 Last, the patient was on continuous hemodialysis during her hospital stay, and this in itself is a predisposing factor for complement activation with subsequent granulocyte aggregation with leukocyte embolization.1–3 One might argue against this mechanism of complement activation, because the patient had complement C4 deficiency. Complement C4 is essential in the classic pathway of complement activation and liberation of the granulocyteaggregating factor C5a. A possible explanation is as follows: the complement system can be activated through two different pathways, the classic and the alternate pathways, both of which have a common final pathway leading to liberation of active complement factors such as C5a.18 Complement C4 is essential for the classic but not for the alternate, also called properdin, pathway. The latter pathway can be initiated by antigen–antibody complex deposition in tissue or by the interaction of material such as insulin and lipopolysaccharide with serum.18 Therefore, we propose that antigen–antibody complex deposition secondary to the patient’s connective tissue disease with vasculitis and the lipopolysaccharide liberated in her serum secondary to complicating infections led to alternate pathway activation bypassing the classic pathway to which complement C4 is essential. The end result is complement C5a liberation and granulocyte aggregation. This was thought to be the mechanism behind the pathogenesis of nephritis in a young boy with severe SLE and deficiency of the fourth component of complement.16 This case, although thought to exemplify the same mechanism, differs from previously reported cases of Purtscher and Purtscher-like retinopathy in several regards. First, the clinical findings noted in all these cases consisted of retinal infarcts (cotton wool spots) and hemorrhages confined to the posterior pole, particularly to an area limited by the optic nerve head and macula,4,5,7,8 although Gass8 noted that the occlusive process can extensively involve the peripheral fundus, resulting in retinal vascular proliferation and vitreous hemorrhage. Fluorescein angiographic findings disclosed a similar distribution of the occlusive process in all

these cases. Although ophthalmoscopic evaluation of our case was not performed immediately before the patient died, there were no histologic findings suggestive of hemorrhages, cotton wool spots, or vascular occlusion in the retina. So one can conclude that the clinical findings would have been different from those seen in Purtscher and Purtscherlike retinopathy had an ophthalmoscopic evaluation been performed shortly before death. Also, one would have expected to have seen peripheral disease that corresponded to the extensive peripheral areas of serous detachment noted histologically at autopsy. Second, in a previous clinicopathologic case report of Purtscherlike retinopathy in a patient with pancreatitis, Kincaid et al4 documented posterior retinal vascular occlusion associated with adjacent areas of retinal edema. The area of transition from normal to edematous retina was abrupt. The occluding material was consistent with fibrin, occasional hemorrhages were present in the outer plexiform layer, and the RPE and choroid were normal except for occasional occluded choroidal arteries. In the case reported here, there was no evidence of retinal vascular occlusion, edema, or hemorrhages, and the retinal layers were unremarkable in both eyes. The major findings were choroidal vascular occlusion involving occasional choroidal veins and choriocapillaris vessels and macular and peripheral serous retinal detachment with RPE changes in both eyes. We hypothesize that choroidal vascular occlusion occurred secondary to complement activation with subsequent in situ granulocyte aggregation in some choroidal veins and granulocyte aggregation and/or leukoembolism to the choriocapillaris. This led to an increase in the hydrostatic pressure and endothelial damage of the choriocapillaris with subsequent increase in fluid egress into the subretinal space, overwhelming the already damaged RPE pump capacity and leading to serous retinal detachment. The focal RPE changes noted could reflect damage to the RPE caused by choriocapillaris occlusion and retinal detachment19,20 or a healing process of the RPE. The possibility of a vasculitic process involving the choroidal circulation21,22 is unlikely because of the absence of histologic signs of choroidal vasculitis. Also the possibility of hypertensive choroidopathy is unlikely because of the absence of suggestive ocular histologic signs of hypertension, such as retinal infarcts, hard exudates, hemorrhages, or arteriosclerosis; optic nerve head edema or infarcts; or choroidal arteriosclerosis with fibrinoid necrosis of arterial walls, the hallmark of accelerated or malignant hypertension.23 Dis-

4 ™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ Figure 1. The macula is elevated by a dense, fibrin-containing, proteinaceous material (asterisk). Focal areas of retinal pigment epithelium detachment (arrows) and hyperplasia (arrowhead) are present, right eye (stain, periodic acid–Schiff; original magnification, ⫻35). Figure 2. A and B, Platelet—fibrin thrombi (asterisks) in choroidal vessels, right eye. Minute areas of retinal pigment epithelium detachment (arrows) are present (stain, A, hematoxylin– eosin; original magnification, ⫻400; B, periodic acid–Schiff; original magnification, ⫻400). Figure 3. Area shows mild vacuolization of the retinal pigment epithelium (RPE) (asterisk), a cluster of hyperplastic RPE (arrows), and an individual pigmented cell (arrowhead) in the subretinal space, left eye (stain, hematoxylin– eosin; original magnification, ⫻400). Figure 4. Area shows a small nodule of hyperplastic retinal pigment epithelium (arrow), right eye (stain, hematoxylin– eosin; original magnification, ⫻400). Figure 5. Choroidal vessel (arrow) has platelet—fibrin thrombus (asterisk) in the lumen, left eye (stain, periodic acid–Schiff; original magnification, ⫻400). Figure 6. Choroidal vessels (arrows) of the right (A) and left (B) eyes have aggregates of granulocytes intermixed with fibrin within the lumen (A and B, stain, hematoxylin– eosin; original magnification, ⫻400).

1275

Ophthalmology Volume 109, Number 7, July 2002 seminated intravascular coagulopathy can lead to exudative retinal detachment associated with RPE changes and fibrin– platelet choroidal vascular occlusion, similar to the case being reported.23 However, the absence of diffuse occlusion of the choriocapillaris by a homogeneous eosinophilic material, the presence of aggregates of granulocytes and fibrin occluding some choroidal vessels, and the absence of signs of disseminated intravascular coagulopathy in the clinical and autopsy findings argue against this possibility. The mechanism proposed here may at least in part explain many of the previously reported cases of SLE in which similar clinical or histopathologic findings were difficult to comprehend.19,21,24,25 Two of these cases are representative and will be briefly discussed. In a similar case report, Graham et al21described a 16-year-old boy with SLE who had cerebral and retinochoroidal vascular changes consisting of typical vasculitis with fibrinoid necrosis of some meningeal and choroidal vessels and occlusion by an amorphous hyaline material without vasculitis in retinal and other meningeal vessels. Clinically, the patient had cotton wool spots close to the optic nerve head, a retinal hemorrhage inferotemporally, and evidence of retinal vascular occlusion. The patient died because of left ventricular failure secondary to hypertension and pulmonary embolism after hemodialysis for pulmonary edema. The authors concluded that the “noninflammatory” vascular disease in the retina could be caused by either an embolus from the heart or by local antigen–antibody complex formation itself blocking the vascular lumen. It is possible as well that the nonvasculitic retinal vascular occlusion was due to granulocyte aggregation and leukoembolism, as in our case, given the nature of the disease (SLE),10 the fact that the patient died after hemodialysis because of pulmonary edema,1–3 and the retinopathy that was seen clinically. Lowder et al24 described two cases with SLE and serous retinal detachment of the macula. Both patients had glomerulonephritis secondary to SLE and elevated blood pressure and had areas of choroidal leakage on fluorescein angiography corresponding to the area of serous retinal detachment. The authors suggested that the fundus changes in SLE might be a reflection of concomitant hypertension and glomerulonephritis; however, the persistence of serous retinal detachment in one of their cases after normalization of blood pressure suggested other causal factors. They concluded that immune complex deposition in the choriocapillaris, which was previously histopathologically documented by immunofluorescence,22 might have caused damage to the vessel wall with subsequent exudation and persistence of the fluid in the subretinal space despite normalization of the blood pressure. This phenomenon, as in our case, can be partially explained on the basis of granulocyte aggregation and leukoembolism. In conclusion, this case represents another example in which tissue damage is secondary to complement activation and granulocyte aggregation despite the presence of complement C4 deficiency. The histopathologic features noted differ from those of Purtscher and Purtscher-like retinopathies and help explain many of the cases of SLE with RPE

1276

abnormalities and serous retinal detachment reported in the literature.

References 1. Jacob HS, Craddock PR, Hammerschmidt DE, Moldow CF. Complement-induced granulocyte aggregation: an unsuspected mechanism of disease [review]. N Engl J Med 1980; 302:789 –94. 2. Craddock PR, Hammerschmidt DE, Moldow CF, et al. Granulocyte aggregation as a manifestation of membrane interactions with complement: possible role of leukocyte margination, microvascular occlusion, and endothelial damage. Semin Hematol 1979;16:140 –7. 3. Arnaout MA, Hakim RM, Todd RF III, et al. Increased expression of an adhesion-promoting surface glycoprotein in the granulocytopenia of hemodialysis. N Engl J Med 1985;312: 457– 62. 4. Kincaid MC, Green WR, Knox DL, Mohler C. A clinicopathological case report of retinopathy of pancreatitis. Br J Ophthalmol 1982;66:219 –26. 5. Jacob HS, Goldstein IM, Shapiro I, et al. Sudden blindness in acute pancreatitis. Possible role of complement-induced retinal leukoembolization. Arch Intern Med 1981;141:134 – 6. 6. Shapiro I, Jacob HS. Leukoembolization in ocular vascular occlusion. Ann Ophthalmol 1982;14:60 –2. 7. Blodi BA, Johnson MW, Gass JDM, et al. Purtscher’s-like retinopathy after childbirth. Ophthalmology 1990;97:1654 –9. 8. Gass JDM. Stereoscopic Atlas of Macular Diseases, Diagnosis and Treatment, 4th ed. St. Louis: Mosby, 1997; vol. 1, 451– 4. 9. Hammerschmidt DE, Weaver LJ, Hudson LD, et al. Association of complement activation and elevated plasma-C5a with adult respiratory distress syndrome: pathophysiological relevance and possible prognostic value. Lancet 1980; 1:947–9. 10. Abramson S, Belmont HM, Hopkins P, et al. Complement activation and vascular injury in systemic lupus erythematosus. J Rheumatol 1987;14(Suppl 13):43– 6. 11. Abramson SB, Dobro J, Eberle MA, et al. Acute reversible hypoxemia in systemic lupus erythematosus. Ann Intern Med 1991;114:941–7. 12. Craddock PR, Hammerschmidt DE, White JG, et al. Complement (C5a)-induced granulocyte aggregation in vitro. A possible mechanism of complement-mediated leukostasis and leukopenia. J Clin Invest 1977;60:260 – 4. 13. Atkinson JP. Complement deficiency: predisposing factor to autoimmune syndromes. Am J Med 1988;85(Suppl 6A): 45–7. 14. Tappeiner G, Hintner H, Scholz S, et al. Systemic lupus erythematosus in hereditary deficiency in the fourth component of complement. J Am Acad Dermatol 1982;7:66 – 79. 15. Minta JO, Urowitz MB, Gladman DD, et al. Selective deficiency of the fourth component of complement in a patient with systemic lupus erythematosus (SLE): immunochemical and biological studies. Clin Exp Immunol 1981;45:72– 80. 16. Schaller JG, Gilliland BG, Ochs HD, et al. Severe systemic lupus erythematosus with nephritis in a boy with deficiency of the fourth component of complement. Arthritis Rheum 1977; 20:1519 –25. 17. Petri M, Watson R, Winkelstein JA, McLean RH. Clinical expression of systemic lupus erythematosus in patients with C4A deficiency. Medicine (Baltimore) 1993;72:236 – 44.

Ghazi et al 䡠 Choroidal Vascular Occlusion in Connective Tissue Disease 18. Ruddy S, Gigli I, Austen FK. The complement system of man (first of four parts). N Engl J Med 1972;287:489 –95. 19. Matsuo T, Nakayama T, Koyama T, Matsuo N. Multifocal pigment epithelial damages with serous retinal detachment in systemic lupus erythematosus. Ophthalmologica 1987;195: 97–102. 20. Kinyoun JL, Kalina RE. Visual loss from choroidal ischemia. Am J Ophthalmol 1986;101:650 – 6. 21. Graham EM, Spalton DJ, Barnard RO, et al. Cerebral and retinal vascular changes in systemic lupus erythematosus. Ophthalmology 1985;92:444 – 8. 22. Diddie KR, Aronson AJ, Ernest JT. Chorioretinopathy in a

case of systemic lupus erythematosus. Trans Am Ophthalmol Soc 1977;75:122–31. 23. Green WR. Retina. In: Spencer WH, ed. Ophthalmic Pathology: An Atlas and Textbook, 4th ed. Philadelphia: WB Saunders, 1996, vol. 2, chap. 9, 1108 –16, 1141–9. 24. Lowder CY, Gutman FA, Zegarra H, et al. Macular and paramacular detachment of the neurosensory retina associated with systemic diseases. Trans Am Ophthalmol Soc 1981;79: 347–70. 25. Jabs DA, Hanneken AM, Schachat AP, Fine SL. Choroidopathy in systemic lupus erythematosus. Arch Ophthalmol 1988; 106:230 – 4.

1277