Eye (1991) 5, 48-- 55

Vaso-occlusive Retinopathy in the Primary Anti­ phospholipid Antibody Syndrome

1. F. ACHESON,! R. M. C. GREG SON,2 P. MERRY,3 W. E. SCHULENBURG! London

Summary We report two patients with contrasting patterns of retinal vascular occlusion associ­ ated with the primary anti-phospholipid antibody syndrome. The immuno-path­ ological features and clinical associations are discussed. This condition is of interest to ophthalmologists because of its association with thrombosis in the , brain and elsewhere and because it provides new insights into the pathogenesis of retinal vas­ cular disease in young patients.

In recent years a strong association has In 1975, age 21, hypertension and classical become apparent between thrombosis and migraine were diagnosed but no specific medi­ the presence of antibodies to negatively cations were given. The oral contraceptive pill charged phospholipids. Anti-phospholipid was stopped. In 1977 she became pregnant, antibodies may be found in patients who have developed pre-eclamptic toxaemia with a systemic lupus erythematosis (SLE), syphilis blood pressure of 200/120 mm Hg and at 34 and more rarely in other conditions. In addi­ weeks gestation had an induced delivery of a tion, a distinct clinical entity has now been still-born macerated 24 week fetus. The cause described in which arterial or venous throm­ of the intrauterine death was placental insuf­ bosis, thrombocytopaenia, and recurrent fiency. In the post-partum period hyper­ abortion occur in the presence of these anti­ tension persisted and she was given bodies without any associated medical dis­ propranolol 40 mg and prazosin 2 mg both order. This is known as the primary three times daily. antiphospholipid antibody syndrome.!-4 In the period between 1978 and 1980 the We describe two patients with contrasting hypertension was well controlled but Ray­ patterns of vaso-occlusive retinopathy in this naud's phenomenon developed and atenolol condition. The firsthad a picture of peripheral 100 mg daily was given instead of proprano­ retinal arteriolar obliteration and capillary lol. Troublesome Raynaud's phenomenon closure complicated by neovascularisation, persisted and from 1982 all beta-blockers and the second had a branch retinal were stopped and she remained on prazosin occlusion. alone. In 1983 she had a spontaneous abortion Case One at 12 weeks gestation and in 1984 a further A 34 year old white female was referred to the pregnancy was lost due to intrauterine death in March 1989 for the assessment of a left uni­ attributed to placental failure at 30 weeks. ocular supero-nasal visual field defect. Blood pressure was well controlled and the

From: 'Department of , Hammersmith Hospital, London W12 and Western Ophthalmic Hospital, London NWl. 2Medical Retinal Clinic, Moorfields Eye Hospital, London ECl. 3Department of Rheumatology, Charing Cross Hospital, London W6.

Correspondence to: J. F. Acheson FRCS FCOphth, Western Ophthalmic Hospital, Marylebone Road, London NW1 5YE. VASO-OCCLUSIVE RETINOPATHY IN THE PRIMARY ANTI-PHOSPHOLIPID ANTIBODY SYNDROME 49 prazosin was changed to Dyazide one tablet Oral prednisolone 10 mg daily was given for twice daily. eight months. In October 1988 ketanserin was From May 1985 she took no anti-hyperten­ given for continuing migraine attacks and the sive medications and was admitted to hospital blood pressure was mildly elevated at 150/95 as an emergency in December 1985 with a standing, 145/100 lying. In November 1988 blood pressure of 135/100 mmHg standing, she underwent detailed reassessment follow­ 185/110 lying and a right spastic hemiparesis. ing recurrent transient ischaemic attacks, There had been no migrainous episodes for a worsening migraines and left sided visual field number of years. A CT head scan without disturbances. The blood pressure was 160/70. administration of contrast showed a wedge Neurological examination showed a new left shaped area of low-attenuation in the left par­ homonymous hemianopia. A magnetic reson­ ietal region consistent with a cerebral infarc­ ance brain scan showed multiple patchy tion. Extensive livedo reticularis and lesions in the periventricular white matter acrocyanosis was noted. An intra-arterial consistent with small vessel cerebral vaso­ digital subtraction angiogram of the aortic occlusive disease and a dark region in the right arch, neck and intra-cranial vessels was occipito-parietal region consistent with an old normal. Doppler ultrasound studies of carotid cerebral infarction (Figs la and b). blood flow, an echocardiogram and intra­ Doppler ultrasound studies of carotid venous were also normal. Full blood flow were again normal. The platelet blood count, platelets, ESR, renal and count had fallen to 88 x 109/L and IgG anti­ hepatic function were all normal. A screen for cardiolipin antibodies remained present anti-nuclear antibodies (ANA) was negative. (Table I). The erythrocyte sedimentation rate The prothrombin time and the partial throm­ and C-reactive protein levels were normal. A boplastin time were within normal limits and repeat partial thromboplastin time was serological tests for syphilis were negative. slightly prolonged at 43 seconds (normal 40). Ig G anti-cardiolipin antibodies were present Repeat studies for antinuclear antibodies in high tit res (Table I). including anti-Ro, La, Sm, and RNP were From February 1986 to May 1987 she was again negative. fully anti-coagulated with oral warfarin, and She underwent her first ophthalmic eval­ the hypertension was well controlled with uation in March 1989. Corrected visual acu­ hydrallazine 50 mg twice daily, pindolol 5 mg ities were 6/5 on the right and 6/6 on the left. daily and Dyazide once daily. From May 1987 Visual fields to confrontation and on Gold­ warfarin was stopped and aspirin 75 mg daily man perimetry showed a left monocular commenced. From July 1987 the pindolol was supero-nasal defect which did not respect the changed to nifedipine slow release 20 mg and vertical meridian. The intraocular pressures pizotifen was given for fresh migrainous head­ were normal and there were no cells in the aches. In November 1987 the platelet count anterior chamber or vitreous. was mildly reduced at 105 x 109/1 and the IgG Fundoscopy showed normal optic discs and titre of anti-cardioliin antibodies remained mild generalised vascular calibre attenuation elevated. with arterio-venous nipping. In the left infero- Table I. Summary of investigations for case one

Platelets Anti-cardiolipin antibody levels 9 (normal 120 - 400 x J0 /L) JgG (normal <9 units) /gM (normal <9 units)

December 1985 140 60 7.4 March 1986 130 3.1 December 1986 18.5 November 1987 146 16 January 1988 54 May 1988 60 1.2 September 1988 101 35 October 1988 99 35 2.8 March 1989 88 58 4.2 50 J. F. ACHESON ET AL.

Fig.la. Fig. lb.

Figs. la and lb. Case 1. Magnetic resonance imaging scans of brain showing periventricular changes in both hemispheres, more marked on the right consistent with but not diagnostic of vascular disease, and evidence of an old right occipita-parietal haemorrhage.

temporal periphery there was collateral vessel Eye Casualty Department in July 1988 with a formation and an elevated extraretinal fibro­ one day history of floaters in the right eye. vascular complex with localised pre-retinal Full ophthalmic examination was normal haemorrhage. apart from an incomplete posterior vitreous demonstrated detachment and a small subhyaloid haemor­ delayed arteriolar filling in the affected quad­ rhage at the right margin. At age of rant with several neovascular complexes and 17 the lupus anticoagulant was first diagnosed . peripheral capillary non-perfusion. The after viral pneumonia complicated by deep remaining retinal periphery showed vascular venous thrombosis. She was given oral anti­ attenuation and further areas of peripheral coagulants and corticosteroids for three years capillary closure. In addition there was patchy after which time all treatment was stopped as hypofluorescence of the pigment epithelium she remained well. throughout both fundi (Figs 2a-f). At routine review three months later the In view of the probable longstanding nature corrected right visual acuity was 6/6 and a of the neovascularisation indicated by the localised area of intra- fibrous component and the absence of any sig­ and retinal swelling within the right temporal vascular arcade was first noted. The blood nificant vitreous haemorrhage, photo­ pressure was 160/95 and nifedipine 20 mg coagulation was not performed and after twice daily was started. twelve months follow-up the vision remains Fluorescein angiography (Figs 3a-c) stable. showed an area of capillary non-perfusion Case Two with venous collaterals in the territory A 30 year old white woman presented to the drained by one macular branch vein . No leak- VASO-OCCLUSIVE RETINOPATHY IN THE PRIMARY ANTI-PHOSPHOLIPID ANTIBODY SYNDROME 51

Fig.2a. Fig. 2b.

Fig.2e. Fig.2d.

Fig.2e. Fig. 2f.

Figs. 2a-f. Fluorescein angiograms fr om one run showing peripheral arteriolar attenuation, capillary non­ perfusion and neovascularisation. See text for details. 52 J. F. ACHESON ET AL.

Fig.3a. Fig.3b.

Fig.3c. Fig. 3d.

Figs. 3a-c. Fluorescein angiograms showing normal Fig. 3d. Red free photograph three months arteriolar filling (a), vein occlusion (b), and normal later showing exudate formation. optic disc (c). age was shown at this stage and there was no treated with sub-cutaneous heparin 10,000 u evidence of neovascularisation. Investiga­ twice daily from eight weeks gestation. The tions showed an abnormal clotting screen, pregnancy was lost following a spontaneous mild thrombocytopaenia, a false-positive abortion at 13 weeks. VDRL, negative antinuclear activity and strongly positive anti-cardiolipin antibody Discussion levels (Table II) . Anti-phospholipid antibodies are a group of Three months later the lesion was sur­ autoantibodies which are detected by precip­ rounded by hard exudates suggesting capil­ itation or complement fixation tests currently lary decompensation (Fig 3d), but she used in standard tests for syphilis, the lupus declined further fluorescein angiography. No anticoagulant test, or by solid phase radio­ specific ophthalmic treatment was given and immunoassay for anti-cardiolipin antibodies. over the next six months the exudates The terms cardiolipin and phospholipid resolved and the visual acuity remained 6/6 . were first linked when it was demonstrated In May 1989 she became pregnant and was that the antigen bound by reagin in patients VASO-OCCLUSIVE RETINOPATHY IN THE PRIMARY ANTI-PHOSPHOLIPID ANTIBODY SYNDROME 53

Table II. Summary of investigations for case two

Clotting screen: Hb 16 gms Platelet 98 x 109/ml (normal 120 - 400) Prothrombin time 15 secs (control 14) PIT 40 secs (control 35 ) Autoantibody screen: VDRL Positive FfA & TPHA Negative ANA Negative Antibodies to Double Stranded DNA Negative Anti-Cardiolipin Antibodies: IgG 212 Units (normal <5 ) IgM 100 Units (normal <3) with syphilis, as detected by techniques such anti-coagulant, but in vivo their effect is para­ as the Wasserman Reaction (WR) and the doxically to promote a thrombotic tendency, Venereal Disease Reference Laboratory floc­ probably because of further actions on phos­ culation test (VORL), was an acidic phos­ pholipid components of platelet membranes pholipid obtained by alcohol extraction of ox and vascular endothelium, as well as on heart muscle .3 thrombolytic factors such as local prostacy­ With widespread screening for syphilis clin, antithrombin III and Protein C activa­ among military personnel during the Second tion .6 World War it became apparent that some A thrombotic tendency is a feature of a sub­ people had a positive test for syphilis without population of patients with SLE7 in whom a any other evidence of the disease . This was vaso-occlusive retinopathy may occasionally known as the biological false positive test for be seen rather than the commoner and milder syphilis (BFP-STS) and had a high prevalence picture of haemorrhages and cotton wool in connective tissue and autoimmune dis­ spotS .8 Asherson et al . have reported the pres­ eases, and a strong association with other ence of lupus anticoagulant and anti-cardioli­ autoantibodies, particularly antinuclear anti­ pin antibodies in this subgroup and showed an bodies (ANA).3 important association with cerebrovascular Later it was shown that patients with the disease .9 The ophthalmic features were not BFP-STS might have a circulating anticoag­ however discussed in any detail . Retinal ulant which because of its association with and vein occlusions associated with ANA became known as the Lupus Anticoag­ antiphospholipid antibodies have also been ulant .5 The lupus anticoagulant has been reported in Sneddon's syndrome (livedo retic­ shown to be an antibody from the IgG or IgM ularis, variable cerebrovascular events and class and reacts with the phospholipid portion labile hypertension) and "lupus-like disease" of the prothrombin converter complex in the (resembling SLE but not including four or coagulation cascade to cause prolongation of more of the 1982 revised American Rheu­ the activated partial thromboplastin time matism Association criteria for the classifi­ (APTT), the Russell viper venom time cation of SLE) .10-14 (RV VT) and, less frequently, prolongation of The association between retinal vascular the Prothrombin time (PT) . 3 occlusions and cerebrovascular disease in It is now apparent that the lupus anticoag­ these patients is of interest in view of path­ ulant and the BFP-STS are anti-phospholipid ological evidence suggesting that not all vas­ antibodies with similar specificityto anti-card­ cular occlusions in SLE are due to vasculitis . iolipin antibodies, and collectively they are all Graham et al .15 have reported post mortem referred to as anti-phospholipid antibodies . details of a patient in whom cerebral and ret­ New solid phase radioimmunoassay tech­ inal vessels were occluded by amorphous hya­ niques for detecting anti-cardiolipin anti­ line material without arteritis . It has been bodies are the most sensitive test for all suggested that in patients who have SLE, members of this family. thrombosis in the brain, the and other In vitro, anti-phospholipid antibodies are sites is due to direct interference with the 54 J. F. ACHESON ET AL.

coagulation system mediated by anti-phos­ sive retinopathy2.8. As discussed above, anti­ pholipid antibodies rather than by circulating phospholipid antibodies have been implicated immune complexes.3 Antibodies to neuronal in the pathogenesis of vaso-occlusive disease membrane antigens may also be important in in SLE. This patient has a hypercoagulabJe some cases.16 state and thrombotic tendency associated It is now apparent that these phenomena with anti-phospholipid antibodies but without are seen in ophthalmic clinical practice in SLE or any evidence of embolisation. We patients with anti-phospholipid antibodies believe that the retinopathy is best explained and a history of thrombosis and who do not on this basis although possibly aggravated by have SLE or any other disease. Three out of hypertension induced arteriolar sclerosis and fivecases reported by Levine et al. 5 presenting attenuation. to the ophthalmologist with a variety of cere­ In the second case, the subhyaloid haemor­ bral and retinal ischaemic events did not have rhage was due to either an acute partial pos­ SLE and did have the lupus anticoagulant. terior vitreous detachment or to They did not report details of specific sensitive breakthrough blood from an acute venous anti-cardiolipin antibody levels. Three occlusion. There was no evidence of neo-vas­ patients described by Kleiner et al. 17 had lupus cularisation or of peripheral retinal pathol­ anticoagulant without SLE or other disease. ogy. On review, the haemodynamic One was negative for anti-cardiolipin anti­ disturbance had the appearances of a macular branch retinal vein occlusion with capillary body. These three patients all had severe decompensation and leakage explaining the visual loss secondary to diffuse bilateral ret­ pattern of hard exudate formation. Mild inal vascular obstructions with rubeosis iridis hypertension was diagnosed at this time. Mac­ and vitreous haemorrhage. Garcia Vicente et ular branch retinal vein occlusion in a 30 year al.18 have reported one similar patient with old patient is a sufficiently uncommon con­ proliferative retinopathy and lupus anticoag­ dition to merit further investigation and it is ulant without details of anti-cardiolipin anti­ clear that in this patient a combination of a body levels. thrombotic tendency and raised blood pres­ Our cases can be added to these first reports sure produced the clinical picture. This case is of vaso-occlusive retinopathy occurring in the of particular interest as, to our knowledge, presence of the primary anti-phospholipid branch retinal vein occlusion has not been antibody syndrome. The interpretation of reported in SLE. both cases is complicated by the fact that both The anti-phospholipid antibody syndrome patients were hypertensive. is of great interest to ophthalmologists firstly The first patient presented after a long because of reports of a poor visual prognosis history of cerebro-vascular events with per­ and secondly because of its association with ipheral retinal neovascularisation straddling thrombosis in the brain and elsewhere. Treat­ perfused and non-perfused retina. This pat­ ment with anticoagulation and immunosup­ tern has been observed in several conditions pression and by control of any associated including sickle cell disease, peripheral retinal hypertension may improve the prognosis in vasculitis and carotid insufficiency. 19 This cases where or recurrent abortion are patient did not have a haemoglobinopathy prominent features. and the angiogram does not suggest vessel Ophthalmologists will want to know when wall inflammation. Carotid insufficiency was it might be worth looking for this condition in explicitly excluded. Peripheral retinal arterio­ patients presenting with retinal vascular lar attenuation with capillary non-perfusion occlusions. There is no evidence to support a with and without peripheral and disc new general screening in all such patients,21 but in vessels has been well described in a small younger patients it may be very valuable to number of patients with SLE8.20 and this look for clotting screen abnormalities particu­ patient shows a similar pattern. Variable larly if there is a history of thrombosis at other hypertension is frequently seen in both SLE sites or recurrent abortion. In the presence of and in Sneddon's syndrome but is not a pre­ any of these features, a search for specific requisite for the development of vaso-occlu- anti-phospholipid antibodies is justified. VASO-OCCLUSIVE RETINOPATHY IN THE PRIMARY ANTI-PHOSPHOLIPID ANTIBODY SYNDROME 55

We are grateful to Dr. M. J. Walport PhD MRCP, , Jabs DA. Fine SL, Hochberg MC, Newman SA, Department of Rheumatology, Hammersmith Hospi­ Heiner GG. Stevens MB. Severe retinal vaso­ tal (case 1), Dr. Makin MRCP. Department of Hae­ occlusive disease in systemic lupus erythematosis. matology, Middlesex Hospital and Mr. A. M. P. Arch Ophthalmol 1986, 104: 558-63. Hamilton FRCS, FCOphth, Medical Retinal Clinic, " Asherson RA. Merry P, Acheson JF, Harris EN, Moorfields Eye Hospital (case 2), for permission to Hughes GRY. Antiphospholipid antibodies: a report these cases and for advice on the preparation of risk factor for occlusive ocular vascular disease in this report. systemic lupus eythematosis and the 'primary' antiphospholipid antibody syndrome. Ann Rheu­ mat Dis 1989,48: 358-61. III We are also grateful to the Departments of Medical Jonas J, Kolble K, Volcker HE, Kaiden JR. Central Photography, Western Ophthalmic Hospital and retinal artery occlusion in Sneddon's disease asso­ Moorfields Eye Hospital for preparing the ciated with anti-phospholipid antibodies. Am J illustrations. Ophthalmol 1986, 102: 37-40. II Elias M and Eldor A. Thromboembolism in patients with 'lupus'-type circulating anticoagulant. Arch Key words: Anti-Cardiolipin antibodies/Lupus Anti­ Intern Med 1984, 144: 510-5. coagulant/Primary Anti-Phospholipid Antibody syn­ 12 Pulido JS, Ward LM. Fishman GA. Anti-phospho­ drome/Retinal vein occulusion/Vaso-occlusive lipid antibodies associated with retinal vascular retinopathy. disease. Retina 1987,7: 215-18. 13 Levine SR, Crofts JW, Lesser R, Floberg J, Welch KMA. Visual symptoms associated with the pres­ ence of a lupus anticoagulant. Ophthalmol 1988, References 95(5), 686---92. 14 Levine SR and Welch KM. The spectrum of neu­ 1 Mackworth-Young CG, Loizou S, Walport MJ. rologic disease associated with anti-phospholipid Primary antiphospholipid antibody syndrome; antibodies, lupus anticoagulant and anti-cardioli­ features of patients with raised anticardiolipin pin antibodies Arch Neuro11987. 44: 876---83. antibodies and no other disorder. Ann Rheumatic 15 Graham EM, Spalton OJ, Barnard RO, Garner A, Dis 1989,48: 362-7. Ross Russell RW. Cerebral and retinal vascular 2 Hughes GRV, Asherson RA. Khamashita MA. changes in systemic lupus erythematosis. Oph­ Antiphospholipid syndrome: linking many thalmol1985. 92(3), 444-8. specialities. Ann Rheumat Dis 1989,48: 355-6. Ib Bonfa E. Golembek SJ, Kaufmal 0, et at. Associ­ 3 Harris EN, Gharavi AE, Hughes GRY. Anit-Phos­ ation between lupus psychosis and anti-ribosomal pholipid Antibodies. Clin Rheum Dis 1985,11(3): P protein antibodies. New Eng J Med 1987,317: 591-609. 265-71. 4 Asherson RA, Khamashata MA. Ordi-Ros J, et al. 17 Kleiner RC, Najarian LV, Schatten S, Jabs DA, Patz The Primary Anti-Phospholipid Antibody Syn­ A. Kaplan HJ. Vaso-occlusive Retinopathy asso­ drome-Major clinical and serological features.. ciated with anti-phospholipid antibodies (lupus Medicine (Baltimore) 1989,68: 366---74. anticoagulant retinopathy). Ophthalmol 1989, 5Feinstein DI and Rapaport SJ. Acquired inhibitors 96(6): 896---904. of blood coagulation. Prog Haemost. Thromb 1< Vicente DCG, Martinez JB, Lopez JR, Castellano J. 1972, 1: 75-95. Retinopathia proliferante y anticoagulante lup­ 6 Cariou R, Tobelem G, Soria C. Caen J. Inhibition of ico. Arch Soc Esp O/taI1988,55: 387-92. 1Y protein C activation by endothelial cells in the Gass JDM. Stereoscopic atlas of macular disease: presence of Lupus Anticoagulant. New Eng J diagnosis and treatment. CV Mosby, St Louis, Med 1986,314(18): 1193-4. 1987. vol 1,p340-53. 7 Boey ML, Colaco, Gharavi AE, Elkon KB, Loizou 211 Sanders MD. Retinal arteritis, retinal vasculitis and S, Hughes GRY. Thrombosis in systemic lupus autoimmune vasculitis. Eye 1987, 1: 441-65. erythematosis: striking association with the pres­ 21 Merry P, Acheson JF, Asherson RA, Hughes GRY. ence of circulating lupus anticoagulant. Br Med J Management of retinal vein occlusion. Br Med J 1983,287,1021-3. 1988,296: 294.