Article Text

Non-caseating granulomas as a cause of ischaemic retinal vasculitis
  1. H E PALMER,
  3. A C E McCARTNEY150,
  1. Ocular Immunology Group, Rayne Institute, St Thomas’s Hospital, London SE1 7EH
  1. Dr Palmer.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Editor,—Retinal vasculitis (RV) can be classified into ischaemic and non-ischaemic forms according to fluorescein angiographic findings: ischaemic RV is characterised by venous or capillary closure, or both. The aetiology of vascular closure in ischaemic RV is not clear, and histological evidence is scarce, but it is generally considered to be due to localised thrombosis. We present a case of ischaemic RV which on postmortem examination was found to be due to retinal vascular occlusion by non-caseating granulomas, consistent with a diagnosis of sarcoidosis.


A 52-year-old woman was referred with a 7 month history of painless misty vision with floaters. She was well and had attended the local ophthalmic unit, where a diagnosis of retinal vasculitis was made. She had no symptoms suggestive of a systemic inflammatory disorder. She was a heavy smoker, and her father had died from a myocardial infarction aged 59 years.

On examination at her first attendance to the St Thomas’s Medical Eye Unit, her visual acuities were 6/24, N12 in the right eye and 6/5, N4.5 in the left. Both anterior chambers were quiet, but there was low grade cellular infiltration in the vitreous bilaterally. Funduscopy revealed occasional peripheral intraretinal haemorrhages in the left eye, and in the right there was an oedematous macula, inferior intraretinal haemorrhages, and some sheathing of peripheral retinal vessels. Fluorescein angiography confirmed the presence of right cystoid macular oedema and bilateral peripheral retinal capillary closure (Fig1).

Figure 1

Fluorescein angiogram of the peripheral retina of the right eye showing capillary closure.

On investigation, she had a lymphopenia, but erythrocyte sedimentation rate, serum ACE, and chest x ray were within normal limits. On this basis, a diagnosis of isolated idiopathic ischaemic RV was made. She was commenced on oral prednisolone, starting at a dosage of 40 mg/day, which was tailored according to the disease activity and was then stopped over the following months, when the RV was quiet. In view of her personal and family histories, an underlying thrombophilia was investigated once she was off treatment: apart from a minimally raised fibrinogen level she had no haemostatic abnormalities. She had no further relapses of RV and never went on to have symptoms consistent with an underlying inflammatory disorder, so prednisolone treatment was never reinstigated. Eight months later she had a myocardial infarction and died.

Postmortem histological examination of the retina showed scattered non-caseating granulomas. There was evidence of these granulomas occluding the lumen of peripheral retinal vessels (Fig 2). There was no evidence of vasculitis or of granuloma formation in any major organs and the cause of death was confirmed as myocardial infarction secondary to arteriosclerosis.

Figure 2

Postmortem histological section showing a non-caseating granuloma occluding a retinal blood vessel (haematoxylin and eosin, × 30).


Retinal vasculitis is a disorder characterised by inflammation of retinal blood vessels and cells in the vitreous. It can occur secondary to infection or neoplasia, or in association with an underlying systemic inflammatory disorder such as Behçet’s disease or sarcoidosis, but is often idiopathic in origin.1

The presence of non-caseating granulomas around retinal blood vessels is highly suggestive of localised ocular sarcoidosis, although the granuloma formation could have been due to chronic inflammation from other causes such as persistent infection. While granulomas are not pathognomonic of sarcoidosis, their presence points towards such a diagnosis and cases of saroidosis confined to the eye or brain have been reported.23 This patient had no symptoms suggestive of an underlying inflammatory condition, which ties in with the absence of vasculitis or granulomas in the major organs on post mortem. In retrospect, it would have been valuable to have performed further investigations, such as a Mantoux and bronchial lavage to prove or exclude sarcoidosis, but owing to the absence of systemic symptoms this diagnosis was not strongly suspected. However, the presence of a lymphopenia was, with hindsight, a useful clue and indicates that sarcoidosis should be considered in a patient presenting with RV at any age.

This patient had the ischaemic form of RV in the absence of a thrombophilia. While sarcoidosis is generally associated with non-ischaemic RV,4 cases with retinal vascular occlusion have been reported.5 These histology specimens show granuloma occluding the retinal vessels, rather than the presence of intravascular thombosis, indicating that granuloma formation may be the direct cause of ischaemia in some patients. It is also of interest that the RV had clinically become quiescent, yet the presence of postmortem retinal granulomas indicates that intraocular inflammation persisted.

This case highlights the necessity to maintain a high index of suspicion of sarcoidosis, and confirms the need for further ocular histological specimens from RV patients to be examined whenever possible.


This work was supported by St Thomas’s (Endowments) Committee and the Iris Fund for the Prevention of Blindness.



  • Dr McCartney sadly died during the writing of this case report, following a long illness.